1° Canadian Bulletin of Fisheries and Aquatic Sciences

DFO Library / MPO Bib iothèque 11R191181M

Bulletin 210 Ottawa 1981

Fee Fishenes Pêches et and Oceans Oceans CanadI ?\D3 m.ps3 eal0

FishiirLs & Ozeans -e:- LIBRARY

MAR 10 198?

BIELIOTHÈQUE Pêchc:s & Océans Physiological Bases of Phytoplankton Ecology The Canadian Bulletins of Fisheries and Aquatic Sciences are designed to interpret current knowledge in scientific fields pertinent to Canadian fisheries and aquatic environments. The Canadian Journal of Fisheries and Aquatic Sciences is published in annual volumes of monthly issues. Canadian Special Publications of Fisheries and Aquatic Sciences are issued periodically. These series are available from authorized bookstore agents and other bookstores, or you may send your prepaid order to the Canadian Government Publishing Centre, Supply and Services Canada, Hull, Que. K1A 0S9. Make cheques or money orders payable in Canadian funds to the Receiver General for Canada.

Director and Editor-in-chief J. WATSON, PH.D. of Scientific Information

Deputy Director and Editor JOHANNA M. REINHART, M.SC. D. G. COOK, PH.D. Assistant Editors LORRAINE C. SMITH, PH.D.

J. CAMP Production-Documentation G. J. NEVILLE B. I. PATTERSON

Department of Fisheries and Oceans Scientific Information and Publications Branch Ottawa, Canada K IA 0E6

Participants in the picture are W.G. Harrison, T. Malone, D. Smith, P. Syrett, B. Thake, J. Raven, J. Feuillade, D. Cushing, P. Holligan, J. McCarthy, V. Ittekkot, S. Weiler, J. Smith, I. Morris, L. Legendre, F. Morel, P. Franco, R. Dugdale, P. Harrison, G. Magazzu, T. Plan, L. Guglielmo, G. Carrada, G. Honsell, G. Socal, T. Thordardottir, D. Marino, C. Descolas-Gros, M. Marzocchi, S. Puiseux-Dao, A. Zingone, M. Modigh, J. Horner, B. Heimdal, T. Smayda, L. Lazzara, C. Tomas, J. Myers, D. Bonin, D. Blasco, J. Gostan, R. Barber, B. Irwin, A. Sournia, E. Sakshaug, M. Estrada, S. Maestrini, R. Jackson, G. Fogg, T. Sertorio, B. Prézelin, P. Falkowski, C. Videau, P. Wheeler, R. Eppley, N. Morel, S. Chisholm, M. Karydis, and L. Mazzella. Other participants, not in the picture, include G. Jacques, W. Gieskes, M. Brogueira, S. PaneIla, C. Andreoli. M. Moita, W. Admiral, H. Glover, J.-C. Therriault, J. Lenz, N. Carr, and E. Toselli. BULLETIN 210

Physiological Bases of Phytoplankton Ecology

TREVOR PLATT [ed.]

Department of Fisheries and Oceans Marine Ecology Laboratory Bedford Institute of Oceanography Dartmouth, Nova Scotia B2Y 4A2

Based on an Advanced Study Institute sponsored by NATO Scientific Affairs Division

DEPARTMENT OF FISHERIES AND OCEANS Ottawa 1981 © Minister of Supply and Services Canada 1981 Available from authorized bookstore agents and other bookstores, or you may send your prepaid order to the Canadian Government Publishing Centre, Supply and Services Canada, Hull, Que. K1A 059. Make cheques or money orders payable in Canadian funds to the Receiver General for Canada A deposit of this publication is also available for reference in public libraries across Canada Canada : $17.95 Catalog No. Fs94-210E Other countries : $21.55 ISBN 0-660-11089-X ISSN 0706-6503

Price subject to change without notice Ottawa

Printed in Canada by K.G. Campbell Corporation

Correct citation for this publication: PLATT , T. [CD.] 1981. Physiological bases of phytoplankton ecology. Can. Bull. Fish. Aquat. Sci. 210: 346 p.

Contents

Foreword viii

Abstract/Résumé ix

Light reactions in photosynthesis B. B. PRÉZELIN 1-43

Dark reactions of photosynthesis B. P. KREMER 44-54

Respiration and photorespiration J. A. RAVEN and J. BEARDALL 55-82

Photosynthesis products, physiological state, and phytoplankton growth I. MORRIS 83-102

Photosynthesis measurements on natural populations of phytoplankton: numerical analysis C. L. GALLEGOS and T. PLATT 103-112

Tracer kinetic analysis applied to problems in marine biology D. F. SMITH and S. M. J. HORNER 113-129

Cell-cycle events in unicellular algae S. PUISEUX-DAO 130-149

Temporal patterns of cell division in unicellular algae S. W. CHISHOLM 150-181

Nitrogen metabolism of microalgae P. J. SYRETT 182-210

The kinetics of nutrient utilization J. J. MCCARTHY 211-233

Adaptation of nutrient assimilation R. C. DUGDALE, B. H. JONES JR., J. J. MACISAAC, and J. J. GOERING 234-250

Relations between nutrient assimilation and growth in phytoplankton with a brief review of estimates of growth rate in the ocean R. W. EPPLEY 251-263

Competition among phytoplankton based on inorganic macronutrients S. Y. MAESTRINI and D. J. BONIN 264-278

Importance of organic nutrients for phytoplankton growth in natural environments: implications for algal species succession D. J. BONIN and S. Y. MAESTRINI 279-291

Some processes and physical factors that affect the ability of individual species of algae to compete for nutrient partition D. J. BONIN, S. Y. MAESTRINI, and J. W. LEFTLEY 292-309

The role of hormones and vitamins in species succession of phytoplankton D. J. BONIN, S. Y. MAESTRINI, and J. W. LEFTLEY 310-322

Allelopathic relationships between phytoplankton species S. Y. MAESTRINI and D. J. BONIN 323-338

Morphological bases of competition and succession A. SOURNIA 339-346

vii Foreword

This Bulletin is a compilation of the major contributions to an Advanced Study Institute, sponsored by NATO with important assistance from the University of Messina, held on the island of Lipari, Sicily, in October 1980. The aim of the Workshop was an exchange of ideas and results between field ecologists and laboratory physiologists working on phytoplankton. It was motivated by the belief that conventional phytoplankton ecology, as practised by biological ocea- nographers, has been slow to assimilate and exploit the progress made by laboratory physiologists and biochemists working on phytoplankton. If the science of the physiological ecology of phytoplankton is to develop rapidly and efficiently, it could profit from a meeting of the two sides. Those who participated, on both sides, found it a stimulating and memorable experience. The idea for the Workshop was first conceived in conversation with Richard Dugdale, and developed by an Organizing Committee including Giuseppe Magazzù, Ian Morris, and Alain Sournia. The magnificent local organization was made possible through the help of Letterio Guglielmo. We also received help, at key points, from Prof. Battaglia of Padova, Prof. Genovese of Messina, Drs N. J. Campbell and J. Watson of Ottawa, and the Mayor and Municipality of Lipari. To all those people, and particularly to the authors of the chapters, I am grateful for their contributions. Finally, it is a pleasure to thank my secretary, Mrs M. Landry, for her valued help at every stage of the organization. TREVOR PLATT Editor and Director of the Advanced Study Institute

viii Abstract

PLATT, T. [aD.] 1981. Physiological bases of phytoplankton ecology. Can. Bull. Fish. Agnat. Sci. 210: 346 p. This book is a collection of 18 essays aimed at the elucidation and exposition of the physiological first principles that underlie phytoplankton ecology. It is based on a successful Advanced Study Institute of which the object was to expose field phytoplankton ecologists to the most recent advances made by laboratory-based physiologists in their understanding of algal photosynthesis, metabolism, and growth, and also to review the physiological principles on which these advances are founded. A second theme is the attempt to incorporate this new knowledge into the interpretation of measure- ments made on natural assemblages of phytoplankton in the field. The subje,ct matter treated includes the light reactions of photosynthesis; the dark reactions of photosynthesis; respiration and photo- respiration; numerical analysis of photosynthesis experiments; application of radioactive tracer techniques to metabolic studies; dynamics of the cell cycle and synchrony; nitrogen metabolism; nutrient uptake kinetics; the relationship between assimilation and growth; adaptation of the metabolic parameters to environmental change; and the physiological and morphological bases of competition and succession. Key words: phytoplankton, ecology, photosynthesis, cell cycle, unicellular algae, growth rate, nutrient assimilation

Résumé

PLArr, T. [ED.] 1981. Physiological bases of phytoplankton ecology. Can. Bull. Fish. Agnat. Sci. 210: 346 p. Le présent ouvrage groupe 18 essais visant à la clarification et à l'exposition des premiers principes physiologiques soustendant l'écologie du phytoplancton. Il est le résultat d'un programme d'études avancées dont l'objectif était d'exposer les écologistes spécialistes du phytoplancton sur le terrain aux plus récents progrès accomplis par les physiologistes en laboratoire sur la photosynthèse, le méta- bolisme et la croissance des algues, et aussi de passer en revue les principes physiologiques sur lesquels reposent ces progrès. Comme deuxième thème, on traite des efforts entrepris en vue d'incorporer ces nouvelles connaissances dans l'interprétation des données recueillies sur des groupements naturels de phytoplancton sur le terrain. Parmi les sujets traités, on note : réactions de photosynthèse à la lumière; réactions de photosynthèse à l'obscurité; respiration et photorespiration; analyse numérique d'essais sur la photosynthèse; application des techniques de marquage par isotopes aux études de métabolisme; dynamique du cycle et du synchronisme cellulaire; métabolisme de l'azote; cinétique de l'assimilation des éléments nutritifs; relation entre assimilation et croissance; adaptation des paramètres métaboli- ques au changement du milieu; et fondements physiologiques et morphologiques de la concurrence et de la succession.

ix Light Reactions in Photosynthesis

BARBARA B. PRÉZELIN Marine Science Institute and Department of Biological Sciences University of California Santa Barbara, CA 93106, USA

Introduction Granum

As approaches to the study of primary produc- Thylakold tivity in algae advance, to include new knowledge about the photosynthetic processes regulating plant envelope growth, biological oceanographers must be impressed Stroma and perhaps feel deluged with the amount of photo- synthetic information now available. Here, an attempt is made to sort the molecular photosynthetic lit- Stroma erature and provide a summary that details present lamellae understanding of the light reactions in photosynthesis Grana lamellae and emphasizes their relevance to studies of algal growth. Because of the wealth of information availa- FIG. 1. Cut-away representation of a chloroplast to show three-dimensional structure. (By permission, Hall and Rao ble, many areas covered had to be simplified. First, 1977.) emphasis was placed on how the light reactions proceed at the molecular level, i.e. the processes of energy is channeled into photochemical reactions that light absorption, excitation energy transduction to the lead eventually to the formation of ATP and reduced photochemical reaction centers, and the generation of NADP, and which in turn are utilized in the bio- electrochemical energy and electron flow between chemical reactions incorporating CO 2 into sugars and two distinct phototraps. The associated reactions of proteins. The light and dark reactions of photo- water splitting, NADP reduction, and transmembrane synthesis are separated physically within the chloro- proton flux which lead eventually to the generation plast. All structural components of the light reactions of ATP, are also discussed. Once the functional are localized within the pigmented membranes, while architecture of the light reactions has been outlined, the soluble enzymatic components of the dark some regulatory aspects of the cellular processes of reactions are found in the aqueous matrix (stroma) photosynthesis on primary productivity are con- surrounding the exterior surface of the thylakoid. sidered. Examples are taken from algal literature Thylakoids have distinct sidedness, being made of two whenever possible, and special attention is given to double membranes joined at the edges to form an topics related to present-day lab/field techniques internal aqueous phase (the intrathylakoid space) which are routinely used or recently introduced to separated off from the stroma. Structural components biological oceanography, i.e. pigmentation, fluores- of the light reactions are arranged in an organized cence probes, photosynthetic unit concepts, and manner across the thylakoid, so that certain steps in the parameters of the photosynthesis—irradiance curve. In photochemistry occur on the inside of the membranes all, it is hoped that a useful framework of information (i.e. the water-splitting reactions) while others are is provided to researchers involved in studies of algal directed toward the exterior (i.e. ATP and reduced physiology and efforts to estimate rates of in situ NADP formation) (Hall and Rao 1978; Kirk and primary productivity. Ti lney-B as sett 1978). Nonphotosynthetic structures also are found in the chloroplast and are localized in the stroma. Plastid DNA, RNA, and ribosomes are present, Algal giving the chloroplast some autonomy in protein, lipid, and pigment synthesis. There are also pyrenoid Plants take their characteristic color from pig- bodies, i.e. spherical regions of high density con- mented membranes termed thylakoids and are found taining one or more enzymes of CO 2 fixation in or within distinct cellular organelles called chloroplasts closely associated with the chloroplasts of some algae. (Fig. 1). It is in this highly colored organelle that Unlike most algal groups, in higher plants/green photosynthesis occurs, i.e. where absorbed light algae the photosynthetic storage products (fixed

1 carbon) accumulate inside the chloroplast as starch not compartmentalized into chloroplasts. (The blue- grains in the stroma or as a shell around pyrenoid green algae are described more accurately as cyano- bodies if present. In addition there can be eyespots bacteria, being true procaryotes. However, as their in the chloroplasts of some flagellated algae and, photosynthetic machinery is similar to eucaryotic depending on the algal class, may be found in asso- algae and quite distinct from photosynthetic bacteria, ciation with the flagellum. Algae also contain other the former term is used here). Their photosynthetic osmiophilic globules besides eyespots although they membranes are formed from invaginations of the cell are not generally very predominate. Although they envelope and extend as single lamellar membranes have no direct function in photosynthesis, these throughout the cytoplasm. As a result, although the globules do appear to be storage pools for such intrathylakoid space is preserved, the stroma does not photosynthetic components as plastoquinone, vitamin exist in these organisms. Thus the components that K, and thylakoid membrane lipid precursors. would be segregated in the stroma of eucaryotic The chloroplast envelope, a semipermeable plants are part of the cytosol of blue-green algae limiting double membrane in green algae and higher and photosynthetic bacteria. plants, keeps the contents of the chloroplast intact The number of chloroplasts per cell varies and separate from other components of the cyto- widely in different cell types and between different plasm. The outermost membrane is believed to plant groups, but generally they increase with cell function primarily in maintaining the structural size. Chloroplasts are most often found around the integrity of the chloroplast. A number of functions periphery of the cytoplasm, close to the cytoplasmic have been assigned to the inner membrane of the membranes. The particular characteristics of chlo- chloroplast envelope, including the control of solute roplast shape and internal structure also varies widely, passage and synthesis of structural components re- but two observations can be described that often quired for the differentiation of the thylakoids (Bi- distinguish the chloroplasts of several algal groups salputra 1974; Kirk and Tilney-Bassett 1978). from those of higher plants or green algae. First, By comparison, the photosynthetic machinery unlike higher plants or red and green algae, the of blue-green algae and photosynthetic bacteria is , chrysophytes, diatoms, dinofla-

AG 2. Chloroplast m the contracted form from GonyaulaA cells fixed at 1800 ct., from a culture held in continuous light. Also shown are an area of the chloroplast containing DNA fibrils and chloroplast ribosomes in the interiamellar spaces. x 38 000. (By permission, Herman and Sweeney 1975.)

2 gellates, Euglenoids, Chloromonads, and brown B CI,! algae all have a third membrane surrounding the RhodospirIllum A existing double membranes of the chloroplast Fcar.-I (Bacferium) B çhl envelope. In some cases, this third membrane that lies outside the chloroplast envelope is referred to as the "chlbroplast endoplasmic reticulum," B Chi as connections between it and the endoplasmic reticulum of the cytoplasm have been observed Chi a in some groups. In addition, ribosomes have been noted on its outer surface in some species and connections with the nuclear envelope have been seen (Kirk and Tilney-Bassett 1978; Bisalputra Chi a 1974). Second, all algal chloroplasts usually contain Anacystis thylakoids that extend the full length of the organelle. (Blue - green) However, algal classes differ in their arrangement of thylakoids within the chloroplast. The red and blue-green algae have the simplest thylakoid arran- gement, with single thylakoids that lie separate and Phycaerythrin parallel to one another throughout the length of the Chl a stroma. algae have paired thylakoids, whereas most other algal groups have sets of threes. Such stacks of thylakoids, referred to as compound o Chl a Maripelta (Red) lamellae, can be seen as bands in cross sections of o chloroplasts (Fig. 2). An elaborate arrangement of _a thylakoids is found in green algae, similar in arran- gement to the thylakoids of higher plants. The com- Chi a çhi c FucoxanthIn Chr■° pound lamellae of green algae may have as many D as seven thylakoids. In some species of greens there Endarachne are clearly recognized regions of stacked thylakoids (Brown) (grana) interdispersed by regions of unstacked thylakoids traversing the stroma and connecting Chi a separate grana stacks to one another (Kirk and Tileny- Chi b Bassett 1978; Hall and Rao 1978). CI,! a

Ulva (Green)

Photosynthetic Pigments

Photosynthetic energy conversion is initiated Pleurochloris when the pigments of the thylakoid membranes absorb light energy. There are three classes of photosyn- thetic pigments: (chl), , and carotenoids. Of these, only chl a must be present 350 450 550 650 750 850 950 for photosynthesis to proceed. The other pigments Wavelength. nm serve a light-harvesting function for chi a, and FIG. 3. Absorption spectra made in vivo for the major expand the spectral range of visible light energy kinds of photosynthetic organisms. Spectrum A was that can be absorbed in the chloroplast and trans- measured by chromatophores of the nonsulfur purple ferred to chl a to drive the photochemical reactions bacterium, Rhodospirillwn spheroides. Other spectra were of photosynthesis. The differential distribution of the obtained with intact organisms. The deep-water red alga, various photosynthetic pigments into specific plant Maripelta rotata , grows in submarine canyons off California groups (Table 1) gives them characteristic whole at depths to 60 m. The approximate absorption maxima cell absorption properties, some of which are re- and, in some cases, the range of absorption of photosyn- thetic pigments are shown. (Car., carotenoid; P.C., phy- presented in Fig. 3. cocyanin.) (By permission, Fork 1977.)

3

TABLE 1. Distribution of photosynthetically active pigments in major plant groups.

Phyco- Phyco- Allophy- b-caro- Major Chl a Chl b Chl c 1 Chl c2 erythrin cyanin cocyanin tene xanthophylls

Blue-green algae + + + + + Myxoxanthin

Red algae + + + + + Lutein

Cryptomonad algae + + + + trace Alloxanthin

Dinoflagellates + + + Peridinin

Brown algae, diatoms + + + + Fucoxanthin

Chrysophytes + + + + Fucoxanthin

Green algae, higher plants + + + Lutein

CHLOROPHYLLS AND —PROTEIN COMPLEXES The chlorophylls include chl a, chi b, chl c 1 , and chi c 2 and give most plants their typical green color. They are characterized by their ring structure, where magnesium is chelated in the CH =CH2 center and liganded at four sites to pyrole nitrogen atoms (Fig. 4). These porphyrin rings are essen- tially planar complexes surrounded by dense clouds of pi-electrons. Polypeptides are believed to attach H3C through linkages with side groups of the planar ring. Through substitutions of different side groups on the ring, electronic states of the molecule are altered and give rise to the characteristic absorption properties of the different chlorophylls (Fig. 4 and H 3C—Ç IV 5). When chl a is freed from the chloroplast and membrane proteins by extraction with acetone, its typical absorption spectrum shows two peaks of approximately equal intensity in the red and blue region of the visible spectrum at 660 and 430 nm, R I R2 R3 respectively (Fig. 5). Chl b under the same condi- — tions has a blue Soret band near 453 nm, which is Chi o CH 3 CH2 CH 3 (H2C) 2 — PHYTOL about 2.85 times more intense than its red absorption peak at 643 nm (Seely 1966). Similarly, the two forms Chi b CHO CH2 CH 3 (H2C)2 — PHYTOL of chl c found in most phytoplankton species are Chl C—1 CH3 CH2CH3 HC=CH 2 —COOH primarily blue light absorption pigments, with the Soret band of chl c 2 at almost 10 times as intense as Chi C 2 CH3 CH=CH 2 HC=CH2— COOH the weak red absorption band around 630 nm (Jeffrey 1969). Thus, the accessory chlorophylls partially fill in the absorption window left by chi a, especially in the blue region of the visible spectrum. While the porphyrin ring gives chlorophylls their absorption properties, the long phytol chain FIG. 4. Structures of chlorophylls. (Redrawn from Seely of chl a and cil b molecules presumably provides 1966; Dougherty et al. 1966.)

4 ber and Barber 1979; Mark well et al. 1979). Specific chlorophyll—proteins are difficult to isolate intact or functionally unaltered because chlorophylls are not linked covalently to proteins and the complexes are located in the water-insoluble membranes of the chloroplast. However, recent efforts with detergent extraction techniques have been successful in isolating a variety of chl a—protein complexes from higher plants and now several groups of algae. The procedures are discussed in some detail in several review articles characterizing plant chloro- phyll proteins (Anderson 1975; Thornber and Alberte 1977; Thomber et al. 1977, 1979; Boardman et al. 1978; Thornber and Barber 1979). One of the best characterized chlorophyll- proteins is the P700—chl a—protein complex, repre- senting photosystem I reaction center and some of its immediate antenna chlorophyll. This complex must be 'present for the photochemical events of 400 500 600 700 photosynthesis to proceed and, therefore, is found WAVELENGTH (nm) in all oxygen-evolving photosynthetic plants (Brown et al. 1974). The P700—chi a—protein shows a char- FIG. 5. Absorption spectra of chlorophylls extracted in acteristic chl absorption peak centered around 675— acetone. (Redrawn and adapted from Hall and Rao 1977.) 677 nm (Fig. 7), which is due to the presence of light-harvesting (LH) chl a molecules specifically associated with each Ps I reaction center. The light a lipophilic side group important in the insertion of energy absorbed by these chl a I molecules is prefer- the pigments into the thylakoid membrane. Phytol is entially directed to the photochemical reaction center a hydrophobic terpenoid (CO0C 21,H3,3 ) attached of Ps I. The amount of light-harvesting chl a, and via an ester linkage to a propionic acid side group on b-carotene (the latter giving rise to the absorption ring IV. By comparison, neither type of chi c has a shoulder around 500 nm in Fig. 7) isolated with the phytol group attached at this position, but the acrylic P700—chl a—protein complex varies with the isolation side group is replaced by propionic acid. With no procedure, but usually the complexes have a chl phytol chain, chi c molecules are only 2/3 the size of a /P700 ratio between 20:1 and 40:1 and a b-caro- chi a or chi b (610 versus about 900 daltons) and tene/P700 ratio of 1:1 (Thornber et al. 1977). It has presumably should be more hydrophilic than either been suggested that b-carotenes function in the chl a or b . But chi c appears to be closely associated complex to protect P700 from photochemical dam- with detergent-solublized thylakoid components. age and that quinones, generally present, may func- The absorption properties of free chlorophylls tion in early photochemical events (Thornber and can be modified significantly when bound to mem- Alberte 1977). brane proteins and inserted into the thylakoid. In At the heart of the complex is P700, a chlo- vivo chlorophylls generally display spectral shifts rophyll dimer that provides the reaction center of of their long wavelength absorption peaks further photosystem I with physical, absorption, and fluo- into the red region of the visible spectrum by any- rescence properties different from antenna chloro- where from 5 to more than 40 nm (Stoke's shift). phyll (Katz et al. 1979). The long wavelength ab- Furthermore, the absorption band of chl a in vivo sorption properties of P700 result presumably from is comprised of several types of chl a—protein com- the chl—chl interactions in the dimer, secondarily plexes, each absorbing light energies at slightly influenced by the protein binding of the complex different wavelengths. Deconvolution of chl a ab- in a specific chl—membrane environment. Both the sorption spectra (Fig. 6) and fluorescence spectral reaction center of Ps Tin vivo and the isolated P700 analyses led to the hypothesis that there are at least complex are characterized by a photobleaching signal four universal forms of chl a in vivo, with low resolved around 700 nm (Fig. 8). This light or temperature (77K) absorption maxima at 662, 670, chemically induced oxidized-minus-reduced negative 677, and 684 nm. The existence of more than one signal at 700 nm arises from a charge separation spectral form of chl b and c has not been documented. reaction in the P700 dimer excited by absorbed light It now appears that virtually all functional chlo- energy and can be used to quantify the number of Ps I rophylls in vivo are conjugated with proteins (Thorn- reaction centers present in a sample.

5 VVavelength,nm 600 650 700 600 650 700 1 Iii1 i 1 I I 1 1 1 1 B CI90A CI89A Fr 2 Fr 1 (Grano) (Stroma)

N

Spinach r- to'N ----•-'--- ,---,----:': o ----

><----- kill t ....,. À - Error X10.67 X20.5I rà) ii iiià A,AT is, Fr o F f I f ,r o c CI878 D C188B o U, Fr 2g Fr 1g (Grana Syst. 2) (Grano Syst. 1 )

Spinach

o

...... ,------2- . ----- -1. . . _...... 11M....k.À. lall. Ille 4...,... X5.93 X8.89

170 165 160 155 150 145 170 165 160 155 150 145 Wavenumber,ce X10-3

FIG. 6. Curve analyses of the absorption spectra of fragments from spinach stroma and grana membranes at — I96°C. The error of fit at each point is shown on a scale below each curve with the designated magnification; the higher the magnification, the better the fit. (By permission, Brown et al. 1972.)

6 350 400 450 500 550 600 650 700 WAVELENGTH (nm)

FIG. 7. Room temperature (300K) absorption spectrum of the 1)71, — chl a — protein isolated from Glenodinium sp. (By permission, Prézelin and Alberte 1978.) Attempts to isolate a single chlorophyll—protein indicator of Ps II activity in higher plants and green complex representing the functional reaction center algae (cf. Butler 1977). Unfortunately, little or no of photosystem II (Ps II) have been more difficult, work is available on the use of this signal in the presumably because of susceptibility of the complex —chl and carotenoid—chl c—chl a pig- to degradation by ionic detergents. Whole cell ab- ment systems found in most aquatic plant groups. sorption studies have suggested that the reaction In studies of whole cell fluorescence excitation spectra center of Ps II contains a special chl a form, which and fluorescence induction curves (Murata et al. 1966; undergoes an absorption decrease at 680 mil when Krey and Govindjee 1966; Govindjee and Yang oxidized and, thus, the phototrap of Ps II was termed 1966; Murata 1969), it has been suggested that the P680. However, this signal is not used to quantify in vivo fluorescence maximum of Ps II is at 685 nm Ps II reaction centers, as it is not resolved simply (room temperature) and 685-695 nm (77K). The and occurs in a part of the visible spectrum where association of these particular chi a forms with the secondary optical signals from P700 and antenna antenna pigments of Ps II have been confirmed in chi a molecules would interfere (Fig. 8). recent studies on isolated Ps II particles (Satoh and It is not clear if P680 represents a chl a dimer Butler 1978; Satoh 1980). structure similar to that described for P700 (Katz The major light-harvesting component from et al. 1979) or a ligated chl a monomer whose higher plants and green algae has been isolated function as a Ps II phototrap is determined by its and the characterization of the chl a—chl b protein membrane environment (Fajer et al. 1979). Closely complexes from all chl b-containing plants are associated with P680, and perhaps reflecting a pri- becoming quite detailed. Functioning as antenna mary or secondary electron acceptor for Ps II, is an pigments for the two photosystems, these chl—pro- unidentified spectral component that undergoes a teins have been shown to contain approximately characteristic light-induced absorbance change in equal concentrations of both chl a and chl b (Fig. 9). the 550-nm region of the visible spectrum. Termed In addition some of higher plant/green algae C550, this optical signal is best resolved at low carotenoids are found associated with these com- temperature (77K) and has proven a strong diagnostic plexes, in particular b-carotene and lutein, as well

7 40

20 Plastocyanin 7 E 0 0 ..., ' ■ ..." 7 .- 2 ./.. E P430 N, Cytochrome f c . cu 5 P700 L;--- It, —40 o c.) c P 700v o N. '4011 —60 .—C 4–, X w —80E- ..-- Cytochrome f

-100 i I I .1 I _I 400 500 600 700 Wavelength, nm

FIG. 8. Oxidized-minus-reduced difference spectra for the electron transfer components P700, P430, cytochrome f, and plastocyanin. The difference spectra for P700, cytochrome f, and plastocyanin are produced on conversion by light from the reduced to the oxidized form. The negative P430 band results from the reduction of an unidentified compound. (By permission, Fork 1977.)

8 Light-Harvesting /b -Protein

672

400 500 600 700 Wavelength (nm) FIG. 9. Room temperature absorption spectrum of light-harvesting chi a/b—protein of green alga, Chlamydomonas reinhardii. (By permission, Thomber and Alberte 1977.) as trace amounts of galactolipids and phospholipids . fucoxanthin (Barrett and Anderson 1977; Anderson The most recent structural model for this complex and Barrett 1979). From dinoflagellates, where would suggest a chl a /chl b—protein monomer chl c always predominates over chl a in whole cells of molecular weight 29 500, comprised of a single (i.e. 1.4:1.0 molar chl c:chl a in low light cells), polypeptide apoprotein of molecular weight 24 000 a single small molecular weight (38 000-50 000) associated through noncovalent linkages to three chl a —chl c 2—protein complex has been isolated molecules each of chl a and chl b (cf. Thornber from Glenodinium sp. This complex contains nearly and Barber 1979). Apparent oligomers of the chl a- all the chl c in the cell, about 20% of the total chl a, chl b —protein complex have also been isolated, no peridinin, but some yellow xanthophylls (diadino- although the characterization of their supposed xanthin and dinoxanthin). Its spectral characteristics aggregate structure is incomplete (Thomber et al. are shown in Fig. 10, and illustrate how the pre- 1979; Markwell et al. 1979). dominance of chl c contributes to the blue light- In most marine phytoplankton, the major light- absorbing capabilities of these phytoplankters . The chl harvesting chlorophyll is chl c, not chl b. Little work c/chl a molar ratio approaches 5:1. The had been done until recently to elucidate the organiza- chl a —chl c2—protein complex of Glenodinium sp. tion of chl c in vivo. Problems arose when detergent- appears to be a functional analog of the monomer form solubilization techniques designed for the chl a- of the chl a —chl b—protein of higher plants/ green chl b pigments system of higher plants/green algae algae, functioning as major light-harvesting com- were applied directly to phytoplankton pigment ponents for the photosynthetic apparatus (Boczar et al. systems. The general result was the rapid dissociation 1979; Prézelin and Boczar 1980a, b; Boczar et al. of chl c and carotenoids from the thylakoid mem- 1980). branes (Prézelin 1976; Prézelin and Alberte 1978; Boczar et al. 1979, 1980; Prézelin and Boczar 1980a, PHYCOBILINS AND b). However, three chlorophyll aggregate structures containing chl a and chl c have been isolated from Red and blue-green algae have derived their brown algae, one of which was also enriched in names from the color of the photosynthetic pigments

9 R- B- R-phycoerythrin Allophycocyanin (APC C-phycocyanin C-phycoerythrin Cryptomonad Allophycocyanin (APC FIG. Cryptomonad chi a—chlc—protein permission, Boczaret 10

4 a RELATIV E ABSORBANCE Values O'hEocha (1976),BennettandSiegelman (1977), 3 PE 555 PE 544 PC 645 PC 630 PC 615 PE 568 E, phycoerythrobilin; 400 10. Roomtemperature are based 500 WAVELENGTH on data complex TABLE 2.Properties al. 1980.) I &B) II &III) PC, ;PU,; PV,phycoviolin. absorption derived mainlyfrom of Glenodinium 600 (nm) Phycobilin PC +PV PE + PE + PC +PV PC +PV PC + spectrum PC PC PC PE PE PE PE of PU PU algal biliproteins.a(AdaptedfromGlazer1977.) PE h sp. (By of the 700 Brooks and Ley 498, 540,565 Absorption (620), 650 545, 563 553, 615 583,645 maxima et 558,630 558,615 (nm) 654 615 565 568 555 544 al. (1977),GlazerZilinskas PE wavelength tional role cryptomonads, containing organisms(Stanier1974) blue-green algae.Cryptomonadalgaealsocontain dominant maximum absorbs lightatslightly complexes, easily isolated that makeup in the species wavelength absorbingphycobilin,with 550 nm spectrum between500 absorbs lightenergyover allophycocyanin (APC), are phycoerythrin by their red.) their photosyntheticapparatus.(However,some algae (MacColl by chl and PCbut Gantt (1973),Glazer All threebiliprotein The light-harvesting apparatus c, of and 650nm.Allophycocyanin color. Thered,blue,and are in the whichisalsopresent of APC centered at650nm. red accessory Fluorescence termed biliproteins, PC (PC645)or 673-680 maxima not red not APC, (nm) and the 635 575 635 it hasbeensuggestedthat 660 637 580 578 660 575 and as — — reds may light-harvesting components blue-green algae,withPEpre- (PE), phycocyanin water-soluble, pigment—protein Berns 1978). and pigments of and PC and Bryant(1975),O'Carra longer have and 570nm.Phycocyanin respectively. Phycoerythrin the a some blue-greenalgae types are broad 24 000-34000 30 800-35000 most perhaps evenmediated been replacedby Molecular wavelengths, between 273 000 224 000 290 000 290 000 226 000 145 000 105 000 and are weight et 27 30 800 37 300 predominant of range ofthevisible central purple biliproteins al. (1978). 800 in most phycobilin- these algae found these unusual (Table 2).In is distinguished component the (PC), and absorption the [(a [(a in Probable structure (a (a (a longest (a (a subunit a in the a long func- most — /n / P)3 /3)3 p in; — P)3 13 9 are are ):1 )3 of ] 2 :, As different groups of algae were analyzed, it became apparent that the spectral characteristics of PE Cyanophytan biliproteins - C phycoerythrin 562 and PCs varied widely (Fig. 11-13). At first, believ- C-phycocyonin Allophycocyonin ing that the different spectral types of biliproteins 615 650 followed a taxonomic pattern, prefixes were added to / designate various taxonomic groups (C, Cyano- / i •, I phyceae; R, Rhodophyceae; B, Bangiophyceae, a ! 1 subclass of lower red algae). Generally, PEs isolated

from the red algae have three major absorption maxi- 0.4 ma, whereas those isolated from blue-green and I i cryptomonads have only one. Similarly, PCs from red h 278 and cryptomonad algae have two absorption maxima 1 /i I 1 0.2 I between about 550 and 650 nm, whereas C—PC has a 1 single broad absorption maximum which varies be- '- 1 \ tween 615 and 620 nm depending on ionic strength. 1 300 "400 500 600 700 APCs can have one or two absorption maxima, with Wavelength (nm) the major peak centered around 650 nm. Initially, it was suggested that the large spectral differences FIG. 11. Biliproteins from blue-green algae: C-phy- observed in each of the individual biliprotein classes coerythrin from Phormidium persicimun (solid line), was the result of conformational stresses imposed on C-phycocyanin from Nostoc muscorum (broken line), the same chromophore by the specific protein en- allophycocyanin from Nostoc muscorum (dotted line). vironments in each of the different algal groups. (By permission, O'Carra and O'hEocha 1976.)

Rhodophytan biliproteins 0.8 R - phycoerythrin R - phycocyanin 568 A llophycocyanin 540 618 498 • 0.6 if / \ 650 ule : 0 553 1 1

_o 0.4

278 0.2 I 307

; ...•

I 300 400 500 600 700 Wavelength (nm ) FIG. 12. Biliproteins from the red alga Cerumium rubrum: R-phycoerythrin (solid line), R-phycocyanin (broken line), allophycocyanin (dotted line). (By permission, O'Carra and O'hEocha 1976.)

11 Cryptomonad biliproteins - Phycoerythrin type IL Phycocyanin type I 0.6 Phycocyanin type IIE

(.) o 0.4 o .0

0.2

300 400 500 600 700 Wavelength (nm) FIG. 13. Some cryptomonad biliproteins: phycoerythrin from Hemiselmis fugescens (solid line), phycocyanin from Hemisehnis virescens (broken line), phycocyanin from Croomonas sp. (dotted line), (By permission, O'Carra and O'hEocha 1976.)

Later, as the biliprotein were PHYCOERYTHROB ILIN chemically characterized, it became apparent that spectral differences were related to the fact that the composition of the chromophores was not always the same in similarly colored biliproteins isolated from l l different algal taxonomic groups (cf. reviews of CH3 CH CH 3 CH2 CH2 CH3 CH3 Cbl Glazer 1977; Bennett and Siegelman 1978; Gantt 1980). When the chromophore composition of four CH 3 CH 2 CH2 CH2 spectrally distinct PEs was analyzed, one representing C=0 C= 0 the PE of blue-greens and three isolated from cryp- tomonads, the chromophore was the same in all cases OH OH and contained a single phycobilin pigment, phy- coerythrobilin (Fig. 14). By contrast, two distinct PHYCOUROBIL IN phycobilin pigments were found in the chromophore of R—PE. In addition to phycoerythobilin, phycou- robilin was also isolated and shown to account for the short wavelength absorption maximum seen in R—PE I I CH3 912 CH3 CH2 CH CH3 CH3 CH2 of red algae, but not seen in the PEs isolated from the 2 I l I two other algal groups (Glazer 1977). CH 3 CH2 CH2 CH3 Similarly, , the C—PC chromophore contains a I I single type of phycobilin (phycocyanobilin), whereas C=0 C=0 both phycocyanobilin and phycoerythrobilin are I l OH OH found in the R—PC chromophore. The cryptomonad FIG. 14. (A) Structure of phycoerythrobilin. Phycocyano- PCs are also unique in their chromophore compo- is similar in structure, with the C2 F1:, group of the sition, containing phycocyanobilin in combination IV ring saturated to become QM,. (B) Proposed structure with phycoviolin (Table 2). By contrast, the chro- of phycourobilin. (Redrawn from O'Carra and O'hEocha mophore of allophycocyanins isolated from red and 1976.)

12 blue-green algae are all the same, containing only complexes (i.e. dimers and trimers in Table 2), the phycocyanobilin. Thus, the composition of the purple extinction coefficents of the phycobilins increase APCs and the blue C—PCs chromophores are iden- and the absorption and fluorescence emission spectra tical, and the difference in color presumably arises of the aggregate alters significantly. These aggre- from differences in the protein environment in which gation states of heterodimer biliproteins can be the phycobilin pigment resides. affected by several factors, including ionic strength, Two different polypeptides, alpha and beta, pH, and the concentration of the biliproteins (Chap- malce up the subunits of the apoprotein of biliproteins. man 1973; Gantt and Conti 1976; Glazer 1977; These subunits usually occur in a 1: 1 stoichiometry, Trench and Ronzio 1978). with at least one chromophore attached to each The biliproteins not only aggregate with like polypeptide. The primary structure of the apoproteins kinds of biliproteins, but these aggregates further can from different biliproteins has been examined and combine with different kinds of biliproteins to form it appears that the amino acid composition is highly large organized aggregate structures on the stromal conserved, regardless of the algal source (Chapman surface of the thylakoid membranes. Aggregates of 1973; O'Carra and O'hEocha 1976; Glazer 1978; PE, PC, and APC are termed in red Gantt 1980). The heterodimer structure of biliproteins and blue-green algae. They can be seen as granules is the alpha—beta configuration, weighing approxi- with diameters to 300 A on the exterior surface of the mately 30 000 and exemplified in the cryptomonad thylakoid membrane (Gantt et al. 1976; Glazer 1977) pigment system (Table 2). (Fig. 15). Light energy transduction occurs from most As the individual biliprotein polypeptides externally placed PE (fluorescing orange-yellow at aggregate to form the higher molecular biliprotein 580 nm) to PC (fluorescing its absorbed light energy

FIG. 15. Model of a as envisioned to exist in the red alga, Porphyridium cruentum. Allophycocyanin as a core (white stipples) is in the center and near the photosynthetic membrane. R-phycocyanin (black stipples) suffounds the allophycocyanin, and phycoerythrin (light gray) constitutes the outer layer (stroma surface). (By permission, Gantt et al. 1976.)

13 between 635 and 660 nm) to APC (fluorescing a deep increase their proportions under bright light con- red with a emission maximum between 660 nm and ditions and function as photoprotective screens or 680 nm) located on the thylakoid surface in contact effective chemical quenchers in photooxidative with chl a. It is from APC that light energy is trans- reactions (Isler 1971; Krinsky 1979). These yellow ferred from the phycobilin antenna pigments to the chl carotenoids may also serve as biosynthetic precursors a molecules making up the antenna nd phototraps of for the major light-harvesting carotenoids present the two photosystems. By contrast, the biliproteins of in photosynthetic algae. The role of carotenoids as cryptomonads are not aggregated into supermolecular photosynthetic light-harvesting components has been structures but are packed evenly on the internal sur- examined extensively only in a few of these pigment faces of thylakoid membranes, between the pairs of systems. The efficient transfer of light energy from photosynthetic membranes. lutein to chl a within a chl —chl b protein complex It can be concluded that the absorption properties has been suggested (Satoh and Butler 1978). How- of different biliproteins are influenced both by the ever, lutein and similar carotenoids have not been chromophore composition and by membrane pig- considered significant contributors to total light ab- ment—protein interactions at higher order organi- sorption in higher plants/green algae where chloro- zational levels. As the hydrophobic tetrapyroles be- phylls dominate the pigment system or in the red come protein-bound, they are shielded from the algae where the biliproteins predominate. external environment and are subject to influences of the internal apoprotein environment. The resulting protein-induced conformational changes account for /3-CAROTENE C40 H56 much of the spectral changes observed in bound phycobilins where different apoproteins modify the absorption spectrum in different ways. Interestingly, , these tetrapyroles only become fluorescent when they become protein bound, suggesting that the apoprotein LUTE IN C40 11 56 °2 also plays an important functional role in providing the OH optimal environment for effective energy transfer to occur. These spectral characteristics are further modified as the different biliproteins assemble to- HO gether to form supermolecular phycobilisomes (Chapman 1973; Gantt et al. 1977; Glazer 1977; Gartt C40 H5202 1980). ALLOXANTH IN OH

CAROTENOIDS AND CAROTENOID- CHLOROPHYLL—PROTEIN COMPLEXES HO The chemistry of carotenoids has been exten- C46 H66 07 sively reviewed (cf. Isler 1971; Goodwin 1976; Moss MYXOXANTHOPHYLL and Weedon 1976). The structure of carotenoids C6 H IA- can be generally described as tetraterpenes made of OH eight isoprenoid (branched five carbon chains) units. HO Carotenoids can be divided into two groups, de- pending on whether they are hydrocarbons (carotenes) C42 H58 06 FUCOXANTHIN or contain some oxygen molecules (xanthophylls) OH (Fig. 16). These pigments can be synthesized de Or novo only by plants and photosynthetic bacteria. •-;" Animals derive their carotenoid pigmentation by grazing on these food sources and altering the chem- CH3C00 OH ical structure of ingested carotenoids, usually through C39 HS O07 the oxidizing effects of their digestive enzymes. PERIDININ OH The carotenoids found in the photosynthetic tissues of plants can play several functional roles. Very small amounts of /3-carotene are found in all photosynthetic membranes, presumably with both a CH3C00' photoprotective role and a stabilizing influence for the highly reactive P700 complex of Ps I. Yellow carotenoids (about 20% total carotenoid content of FIG . 16. Structures of representative plant carotenoids. most phytoplankton algae) have been shown to (From Isler 1971.)

14 The major carotenoids of most phytoplankton variety of free-living and symbiotic dinoflagellates are fucoxanthin and peridinin, present in the carote- (Haidak et al. 1966; Haxo et al. 1976; Prézelin and noid—chl a—chl c pigment system of dinoflagellates, Haxo 1976; Siegelman et al. 1977; Chang and Trench diatoms, brown algae, and chrysophytes in amounts 1981; Meeson and Sweeney 1981). Characterized that exceed that of either chi a or chi c. These in detail in Glenodinann sp. and Gonyoulax polyedra , two carotenoids are similar in structure (Fig. 16), the PCPs of these organisms appear to contain all differ from most other carotenoids in their high the peridinin of the cell in a 4:1 molar ratio with oxygen content, and have absorption properties that chl a and are noncovalently attached to either one or make them excellent photoreceptors for the blue- two apoproteins with a total complex molecular green light that penetrates to depth in the ocean. weight around 35 000. The four peridinins of the Photosynthetic action spectra of oxygen evolution chromophore occur as two dimers around a monomer in phytoplankton have long indicated the effective of chi a (Fig. 17). Presumably it is this combination of role peridinin and fucoxanthin play in gathering dimerization and protein-binding environment that light energy for photosynthesis (Tanada 1951; Haxo accounts for the absorption properties of peridinin 1960; Prézelin et al. 1976). within the pigment—protein complex (Fig. 18). Like The protein binding of fucoxanthin was estab- the biliproteins, PCP is a water-soluble complex with lished by Margulies (1970), but attempts to isolate its chromophore buried inside a hydrophobic pocket of a fucoxanthin-protein by higher plant extraction the apoprotein (Song et al. 1976). Unlike the bili- techniques have been generally unsuccessful (Mann proteins, PCP contains a single chl a molecule in its and Myers 1968). Barrett and Anderson (1977 , 1980) chromophore which receives light energy from peri- and Berkaloff et al. (1980) recently isolated a chl dinin within the chromophore with 100% efficiency a—chl c fraction enriched in fucoxanthin, but the (Prézelin and Haxo 1976; Song et al. 1976). It is detailed structural information on the organization of tempting to suggest that fucoxanthin, so structurally fucoxanthin and its binding to protein in these com- and functionally similar to peridinin, also might be plexes is not yet available. localized in a pigment—protein complex similar to There has been major success in isolating PCP. peridinin—chl a—protein (PCP) complexes from a

FIG. 17. A probable molecular arrangement of chlorophyll a and peridinins based on relative orientations of transition moments (double arrows of Qy fluorescence and B+ exciton transitions. (By permission, Song et al. 1976.)

15 400 500 600 700 nm FIG. 18. (A) Low temperature (77K) absorption spectrum of purified pl 7.3 PCP from Glenodinium sp. in 5 mM Tris pH 8.4 buffer. (B) Room temperature absorption spectrum of purified pl 7.3 PCP from Glenodinium sp. in 5 inM Tris pH 8.4 buffer. (By permission, Prézelin et al. 1976.)

THE PHOTOSYNTHETIC UNIT CONCEPT ferent algal groups would then reflect differences only in the pigment—protein complexes that make up their The chemical fractionation of chloroplasts light-harvesting components, i.e. the chl a —chl b has revealed different structural subunits for antenna protein complex in higher plants and green algae, the pigments. In addition, each phototrap is bound in composite of biliproteins in red, blue-green, and cryp- apparent protein complexes with discrete groups of tomonad algae, and the carotenoid—chi a—protein chl a molecules that absorb light energy for that complexes in association with chl a— chl c protein reaction center. Within Ps I there are an estimated and brown 40-120 chl a1 molecules bound in a protein complex complexes in diatoms, dinoflagellates , simplest functional with P700, as compared with an estimated 20-60 chl algae (Fig. 19). The minimal and comprised two photosystems a11 molecules associated with P680 in Ps II. A third chl composite structure, of and their light-harvesting components, has been a—protein complex , containing only chl a, also has termed a photosynthetic unit (PSU). This organization been hypothesized to link the reaction centers of Ps I implies a 1: 1 relationship between closely spaced Ps I and II, serving no photochemical function but regu- and II reaction centers, although recent experimenta- lating light energy distribution between the two photo- the case (Haehnel 1976; systems in the spillover process (Thornber et al. 1977; tion suggests this may not be Anderson 1980). Hayden and Hopkins 1977; Anderson et al. 1978; Kawamura et al. 1979; Miles et al. 1979; Anderson 1980). This chl a core is thought to be conservative in pigment composition FUNCTIONS OF THE PHOTOSYNTHETIC under most conditions and has been suggested to be the APPARATUS same in all organisms that contain chi a (Oquist 1974; Thornber et al. 1977; Vierling and Alberte 1980) (Fig. In the first steps of photosynthesis , light is 19). If so, differences in pigment composition of dif- absorbed by antenna pigments and transferred to the

16 hv hv

LIght - harvesting ; pigment -protein complexes

Light- harvesting Ohl o-protein

Vreachon FD—FRS P700- Chi o- center of protein ylotosystem / 116 :2e \ ; 2e - r 2e PC-- cyt f b

FIG. 19. Schematic model of the photosynthetic unit of chl a-containing plants. Organization of photosystems and light-harvesting components based on model proposed by Thornber et al. (1977). phototraps of the photochemical reaction centers. (Fig. 20; Table 3). In a well-coupled photosynthetic There, absorbed light energy drives electron flow from water molecules to NADP and is tightly coupled TABLE 3. Excited-states reaction scheme (by permission, to the formation of ATP. The reducing power of Malkin 1977). NADPH and the chemical potential of ATP are then Typical values* used in the carbon dioxide fixation reactions in the of the rate stroma. Although the photosynthetic events of the constants light-driven reactions are described most easily as (s- ') a sequence of events, it should be kept in mind that these are interdependent processes. For instance, M +hv (absorption of the nth — excited singlet state) there are possible exchanges of electromagnetic 'M,t 1m* (cascading to the lowest 10'2 energy between antenna serving different PSUs, as excited singlet by radia- well as electron flow between similar electron carriers tionless transitions) of different electron transport chains linking Ps II and kF 'Mt M + hv' (fluorescence) 109 Ps I reaction centers. Moreover, it is the combined kH photochemical events of all functional reaction centers 'Mt (radiation transitions) 101-10" that establishes the electrochemical gradients within Imr kT 341? the thylakoid membrane and are crucial to the regula- (transition to the lowest 102-10" excited triplet state by tion of overall photosynthetic events (Junge 1977; radiationless transition — Witt 1979). intersystem crossing) Light absorption and energy transduction — kp 'Mt (formation of photo- When light energy is absorbed into the molecular chemical products from structure of a pigment, electrons become redistributed the excited singlet state) into a set of excited singlet states (Fig. 20). The 3110, k'F absorption spectrum of a pigment molecule or a pig- M + hv" (phosphorescence) 1 k'H ment—protein complex reflects the light energy 'Mt (radiationless transition to 102-102 levels most effectively absorbed into the electronic the ground state [singlet] (10-1 -10-2 states characteristic of the chromophore structure intersystem crossing) at 77K) — k'p and conformation (cf. Clayton 1970; Sauer 1975; P' Malkin 1977; Knox 1977). The increased energy of the Mt —.- (photochemistry) excited singlet state can be dissipated in several ways

17

1 M3

I 3 1 M 2 I 1 1 1 1 1 ' M ,

"I I I I I 3* AE =h(vi-v") 1 1 I M i 1 I 1 I I 1 1 1 I I I I 1 I I 1 I I 1 1 I I 1 r■A _L_L Tt Singlet manifold Triplet manifold FIG. 20. Diagram of electronic state transitions for a typical polytomic milecule. M represents the ground-state (singlet). 'Me represents the n-excited singlet state. 'We represents the n-excited triplet state. Light absorption is mainly restricted to either singlet or triplet manifolds. Broken arrows indicate absorption; wavy arrows indicate nonradiative transitions; solid arrows indicate radiative transitions: fluorescence (frequency v'—photon energy hp') and phosphorescence (frequency v"—photon energy hv"). (By permission, Malkin 1977.) apparatus, most energy of the excited singlet state of most antenna pigments (Malkin 1977). However, antenna pigments is transduced between neighboring it has been suggested that the long-lived triplet state pigment molecules until it reaches the phototraps of may be important in the reaction center phototraps, the reaction center. The energy transfer process where excitation energy is quenched by primary usually is complete within a nanosecond, thereby electron acceptors (cf. Sauer 1975) (Fig. 20, Table competing successfully with dissipating processes 3). (Table 3). Triplet—triplet state conversions among excited A second excited state also can be produced state molecules also can occur and serve a photo- from the excited singlet state by nonradiative tran- protective function. In the presence of bright light, sition processes. Termed the triplet state, it is char- more energy is absorbed into excited singlet states acterized by decreased energy and a longer lifetime than can be used in the photochemical reactions. (10 ms) than the excited singlet state. The transition This situation seems relevant when algae with large from the singlet to triplet state involves the unpairing amounts of light-harvesting pigments (i.e. low light of an electron spin in an outer orbital of the pigment cells) experience a large and sudden increase in light molecule. The probability of a singlet to triplet intensities in their environment. Under these con- conversion and vice versa is considered small among ditions, even though some excess light energy is dis-

18 sipated through light emission and nonradiative decay, states has been described in some detail for those triplet states of antenna chl a can be formed and com- yellow xanthophylls found in most algal groups, bined with oxygen molecules to become chemically whose concentrations increase at high light intensities altered. A loss of chl molecules can result and seri- but not noted for their photosynthetic light-harvesting ously affect photosynthetic potential of the algae at capabilities (cf. Krinsky 1979). There is also evidence high light levels. One way to avoid the photooxidation that the light-harvesting peridinin molecules also of chi a molecules is to transfer quickly the energy of effectively protect the chi a monomer within the PCP the chl a triplet state to neighboring carotenoids , where chromophore from photodestruction (Song et al. the excitation energy is released as heat (Fig. 20, 21). 1980). This process of carotenoid quenching of chi a triplet

hv –500 antennae pigments

c oil ec t in g S-S Energy migration

—11 2 0

d iSs ipating T-T Energy migration

Pigmentsystem 1 Pigmentsystem 11

_ F,0 _ _ Ch1-4 NADP• n — — — — — — — 2

Î. t o s Cars —\/ s -- collecting Chl - a Chl a _ antennae S-S Energy e-N-1"\-.4 \-+ ( Chl -b) s) pig ment s migration echt -a

fluorescence heat dissipating - - T-T Energy T / migration C ar • 3e s Car' (Chi -b) Chl -a --/

REG. 21. Collection and dissipating energy migrations in the antennae pigments of photosynthesis. (By permission, Witt 1979.)

19 However, the dominant means of dissipating or alter energy transduction between the pigment mol- excess light energy from the excited states of photo- ecules and can enhance the dissipation processes. synthetic pigments is through luminescence. Light Charge separation and electron transport — emission is most easily observed in solutions of free Several excellent reviews have been written that chlorophylls, where coupled energy transfer, photo- thoroughly detail present knowledge of the primary chemistry, and membrane-mediated thermal tran- photochemical events (Sauer 1975; Malkin and Ke sitions are eliminated. However, whereas chlorophylls 1978) and electron transport reactions of photo- fluoresce, extracted carotenoids and phycobilins re- systems I and II (Trebst 1974; Bearden and Malkin moved from biliproteins do not fluoresce. Fluores- 1975; Avron 1975; Golbeck et al. 1977; Junge 1977; cence is the light-emitting process by which pigments Bolton 1977; Amesz 1977; Ke 1978; Knaff and in the excited singlet state can return to ground state if Malkin 1978; Crofts and Wood 1978; CIB A Foun- their excess energy is not funneled through other pro- dation Symposium 1979; Muhlethaler 1980; Velthuys cesses within the fluorescence lifetime of the molecule 1980). Presented here is a summary of the latest views (Fig. 20, Table 3). The transfer time between closely on charge separation and electron transport reactions. neighboring chl a molecules is estimated at less than a An attempt is made to highlight differences between picosecond, while the fluorescence lifetime of chl a is algae and higher plants and indicate where concurrent on the order of 5 ns. Fluorescence occurs only from the information on algal groups is lacking. lowest excited singlet state and, because of thermal The primary photochemical events in photo- relaxation prior to light emission, the fluorescence synthesis occur when reaction centers of Ps I and II maximum is a few nanometres longer than the longest are excited by absorbed light energy (Fig. 21, 22). absorption maximum of the chromophore. In vivo, the The charge separation reaction can be visualized majority ( > 90%) of fluorescence at room tempera- as the following: ture arises from Ps II, as Ps lis much more likely to dissipate excess light energy through nonradiative H 5 ps DPA Di (P *A 1 ) ■ Di (P +A 7) thermal conversion processes (Fig. 21). , 200 ps Light emission as the triplet state returns to D' (P +A DPAA ground state is also possible and is termed phos- where P is the photoactive P680 and P700 chl a phorescence. It occurs much later (about 1 s) and with pairs, and D and A represent electron donor and much less energy release than fluorescence (Table 3). acceptors, respectively. In the excited singlet state, A related process is delayed light emission, which both P680 and P700 eject an electron which is rapidly can only occur in vivo and is closely related to the transferred (5 ps) to a closely associated primary photochemical properties of Ps II and the energized electron acceptor (A). A conversion from the singlet state of the thylakoid membrane (cf. Malkin 1977). to a triplet state of the radical ion pair results and, It is a complex phenomenon, occurring over a time from the triplet state, electron transfer to secondary scale from 1 gs to 1 min, and involves a triplet-to- singlet reconversion with the subsequent fluorescence of light coming from the excited singlet state. Ch -or :volt Generally, , the light energy absorbed by antenna Chl-ae pigments reaches the phototraps with high effi- ciency, suggesting the mechanism(s) of energy trans- duction to the reaction center must be very rapid X compete successfully against the wasteful - piFd and hV processes mentioned above. The speed of electro- I X-?20: NADP + PG811 magnetic transfer between pigment molecules PG pool hVil depends on the strength of their resonant transition — 0 dipoles, which in turn depends on the distance be- Ese- tween the pigments and their relative orientation to one another (Junge 1977). For sufficient resonance Chl-a1 H20 to keep energy transfer times less than the fluo- (P700) rescence lifetime, chl a molecules need to be closer than 100 À. It appears that the proteins in the mem- Chl-a • brane and pigment complexes serve to optimally space (P680 ) chl molecules in vivo. Any membrane changes that alter the distances and orientation of pigment mole- cules so as to interfere with the strength and speed FIG. 22. Energy diagram of the electron transfer and mid- of their interaction will subsequently disrupt, weaken, point potentials. (By permission, Witt 1979b.)

20 acceptors occurs in about 200 ps . At this point, viously described as indicative of Ps II photochemical back reactions are less likely. Where recombinations events. do occur, delayed fluorescence from Ps II may be In Ps I of higher plants and some algae species, observed. Secondary electron donors and acceptors the primary electron donor is plastocyanin and the complete the charge separation and the reaction primary acceptor is possibly either a bound iron- center returns to its original condition through the sulfur protein or flavoprotein (Ke 1978; Croft and oxidation of electron donors. In both photosystems, Wood 1978). The acceptor was first identified on the primary electron donors lie near the inner surface the basis of spectral absorbance changes observed of the thylakoid, the reaction centers are buried in at 430 nm during the photochemical oxidation/ the thylakoid membrane, and the electron acceptors reduction of P700 and was accordingly termed P430. are found near the outer stoma surface of the thy- Following electron transfer from plastocyanin to lakoids. Thus, the photochemical events transfer one P700, the plastocyanin is rereduced by electron flow electron from each phototrap across the thylakoid directed at photosystem I through a series of trans- membrane. In this manner, light energy is stabilized membrane electron carriers from the acceptor side as stored chemical potential to be used in the for- of Ps II (Fig. 22, 23). From the acceptor side of mation of oxidizing and reducing compounds, trans- Ps I, subsequeM - oxidation/reduction reactions membrane electric fields, ion gradients, and mem- transport electrons to the final electron acceptor brane conformational changes. The universality of the NADP. charge separation reaction in diverse pigment systems The electron transport chain can be discussed of different plant groups has yet to be demonstrated, in three sections; the donor side of Ps II, the electron as most studies to date have been done with higher carriers between Ps I and II, and the acceptor side plants and green algae. However, it is believed that of Ps I (Fig. 22). The donor side of Ps II is least differences may pertain more to details of molecular characterized and it is here that oxygen evolution architecture and arrangement than to the fundamental and proton release from water occurs. Kok et al. nature of the charge separation process (Sauer 1981). (1970) observed oscillations in oxygen yield from In Ps II, the primary electron donor is water cells illuminated with a series of flashes. Four suc- and the acceptor is most probably a specialized cessive flashes, and thus four electron transfers to plastoquinone molecule, historically termed Q for P680, were required to produce a single oxygen quencher or X-320 for the spectral changes asso- molecule from two water molecules. The chemically ciated with the charge separation phenomenon uncharacterized water-splitting enzyme, known as (Amesz 1977; hinge 1977; Knaff and Malkin 1978). S, Z, E, M, or Y by different workers, very effi- Q is closely associated with the C550 signal pre- ciently donates electrons to P680 within 30 ns. To

NADP• H +

outside soÀ Plastoquinones X-320

se r- inside

Fm. 23. Preliminary topography of the molecular machinery of photosynthesis based on functional experiments. The two black "trunks" symbolize the two photoactive centers, consisting of chl a l and chl au , which probably are complexed with proteins. The porphyrin rings are located toward the inner surface. (By permission, Witt 1979a.)

21 form oxygen from water, the four electrons are the rate-limiting step in electron transport, setting extracted and stored in successive stages on the S the maximum quantum supply rate needed to drive complex, with the release of four protons to the electron flow (turnover time). As discussed by Crofts intrathylalcoid space. The work begun by Kok et al. and Wood (1978), PC lias been found in so many (1970) has shown the S complex to include a charge higher plants it is considered ubiquitous in all plant storage enzyme with four oxidation states (S states) groups. However, the presence of PC has not been where one electron is removed from one of two demonstrated in several algal species and the universal associated water molecules within 0.2 ms of each distribution of this electron carrier in photosynthetic flash (cf. Harriman and Barber 1979). All attempts algae has been questioned. Connected with these to isolate the S complex have been unsuccessful observations, an interesting suggestion has been although Spector and Winget (1980) have isolated made. One of the biggest differences known to occur a maganese-containing protein (molecular weight between higher plants/green algae and other algal 65 000) which appears to be involved in oxygen evolu- groups is the nature of another electron carrier, cytf. tion. It has been suggested that manganese is the Found in higher plants, cyt f is not always part of charge accumulator in the S complex, although cal- the electron transport chain. The cyt f of algae is cium and chloride also are knoWti to be involved in very abundant and known not to have the same oxygen evolution. spectral or molecular weight characteristics as that Localized between Ps I and Ps II are a series isolated from higher plants (Wood 1977). This algal of chemical electron carriers. There has been contro- cytf is also called cyt c-552 and is known to replace versy over the order and composition of electron the function of PC in Euglena and to be present with acceptors in the chain, but the most favored sim- PC in other algal species. It has been suggested that cyt plified pathway is shown in Fig. 22. Following the c-552 serves a dual role for PC, especially under primary acceptor, Q or X-320, electron transport conditions of copper limitation when the PC content of moves out of Ps- II via a second acceptor, R (or PQ,,,). plants declines. R is believed to be a pastoquionone (PQ ( ,)), a lipo- An analogous situation also appears to occur philic quinone with an isoprenoid side chain, bound on the acceptor side of Ps I. In some algae, the to a special protein. It is between Q and R that the primary electron acceptor, the iron—sulfur protein photosynthetic inhibitor 3—(3, 4-dichlorophenyI)-1, ferredoxin, is partially replaced by flavodoxin. As 1 dimethylurea (DCMU or diuron) acts to block it is able to substitute for almost all reactions involving electron flow. When R accepts electrons from Q ferredoxin, flavodoxin synthesis often is promoted it also takes up one proton from the stroma. However, by iron deficiency. It is suggested that ferredoxin since the addition or removal of two hydrogen atoms is prefeired in rich medium because it has better is needed to reduce or oxidize plastoquionones, reaction properties than flavodoxin. But, because R passes two pairs of electrons and protons on ferredoxin requires significant amounts of the cell's when subsequently oxidized by the neighboring iron supply, flavodoxin is a useful alternative for plastoquinone (PQ) pool. growth when iron is limiting (cf. Croft and Wood Perhaps the most important set of electron 1979). carriers is the PQ pool. Where the other electron Ferredoxin reduces NADP with the aid of ferre- carriers appear present in ratios of 1 or less per PSU, doxin-NADP reductase, a FAD-containing flavo- there are 5-10 PQ/electron chain. As such, the protein (Fig. 23). Cyclic electron flow around Ps I PQ pool innerconnects more than one electron trans- in vivo also originates from ferredoxin, which passes port chain, presumably providing alternate paths for electrons via cytochrome b -563 to PQ and back electron flow when electron 'acceptors in one chain down the noncyclic portion of the chain to P700. already are reduced. In this manner, PQ can provide Thus, during cyclic electron flow, additional protons an electron buffer in the transport systems and can are pumped to the intrathylakoid space. Cyclic regulate electron flow direction among different electron flow, leading to cyclic photophosphoryla- electron chains. In addition, as PQ is oxidized and tion, is stimulated when carbon dioxide is limiting. passes its electrons along to plastocyanin (PC) or cyt Reduced NADPH cannot be reoxidized when the c-552, two protons are released per molecules of PQ Cavin-Benson cycle activity is diminished and alter- to the intrathylakoid space. The proton release from nate electron acceptors from the acceptor side of both PQ oxidation and water-splitting events to the Ps I must be utilized. In addition to cyclic flow, interior phase of the thylakoid establishes the proton oxygen can also serve as a terminal electron acceptor gradient across the thylakoid membrane, the driving in a process described as pseudocyclic photophos- force for photophosphorylation (cf. Junge 1977; phorylation (cf. Gimmler 1977) (see Raven and Witt 1979). Beardall 1981). Cyclic electron flow also appears to The oxidation of reduced PQ by plastocyanin increase when Ps II activity is enhanced over Ps I, (a blue copper protein) is the slowest and, therefore, which can occur either when light energy distribution

22 is not balanced between the two photosystems or when the number of Ps II centers predominates over the number of Ps I centers. Photophosphorylation —The most accepted view of the coupling of electron transport, proton translocation, and ATP formation is based on Mit- chell's (1974) chemiosmotic hypothesis and can be summarized as follows. Photophosphorylation can occur only within intact membrane vesicles imper- meable to protons and with sufficient ADP and Pi available. Membrane potentials build up as a result of the primary photoact followed by the vectorial transport of protons (within 10 ms) to the interior of the vesicle (Fig. 23). Once a higher membrane potential is created by the proton flux, other ions can diffuse passively. Generally, chloride ions move in with protons and magnesium divalent ions move out. It is possible for most counterions present to pass through the thylakoid membrane to balance proton uptake. However, anion flow (i.e. C1- ) inward results in membrane swelling and volume changes that can lead to membrane conformational changes uncoupling ATP synthesis and photosyn- thetic light reactions. Such alterations of mem- FIG. 24. Thermodynamic relations in ATP synthesis by brane activity do not occur when cations (i.e. Mg") membrane-bound, reversible, vectorial ATPase according are pumped out (Avron 1981). to the chemiosmotic hypothesis. (By permission, Jagendorf A combination of the higher proton concen- 1977.) trations and net positive charge on the inside of the Membrane Localization of Photosystems thylakoid tends to provide a "proton motive force" across the membrane to drive the protons out. The magnitude of proton movement depends on the light In recent years, using techniques of freeze- intensity (the driving force), the pH gradient in steady fracture and freeze-etching, surfaces of the thylakoid state, and the internal buffer capacity of the mem- membranes have been exposed to reveal the presence brane. It appears that the internal proton concentration of particles embedded in or located on this mem- can reach magnitudes 10' times as great as the external brane (Fig. 25). Presumably proteinaceous in nature, proton concentration, with the pH of the intrathylakoid the particles are of different size and distribution space being as low as 4.0. It is estimated that between and are increasingly regarded to represent discrete 2.5 and 3.0 protons are pumped for each ADP phos- components of the photosynthetic apparatus. Detailed phorylated. It is the proton motive force that drives discussion of the freeze-fracture technique, descrip- ATP synthesis via a membrane-localized reversible tions of the substructure of the thylakoid membrane ATPase, which is similar in both chloroplasts and of green plants, and the molecular interpretations bacterial chromatophores (cf. Jagendorf 1977; Avron of observations are available in the review articles 1981). of Arntzen et al. (1977), Staehelin et al. (1977), Kirk and Tilney-Bassett (1978), Staehelin and The actual mechanism by which ATP is formed Arntzen (1979), Staehelin et al. (1980). The sub- from the dehydration of ADP and binding of Pi is structure of the thylakoid membranes of phycobilin- not known. One scheme is Fig. 24. Here, hydroxyl containing organisms has been discussed in articles and protons are released in ADP dehydration reac- by Gantt and Conti (1966), Lefort-Tran et al. (1973), tions, with the hydroxyl ions drawn toward the Gantt et al. (1977), and Gantt (1980). Until recently, proton-rich interior to combine with the protons to no clear freeze-fracture micrographs of chl c-con- form water. As a result the protons from the dehy- taining algal groups (browns, diatoms, dinoflagel- dration reaction are released in a stoichiometric 1:1 lates) have been obtained. A single study on the red ratio with each proton hydrated on the interior, and tide dinoflagellate, Gonyaulax polyedra, is now thus a net proton release across the thylakoid mem- available (Sweeney 1981). brane results (cf. Gimmler 1977). The reverse process is hypothesized to occur when ATP reserves are Freeze-fracture techniques (Fig. 25a, b) reveal mobilized in the stroma, to allow the hydrolysis both the inner fracture face (PF) and the outer fracture of ATP to ADP and Pi. face (EF) of the thylakoids. In green plants and algae

23 ss. • EFe Efu V"- _.• it-erattleeren

, e • ______ w . . • • • • • • • • emx...kv.,.eees.. . .

FIG. 25. (Top) Freeze-fractured isolated thylalcoids of spinach illustrate the four types of fracture faces typical for such specimens. The faces EFs and PFs belong to stacked membrane regions, faces EFu and PFu to unstacked ones. x 85 000. (Bottom) Illustration of how the membrane faces EFs, EFu, PFs, and PFu in top arise during the fracturing of thylakoid membranes. (By permission, Staehelin et al. 1980.) with recognized granum, a distinction is made be- (u). Thus, in higher plants, where grana stacks are tween thylalcoid surfaces of the grana and stoma most evident, four fracture faces (EFs, EFu, PFs, lamellae. Granum membranes are considered stacked PFu) are recognized (Fig. 25). (s) and stroma membranes are considered unstacked

24 The best characterized thylakoids are those of nificant number of these particles may represent higher plants and green algae. On the inner fracture Ps I reaction centers (Staehelin et al. 1980), although face of the granum thylakoid (EFs), there are char- the exact structural relationship is debated strongly. acteristic large particles up to 164 A in diameter. In addition, Staehelin (1976) found the EFs particles The largest of these particles are not found on similar in one membrane to line up with rows of PFs particles fracture faces (EFu) of the stroma lamellae. Armond in the other fracture face, suggesting the structurally et al. (1977) observed the large particles to increase close association of Ps I and II reaction centers. from a core diameter of 80 À in etiolated pea leaves Along with other findings, Staehelin et al. (1977) and to three large size-classes of particles (105, 132, Arntzen et al. (1977) used these observations to design and 164 À) in fully greened leaves. The only new a model for intramembrane particle association in polypeptide incorporated during this time was the thylakoid stacked regions of chloroplast membranes chl alb—protein of the light-harvesting component. (Fig. 26). Armond et al. (1977) suggested the inner fracture In electron micrographs of negatively stained face oparticles represent a Ps II reaction center core thylakoids, 100 À particles can be observed in large (80 A) in association with either 1, 2, or 4 aggregates number over the outer surface (cf. Kirk and Tilney- of the chi a /b—protein complex. These observations Bassett 1978). These particles have been identified are in accord with earlier studies of Sane et al. (1970) as the coupling factor for photophosphorylation. suggesting Ps II activity was absent from the stroma of Likewise, both biochemical and immunological higher plants. evidence is available to indicate the presence of the The outer fracture face (PF) is similar in both RUBP carboxylase as a 100-120 À particle on this the granum and stroma region, characterized by a same surface. These particles are only found on high concentration of smaller sized particles ranging external surfaces exposed to the stroma (Fig. 26). in size from 60 to 110 À. These particles extend This is in contrast with outer surfaces of thylakoids through the outer membrane surface slightly (Kirk closely aligned to other thylakoid membrane surfaces, and Tilney-Bassett 1978). It is believed that a sig- i.e. stacked regions of higher plant chloroplasts or • • • • • • • • 49exeremoggremormyedowebuyeedaes•• ee,

PS (80 EF particles) e, PS I, cytochrome.complexes,"free" Chl " a /b LH (?) (80 4 PFs and PFu particies)

el% PSI! + full cprriplement of Chl a/b e PS I +LH (?)(II5A PFu particles) LH (>140 A EFs particles) 0 coupling factor PSI! + partial copplement of Chi ribulose I,5 -diphosphate carboxylase A EFs and EFu a/b LH (<140 particles )

FIG. 26. Schematic illustration of the supramolecular organization of thylakoid membranes of higher plants and green algae. Note the different composition of the stacked (grana) and unstacked (stroma) membrane regions. This differen- tiation appears to result from the adhesion between LHC—PS II complexes in adjacent membranes, and the concomitant physical exclusion of components not directly associated with the electron—transport chain. (By permission, Staehelin and Amtzen 1979.)

25 compound lamellae of overall algal classes. Only Photosynthetic Unit Size and in the reds and blue-green algae, where thylakoids Density Determinations occur singularly, would a homogenous distribution of photosynthetic particles be expected along the As the discrete components of the two photo- external thylakoid surfaces. systems were identified and the concept of an organ- In red and blue-green algae, where thylakoids ized arrangement of photosynthetic particles in the occur singly so no distinction is made between contact thylakoid membrane developed, attempts were made and noncontact regions, a regular array of large to estimate the pigment cross section (size) and particles (120-170 À) can be observed on the ex- density of photosystem reaction centers or combined terior surfaces. They represent the phycobilisomes , PSUs in green plants. Such information is useful in whose arrangement may be ordered with underlying studies which characterize photosynthetic particles Ps II reaction centers (Fig. 27). As no light-harvesting and submolecular arrangements of thylakoid mem- chi a /b—protein complex is present in these organ- branes, assess the regulatory responses of the photo- isms, large particles on the inner fraction face are synthetic apparatus to environmental change (ie. light, not present. The overall distribution and sizes of nutrients, temperature, etc.), and even attempt predic- particles within the thylakoid membranes are other- tions of photosynthetic capacity. The measurement of wise similar to that of agranual chloroplasts of higher size or number of PSUs is not a direct measure of plants (Lefort-Tran et al. 1973). photosynthetic activity. However, because the PSU represents a probable minimal unit of photosynthesis, Very little detailed information on the molecular it has been reasoned and debated that the number of substructure of thylakoids in the major marine phyto- PSUs should be directly correlated with photosyn- plankton groups is available. However, the freeze- thetic capacity (Alberte et al. 1976a, b, 1977; Prézelin fracture study of G. polyedra does suggest some 1976; Terri et al. 1977; Armond and Mooney 1977; consistency of organization between the different Malkin et al. 1977; Prézelin and Sweeney 1978; Pré- algal groups (Fig. 28). Like other plant groups, zelin and Alberte 1978; Perry et al. 1980). One in- the PF faces show many more particles than the stance where this correlation has not held is with the EF faces (Sweeney 1981). Whereas the 160 À particle many phytoplankton species exhibiting a daily pe- correlated with the chl a /b—protein complexes is riodicity of photosynthesis in P max and in situ photo- absence from the EFs face, the physical location synthesis, Pi , which is independent of both pigmen- of the peripheral light-harvesting peridinin—chl a- tation and the shape of the photosynthesis-irradiance protein complex could not yet be identified (Swee- curves (Prézelin and Sweeney 1977; Prézelin and Ley ney). 1980; Harding et al. 1981). Under such circumstances,

otp B u, I

Cur' OFF 0 di e \- 1 (34‘5 n ra / 411111L/ - • s., ere e..4e.Qr e u_dja h frk t ,cl 10..be • - erre ve. • 4„.14c,:ed•, liwo/.06 lbee.iiniom ,relepirexe .‘

_

b6Onml Phycobilisomes

Flo. 27. Schematic representation of phycobilisome arrangement on thylakoids of red and blue-green algae, and the relationship to particle organization within the thylakoid membrane. (By permission, Lefort-Tran et al. 1973.)

26 FIG. 28. Thylakoid membrane faces of the chloroplast of Gonyaulax polyedra exposed by freeze-fracture. All 4 faces of the thylakoid are x 120 000. (Courtsey of Sweeney 1981.) the apparent PSU size and number stay relatively con- branes, and that measuring either the number of stant while the photosynthetic rate varies significantly Ps I or Ps II reaction centers should reflect accurately over the day. In these cases, knowing the density of the total number of functional PSU present. In higher photosystems alone would add little and even be mis- plants the numbers of Ps 1 and II reaction centers does leading in attempts to estimate daily rates of primary appear to be almost equal (Haehnel 1976). However, productivity. recent studies on blue-green algae indicate the Ps II:1 Most work to date has been based on the as- ratio is not always unity, ranging from 1.2 to 3.9 sumption that an equal number of both Ps I and II depending on the species and light intensity used in reaction centers occur in most chloroplast mem- culturing (Kawamura et al. 1979). Because only a few

27 studies have been completed and error associated with buffering capacity of the extract (Markwell et al. the different measuring techniques is undefined, it is 1980). In addition, if no detergent is used to solubilize not possible at this time to assess what impact such large chloroplast fragments, scattering in the sample results will have on the concepts of photosynthetic unit increases by an unknown amount, and increases determinations and their usefulness in productivity sample absorption and the apparent P700 signal estimates. (Markwell, personal communication). Also, samples Also subject to some debate is which technique have to be temperature-equilibrated or large spectral of measuring individual photosystem size and number shifts in the absorption spectrum result and interfere is most accurate in reflecting total PSU size (absorp- with the P700 signal. tive cross section) and number. In the past, the most Another major problem with the chemical assay commonly used methods have been estimates of P700 is that the chemical oxidants often cause irreversible content for Ps land oxygen flash yields for Ps II. These oxidations of antenna chlorophyll close to P700 and two procedures are outlined below, as they are techni- enhance the apparent P700 signal significantly. How- ques of choice for most biologists. They require less ever, a procedural method outlined by Markwell investment of time and/or instrumentation than more et al. (1980) and involving a spectrophotometer intricate procedures, which include fluorescence- coupled to a microprocessor has been successful in induction measurements (Malkin et al. 1977) and reducing the spectral interference of antenna chlo- plastoquinone and chl a flash absorbance measure- rophyll around P700. Similarly, the type of solu- ments (Haehnel 1976). bilizing detergent can enhance or diminish the P700 signal, presumably altering redox interactions of P700 ASSAY antenna chlorophylls (Markwell et al. 1980). P700 can be detected in plant samples either by The maintainance of uniform measuring condi- photochemical or chemical assays. Both approaches tions insures that the estimates of the relative change in are based on measuring the oxidized-minus-reduced P700 content is possible in the same algal species difference spectrum of P700 (Fig. 8) and applying under differing environmental conditions. However, it a differential molar extinction coefficient to quantify appears that great care must be taken to achieve a close the signal change induced at 700 nm. Coefficients estimate of the absolute number of Ps I reaction centers varying from 64 to 70 mequiv. cm have been re- present. Also, few independent measurements have ported for P700 (Hiyama and Ke 1972; Shiozawa et al. been made to assure that total exposure of P700 centers 1974; Ke 1978). PSU density is then expressed at is achieved when the same extraction procedures are P700/cell or area, and PSU size expressed as chl applied to plant groups with different membrane solu- a /P700 ratios. bilization properties. Thus, while comparison of rela- Light-induced oxidation/reduction changes tive changes in PSU size and density within a single can be measured in whole cell suspensions of many plant species is easily possible, it is not yet clear how algal species. However, P700 signals apparently much error may be involved when comparisons of cannot be detected in whole cells of all phytoplankton absolute PS I number or PSU size and density are (notably dinoflagellates , see Prézelin 1976; Govindjee attempted within and between different plant groups et al. 1979). It is not clear why this is the case, as (see Markwell et al. 1980; Perry et al. 1980). P700—chl a—protein complexes are known to be present (Prézelin and Alberte 1978). It is possible OXYGEN FLASH YIELDS that recombination of light-induced charges sepa- ration in Ps I occurs so quickly it is not detected in The number of Ps II reaction centers has been the assay procedure. The instrumentation required estimated in whole cell algal suspensions by illu- for the photochemical assay is not readily available minating them with consecutive brief flashes of to most biologists and so the chemical assay of P700 saturating light and determining the oxygen yield/ more commonly is employed. flash/chlorophyll. These measurements were first In the chemical assay, oxidants (i.e. potassium made by Emerson and Arnold (1932) and are based ferricyanide) and reductants (i.e. sodium ascorbate) on the concept that all reaction centers of a plant are used to generate an oxidized-minus-reduced exposed to a brief intense light flash will react only absorption spectrum of P700. However, several once (turnover) during the flash. In the work on precautions must be taken to insure an accurate Morelia, optimal oxygen yield was observed when determination of P700 content (Markwell et al. 1980). the flash duration was about 1 gs and the time between First, cell fractionation and detergent solubilization flashes greater than 40 ms. Thus, the average light of thylakoid membranes are required to insure expo- intensity is low and the dark period relatively long. sure of all P700 complexes to redox reagents. Extracts The dark time is required to pass on the products of whole plants not clarified by initial centrifugation from one flash and return all the reaction centers give poor results, presumably due to the redox- to a state that can fully utilize the energy of the next

28 light flash. This reasoning appeared confirmed when the photosynthetic characteristics at the whole plant it was determined by Emerson and co-workers that level. maximum flash yield was temperature independent There are several ways the photosynthetic and the dark time needed to complete the cycle of machinery may respond to external or endogenous necessary enzyme reactions was temperature de- environmental changes. The most obvious responses pendent. In Chlorella there are about 3000 chl a include alterations in (1) PSU size or the average molecules present for each oxygen molecule evolved amount of light-harvesting pigments per PS I/PS II (Emerson and Arnold 1932) and this PSU size does pair; (2) PSU or individual photosystem density, (3) not appear to vary with environmental conditions thylakoid membrane state leading to complete cou- (Myers and Graham 1971). pling/uncoupling of energy transduction and electron The process can be time consuming, as flash flow in discrete PSUs, and (4) reduction of dark yield as a function of both flash speed and intensity enzyme activity and/or rates of electron flow from one needs to be determined before the maximum flash electron carrier to another in Ps I and/or Ps II. A yield can be obtained. These parameters are not the change in any one of the above processes should have a same for different species of phytoplankton (Kawa- predictable effect on measureable photosynthetic mura et al. 1979). Also, significant respiration is pos- components and their activity, with possible con- sible during the relatively long dark period and may be sequences for the magnitude (i.e. photosynthetic po- accentuated over time by the relatively low average tential or capacity, P.„,) and shape (i.e. a, the light- light level the algae can absorb during the measure- limited slope; I = 1/2 Pin„ x , the light level at which ment time. Interference with oxygen yield measure- photosynthesis is half saturated) of the P—I relation- ments is accentuated in those marine phytoplankton ships. Each possibility is discussed briefly, with cited for which respiratory demands are proportionally examples of the environmental/endogenous changes much higher than those of green algae and higher that bring about these cellular modifications in dif- plants (i.e. diatoms, and especially dinoflagellates). ferent plant groups. There always exists the possibility Respiration changes presumably can be corrected for that more than one of the photosynthetic responses by knowing the dark oxygen consumption rates and mentioned might result as a single external variable assuming they are unchanged during the brief intense change, or that more subtle alterations in the photo- light flash. synthetic apparatus might occur (i.e. changes in the Ps II/Ps I ratio). Their detection would require much more detailed measurements than those outlined Regulation of Photosynthetic Cellular below. However, it is still evident that development Processes which Alter Photosynthesis- and usage of photosynthetic techniques designed to Irradiance Relationships characterize the photosynthetic apparatus can be incorporated into biological oceanographic techniques The preceding sections attempted to present to improve significantly the capabilities to both under- the current views on the structural and functional stand the mechanisms behind environmentally in- organization of the photosynthetic light reactions, duced physiological changes in phytoplankton and to but the emphasis now is shifted to highlight some use that knowledge to improve the ability to predict regulatory aspects of the cellular processes of photo- the direction and even time course of photosynthetic synthesis and their effect in altering rates of primary change during environmental fluctuations. productivity at the whole plant level. While certainly A) PSU size (the average number of light- not exhaustive, the following brief discussion is harvesting pigments per photosystem) changes can aimed at establishing that (i) discrete changes in be estimated by looking for changes in either chl al the organization and/or activity of the photosynthetic P700 or chl a/0 2 ratios. apparatus are induced by changes in environmental Should such analytical procedures not be avail- and endogenous factors, 'OD the biochemical and able, a change in PSU size (photosystem cross structural nature of the cellular changes can now section) can be inferred from disproportionate be identified in a fairly precise way by combining changes in cellular pigmentation. Since the chl a techniques which measure the amounts of various core of the photosynthetic reaction centers and their photosynthetic components (i.e. , pigmentation, P700, spillover component are assumed fairly uniform in and RUBP carboxylase) with the abundant infor- composition (Oquist 1974; Thornber et al. 1977; mation inherent in the photosynthesis—irradiance Vierling and Alberte 1980), changes in PSU size (P—I) relationships (cf. Bourdu and Prioui 1974), and are evidenced through changes in the amount of (iii) knowing the cellular response mechanism(s) to light-harvesting component (LHC) associated with environmental/endogenous variables considerably each PSU. Since the LHC of most marine plants advances the knowledge and ability to predict the (with the exception of green algae) is dominated by physiological consequences of the cellular changes on pigments other than chi a, a change in PSU size

29 curves similar to those in Fig. 29 are seen when bright- AFTER BEFORE BEFORE light cultures are low-light adapted. The changes occur within a generation time and are associated with major LH Chl a increases in pigmentation and specifically in associa- 0 0 tion with increased amounts of LHC, i.e. PCP and the chl a /chi c—protein complexes (Prézelin 1976; Prézelin and Sweeney 1978, 1979; Prézelin and VARIABLE IRRADIANCE —› data are not I PREDICTS Matlick 1980). Although all the necessary CHANGE available, pigmentation and P—I data suggest similar BEFORE AFTER low light responses may also occur in the blue-green LHC alga, Synechoccocus elogatus (Jorgensen 1969), the LH Chl diatom, Skeletoneina costattun (Brooks 1964), the AFTER green algae, D. tertioleeta, and to a lesser extent in Chlorella vannielli (Reger and Krauss 1970) and in p_ some seaweed (Ramus et al. 1967a, b). IRRADIANCE —> B) PSU (or photosystem) density changes should be seen in changes in both P700/cell and total 02 Fia. 29. Schematic representation of relationship between flash yields, while c hl a /P700 and chl a 10., remain altered photosynthetic unit size and changes in the P—I unaltered. Amounts of pigmentation would change, curves, expressed either on a cellular basis (upper curves) although pigment molar ratios do not, as the composi- or a chl a basis (lower curves). (See text for discussion.) tion of PSU is the same. PSU density changes should not affect PSU/chl a values (Fig. 30), therefore, no should lead to a change in whole cell pigment molar significant changes in the P—I chl a curves should be ratios. evident. But, since P, /cell should change in direct cell The predicted effect of altered PSU size is proportions to P700, the magnitude of the P—I/ illustrated in Fig. 29. When the P—I relationship curves should change in direct proportion to pigmenta- is expressed on a chl a basis, P,0„, would be expected tion. to change in direct proportion to the change in asso- Although not often documented, this strategy of a few ciated LHC chl a. In phycobilin systems, where no has been seen in photoadaptive responses , (Pré- chl a is in the LHC, changes in the P—I relationship dinoflagellates, notably Peridinhun chianti) would only be seen if one or more of the phycobilins zel in and Sweeney 1979) and Cerathunfitrca (Meeson photo- were used to standarized photosynthetic rates. On and Sweeney 1981), and suggested by the a cellular or area basis, P,„„, would not be expected synthetic data available for the green alga, Chia- to change as the density of PSU remained constant. mydomas inoewussi, and the xanthophyte, Monodus However, since the size of the LHC was altered, subterraneus (Jorgensen 1969, 1970). Changes in the light intensity required to saturate photosynthesis P700/ cell have been seen in light-induced responses should change in some inverse proportion to the in diatoms (Perry et al. 1980) and cotton (Patterson size of the PSU and be reflected in both slope and AFTER half-saturation constant changes.

Unfortunately, not all these measurements BEFORE have been made simultaneously in most studies of environmentally induced changes in photosynthesis. Increases in chl a /P700 ratios, with no alterations

in P700/cell or area ratios, have been observed when IRRADIANCE VARIABLE PREDICTS higher plants are grown under lowered light con- CHANGE ditions (Brown et al. 1974; Alberte et al. 1976a), BEFORE AND when corn is subjected to decreased water stress AFTER (Alberte et al. 1977), when the halophytic green LH Chl algae Dunaliella is released from hypotonic salt 00 stress (Brown et al. 1974), and when diatoms gen- erally are exposed to lowered light levels or increased AFTER nutrient supply (Perry et al. 1981). Decreased PSU IRRADIANCE size is commonly associated with aging and senes- FIG. 30. Schematic representation of relationship between cence in higher plants (Alberte 1981). In the majority altered photosynthetic unit density and changes in the of these examples the effects on photosynthetic P—I curves, expressed either on a cellular basis (upper characteristics is not available. However, in dino- curves) or a chl a basis (lower curves). (See text for dis- flagellates, changes in photosynthesis-irradiance cussion.) 30

et al. 1977), as well as in mutants of three hybrids -› BEFORE of corn (Terri et al. 1977). In all cases there was BEFORE CELL / good correlation between P700/ cell or area and CO, AFTER LH Chl a Pn,„,/ cell or area. An exception is a recent study • • THESIS in blue-greens, where P700/cell doubled at low Wil SUGAR light, phycocyanin concentration tripled, P,„„„ /cell TOSYN C stayed the same, and P. x /chl a declined (Vierling V VARIABLE IRRADIANCE —› and Alberte 1980). It appears to be an example of PREDICTS CHANGE simultaneous increase in Ps I density and LHC size BEFORE of Ps II at low light levels. AFTER C) Changing photosynthetic efficiency refers CO, / here to the coupling/uncoupling of energy trans- AFTER duction or electron flow between a discrete number of "4 SUGAR total photosystems. In other words, the energy flow through existing PSUs can be reversibly interrupted IRRADIANCE without disturbing the size and density of Ps l and Ps II per cell. In this case, no pigmentation changes are FIG. 32. Schematic representation of relationship between observed, but both light-limited and light-saturated altered photosynthetic enzymatic reactions and changes in rates of photosynthesis are dramatically altered (Fig. the P—I curves, expressed either on a cellular basis (upper for 31). This phenomenon appears basic to thermal liabil- curves) or a chl a basis (lower curves). (See text discussion.) ity in higher plants and algae (Bjorkman 1972), daily photosynthetic periodicity (often controlled by a bio- in proportion to decreased enzyme activity. logical clock) in dinoflagellates (Prézelin et al. 1977; decline been linked to low Prézelin and Sweeney 1977, diatoms (Prézelin and Such responses generally have plants and green algae Ley 1980; Harding et al. 1981), Euglena (Lonegran light responses in higher (Bjorkman 1972), carbon dioxide availability studies and Sargent 1978), and Acetabularia (Terbourgh and (Kelley et al. 1976), and Mcleod 1967), and life cycle changes in the photo- for several groups of plants unpublished synthetic activity of blue-green algae (Senger 1970a, aging in plants (Thimann 1980; Prézelin b). results). D) Changing enzymatic rates (rates of electron transport and carbon dioxide fixation) are detected in changes in specific enzyme activities and not in In Vivo Fluorescence as a pigmentation, PSU size or density, or light-limited Photosynthetic Probe rates of photosynthesis (Fig. 32). The major effect is on light-saturated rates of photosynthesis, which Here, an effort is made to introduce a few general types of fluorescence parameters routinely used in photosynthesis studies today. Some discussion of -> BEFORE regulating parameters is included to indicate com- BEFORE CELL signals which are LHC / plexities in interpreting fluorescence ChI IS information. The following is AFTER rich in photosynthetic 18 Q THES far from a complete review, as fluorescence studies in SYN photosynthesis have expanded considerably during the past 15 yr. Those wishing more detailed information IRRADIANCE —> VARIABLE are referred to the review articles of Papageorgiou CHANGE (1975), Lavorel and Etienne (1977), Butler (1977), the CIBA Foundation Symposium (1979), and the many AFTER articles on fluorescence in the Proceedings of the 5th International Congress on Photosynthesis (1981).

FLUORESCENCE INTENSITY + / — DCMU INDICES Fi G. 31. Schematic representation of relationship between altered photosynthetic energy transduction and changes Since the early 1960s, the fluorescence intensity in the P—I curves, expressed either on a cellular basis (upper of extracted chl a has been used as a routine field curves) or a chi a basis (lower curves). (See text for measurement reliably indicating the quantity of chlo- discussion.) rophylls and their breakdown products in dilute field

31 samples. With the advent of flow-through fluori- Fluorescence intensity (F) instantly monitors meters, in vivo chl a fluorescence profiles could be all the competing processes in primary photosynthetic measured easily. Assuming a reasonable correlation events. At room temperature, 90% or more of the between in vivo and extracted chl a fluorescence in- fluorescence intensity arises from back reactions of tensity, in vivo fluorescence profiles are now com- primary photochemical events occurring in the reac- monly used to indicate the spatial distribution of plant tion centers and antenna chlorophyll of photosystem biomass in aquatic habitats. This approach not only II. Membrane state changes, which reflect physio- assumed that the in vivo fluorescence intensity (F) was logical state changes and which alter F, are removed a consistent and accurate index of chi a content, but when fluorescence measurements are made at low that the chl a content of the sample also was an ac- temperature (i.e. in the presence of liquid nitrogen, curate and consistent index of plant standing crop. The at 77 K). For instance, the daily changes observed latter assumption was discredited early on, as widely in F/chl a and Fix.m achl a associated with circadian varying chl a/biomass ratios were observed under a rhythmicity in photosynthesis of phytoplankton are variety of changing environmental conditions. And abolished when measurements are done at low tempe- recently, the constancy of chl a fluorescence yield rature, suggesting the regulation of diurnal periodicity (F/chl a, referred to as R in some literature) also has of photosynthesis of many phytoplankton species is been challenged. As a result, it has been suggested closely related to endogenous regulatory changes in now that the addition of the photosynthetic inhibitor, the membrane state of the thylakoids (Prézelin and DCMU, would release in vivo F/chl a from all modi- Sweeney 1977; Govingjee et al. 1979; Sweeney et al. fying effects of the cell and thereby be a more true 1979; Prézelin and Ley 1980). Also, low temperature index of chl a concentrations (Slovceck and Hannon usually but not always increases F as fluorescence 1977). This is challenged also. While blocking elec- contributions from Ps I become more significant. For tron flow and thus increasing fluorescence, DCMU biological oceanographers working with in vivo does nothing to control the natural dynamics of the fluorescence measurements of natural phytoplankton thylakoid membrane which determine the changing populations, F may be considered a probe of Ps II physical/chemical environment in which chl a mole- activity, which can be strongly influenced by several cules reside. The modifying effects of a variety of variables. Summarized from the review of Lavorel parameters, including redox potential, ion flow, and Etienne (1977), fluorescence intensity can be membrane stacking, protein phosphorylation, and described generally by: other aspects of membrane state changes, on fluo- rescence yield have been documented in recent years (2) F = I • (Pf (It,t Xe., Xi', state) (Wraight and Crofts 1970; Telfer et al. 1976; Mills et al. 1976; Murata 1969; Barber 1979; Hamann 1979; where I = absorbed light energy, and cf. Proceedings of the 5th International Congress on Photosynthesis 1981). (3) yof — fluorescence yield = So, while fluorescence intensity indices may not kfl(kf + k„ + kd + ki) be reliable measures of chl a concentrations, changes in these indices often reflect physiological state where kf = fluorescence rate constant, and k„ = changes within the plant and, in some instances, corre- photochemical rate constant, and kd = nonradiative late well with changes in photosynthetic activity. For de-excitation rate constant, and k t = excitation trans- instance, changes in F/chl a have been associated with fer rate constant, and (pf depends on several factors light and nutrient stress (Kiefer 1973a; Loftus and affecting photosynthetic rate, including t = time Seliger 1975; Cullen and Renger 1979; Prézelin and of illumination, it = density of illumination, X, = Ley 1980) as well as with diurnal periodicity in photo- excitation wavelength, Àf = fluorescence emission synthesis (Prézelin and Sweeney 1977; Prézelin and wavelength, and state physiological conditions of Ley 1980; Vincent 1980). In some but not all cases, the sample due to pretreatment or growth status good correlations also have been recorded between (i.e. temperature, dark adaptatime, addition of various fluorescence indices and photosynthetic rates oxidants/reductants/inhibitors, light, and nutrient (Samuelsson and Oquist 1977; Samuelsson et al. 1978; status, etc.). Prézelin and Ley 1980; Kulandaivelu and Daniell If the light source (I) is kept constant, then 1980). To analyze critically the usefulness of these changes in detected F should reflect changes in indices in future phytoplankton studies, it is first çof . This is generally the case in fluorescence intensity necessary to review the parameters which comprise studies where the light source is the actinic beam of and regulate the measurement of in vivo fluorescence a fluorimeter, and is not the case where fluorescence intensity and, second, to then review the studies where is detected off the surface of the ocean by remote such indices have or have not been successfully em- sensing. When cells of a single type with identical ployed. pretreatment are placed in a fluorimeter and a constant

32 I used to detect fluorescence, it is then possible to fluorescence indices and photosynthetic activity are study changes in F that accurately reflect changes found. A few examples follow. in (pf . Such conditions have been used widely in Studies of diurnal periodicity in photosynthetic molecular studies to analyze the effects of various capacity (P,„„ x ) in laboratory cultures of dinoflagel- chemicals and environmental conditions (i.e. reduced lates and field samples of mixed diatom populations temperature, ion changes, nutrient stress, etc.) on the have shown that correlated changes in fluorescence photochemical events in photosynthesis. It should be indices F/chl a, Fix.mulchl a, and F/FDCNiu occur cautioned that while the uniform state of the same cells over the day (Prézelin and Sweeney 1977; Sweeney placed in such conditions can be assured, the same et al. 1979; Prézelin and Ley 1980). The mechanism state can not be assured in the same cells from different of regulation of the photosynthesis and fluorescence conditions or from different cells under the same con- periodicity appear closely linked, unrelated to pig- ditions (Papageorgiou 1975; Lavorel and Etienne mentation, and subject to regulation of the thylakoid 1977). membrane activity of a fixed number and fixed-size PSUs by a biological clock (Prézelin and Sweeney The light source (I) also photochemically alters 1977; Sweeney and Prézelin 1978; Sweeney et al. the F signal with time, giving the well-known fluo- 1979). The apparent inverse correlation between rescence induction curve. Generally, standard fluo- and P„,„, resulted from the larger changes in rimeters are not equipped to keep I small enough Fpcmu /chl a than in F/chl a occurring over the day (< 1 ,u,W cm-2 ) or illumination time short enough (Sweeney et al. 1979; Prézelin and Ley 1980). Thus, to keep (pf independent of /t, and, therefore, the time in the case of diurnal periodicity of photosynthesis , the at which measurements are made after the onset of F/Fpcmt: ratio does monitor accurately diurnal illumination become important in determining the changes in photosynthetic activity in these popula- intensity of the F signal. Equally important in de- tions, but does not reflect increased kJ- in the presence termining the intensity of F are the emission wave- of decreased k„ as was earlier predicted (Prézelin and lengths monitored and the excitation wavelengths Sweeney 1977). Instead, when k„ decreases at night used to promote fluorescence. apparently so does kf , suggesting added de-excitation When the photosynthetic inhibitor DCMU is energy is dissipated more through membrane pro- added to algal suspensions, an eventual block to cesses at night than during the day (Govindjee et al. electron flow between Ps II and I occurs and usually 1979; Sweeney et al. 1979). These membrane-asso- can be monitored as an increase in fluorescence ciated changes should be reflected in the rate constants intensity (Fp cmu ) to a new higher steady-state level. kd and k,, which unfortunately are not easily quan- (The time for the F1)(11 signal to maximize appears tified (cf. Papageorgiou 1975; Lavorel and Etienne to vary in different algal preparations, from less 1977; Butler 1977). than 30 s (Cullen and Renger 1979) to 30 min, and These studies reinforce the growing awareness may well reflect differences in membrane permea- that Fi cil a is not a constant that can be used reliably bility or sensitivity to the DCMU block (Prézelin in mixed phytoplankton biomass estimates based on and Ley 1980)). With k„ now presumably equal in vivo chl a fluorescence profiles (Kiefer 1973a, b; to zero, the Fix:mu signal then should reflect the total Loftus and Seliger 1975; Heany 1978). Likewise, light energy absorbed by the photosynthetic apparatus in situations where strong photosynthesis rhythms of Ps II when both kd and kl are considered small are present, the addition of DCMU to uncouple chl and unchanged. Under such conditions, if the fluo- fluorescence from photosynthesis may in fact enhance rescence yield of cil ((pf /chl) has not been altered the fluorescence rhythm. Lastly, it should be men- by the addition of membrane-bound DCMU and the tioned that the correlation between photosynthesis and photosynthetic activity of Ps I and II are generally fluorescence over the day decreased with light limita- equivalent, then the fluorescence ratios F/Fp cmi: tion and sometimes with the age of the cultures. Thus, should reflect that proportion of and I-F/Fn1u although good correlation may exist within any one set absorbed light energy which is reemitted as fluo- of measurements, the nature of the correlation or the resced light or used in the photochemical events of degree of the correlation do not appear to occur re- Ps II, respectively. If all assumptions hold true, producibly between different populations or the same then F/Fp c.mc and 1-F/Fpcmu , respectively, vary population under different conditions. Handling, nu- inversely and directly with the photosynthetic activity tritional and light state of the culture, detection of untreated algal suspensions. There are several procedure, and time of day all combine to affect the cases where one or more of the above assumptions fluorescence measurement. do not hold true and the reliability of the fluorescence Good correlation between fluorescence and indices is questioned (cf. Lavorel and Etienne 1977; photosynthesis also has been observed in aging Proc. 5th Int. Cong. Photosynth. 1981). However, cultures of green and blue-green algae (Samuelsson examples do exist where good correlation between and Oquist 1977; Samuelsson et al. 1978) and higher

33 plants (Kulandaivelu and Daniel! 1980). In the first state can alter the sensitivity of fluorescence yield case, the integrated signal of F/chl a and Fix NIL / to DCMU adds one more consideration to the inter- chi a within the first 5 s of illumination is determined pretation of fluorescence indices. and the calculated 1—F/F iwmc value is shown to be In conclusion, it appears that fluorescence strongly correlated with P., when compared in cells indices provide a wealth of photosynthetic infor- of different culture age. However, the components mation, only some of which can now be applied of this ratio are F/chl a and Fiwmt./chl a and they clearly to field studies. The fact that F/chl a and appear to follow the photosynthetic rate during Fix-mu /chi a vary widely under different conditions exponential growth and only become inversely related makes them poor indices for chl a or photosynthetic as the cells enter the stationary phase. Again, as with activity calculations. However, changes in these the daily periodicity studies, the fluorescence ratios in indices and their ratios, which are susceptible to the presence and absence of DCMU do appear to environmental regulation, make them possibly good reflect accurately but not predict changes in photo- indices of physical/chemical gradient changes synthetic activity. Furthermore, the studies on labora- between different water masses or physiological state tory cultures of green and blue-green algae indicated changes between different populations within the that the largest increases in DCMU-induced fluo- same water mass (Kiefer I973a; Loftus and Seliger rescence (Fm•Nic—F) occurred in cell suspensions 1975; Heany 1978; Cullen and Renger 1979; Vincent where photosynthetic activity was the highest. Similar 1980). But, before fluorescence indices of photo- observations were made by Rey (1978) and showed a synthetic rate can be used reliably, it appears neces- remarkable degree of correlation in studies of fresh- sary to improve our present very limited under- water and marine mixed phytoplankton populations. standing of in vivo fluorescence prperties, as well Together, these workers have suggested the possibility as to standardize the varied sampling and measuring of using "variable" fluorescence (here defined as techniques being used presently. FI)um u—F) as an indirect field measure of primary production and photosynthetic capacity. FLUORESCENCE INDUCTION CURVES The observation that DCMU does not always Fluorescence induction curves measure the increase fluorescence intensity (F) to the sanie degree time-varying fluorescence signal following the onset has been studied in some detail in the blue-green of constant illumination and provide information on alga Oscillatoria chalybea by Bader and Schmid the activity of Ps II and its interaction with Ps I (Fig. (1981). They recently reported that ammonium 33, 34). Letters (0, I, D, P, S, M, T) designate the sulfate-grown cells of O. chalybea exhibited DCMU- main features of fluorescence kinetic changes des- enhanced fluorescence in the presence of DCMU. cribed by the curves and often are used with the They argue that, unlike higher plants, the site of the assumption that each phase corresponds to specific DCMU block shifts in nitrate-grown algae from the aspects of the basic photochemistry associated with Ps acceptor side to the donor side of Ps II. That nutrient II (cf. Papageorgiou 1975; Lavorel and Etienne 1977).

FIG. 33. The various stages of the fluorescence time course of chi a invivo: (A) fast rise in isolated broken (class 11) spinach chloroplasts in the absence (lower curve) and presence (upper curve) of the electron transport inhibitor DCMU; (B) fast rise, and ensuing fast decay of chl a fluorescence in the red alga Porphyridium eruentum (redrawn from Mohanty et al. 197 lb); (C) slow fluorescence change in the green alga Chlorella pyrenoidosa. (By permission, Papa- georgiou 1975.)

34 r \ 1

I 50 1 I t

750nm I I

AT 40 TY SI 30

INTEN t ■

nce c - 196 ° I \ 20 t

resce il

o I 1 , Flu ' 10 ORESCENCE FLU

0 0 15 30 45 60 75 90 b - 70° \ TIME OF ILLUMINATION (seconds)

a - 25 ° • FIG. 34. Chl a fluorescence transients of G. polyedra ...... cultured on a light-dark cycle (LD, broken line) and 650 760 750 800 transferred to continuous dim illumination (LL, solid line); temperature, 22°C; dark adaptation, 10 min. (By permis- Wavelength (mp) sion, Govindjee et al. 1979.) FIG. 35. Fluorescence emission spectra of spinach chlo- roplasts. Excitation light, 475 nm. (A) 25°C; (b) -70°C; Usually, upon illumination with bright light (cf. (c)- - 196°C. Emission measured with a 5-nm half band Munday and Govindjee 1969; Mohanty and Govindjee width. (By permission, Murata et al. 1966b.) 1974; Papageorgiou 1975; Lavorel and Etienne 1977), chl a fluorescence intensity (1) rises from an initial Peru:Junin-Chi - level labeled "0" (for origin) to an intermediary peak Proteun ( PCP) labeled "I" (due to the reduction of the primary elec- tron acceptor of Ps II, Q, to Q- ), (2) declines to a dip, Light-harveshng "D" (due to interaction with Ps I which reoxidizes Q - Chl a/c - proteln 2680 back to Q), (3) rises to the peak "P" (due to filling up of the plastoquinone pool, thus preventing the oxidation of Q- to Q), (4) declines to a quasi steady state "S" (due to the conversion of Q - to a quenching but non- active form of Q, labeled Q', and possibly changes in the thylakoid membrane following changes in pH and/or spillover of energy from the strongly fluo- rescent Ps II to the weakly fluorescent Ps I), and (5) the "S MT" transient, where M refers to a maximum and T to a terminal steady state (due to changes in the thylakoid membrane caused by changes in photo- phosphorylation , redox potential, etc.). The time FIG. 36. A working model for the possible assignment courses for the fluorescence phases can vary consi- of absorption (A) and emission (F) maxima, in nm, to the derably with algal species and preconditions (Fig. 33, various chl a complexes in dinoflagellates. (By per- mission, Govindjee et al. 1979.) 34), but have been described to fall in the following ranges: IDP(-1 s), PS (5-10 s), SM (-0.5 min), MT phases of ihe fluorescence induction curve (the SMT (— 1.5 min) (Lavorel and Etienne 1977). stages run parallel), and argues strongly that the early Following suitable dark adaptation, the time stages of fluorescence induction accurately reflected course of light-induced 0 2 burst and evolution rates the photochemical activity of Ps II (cf. Joliot 1981). It were found to be antiparallel to the IDP and PS should be made clear that this observation (the

35 Kautsky effect) is distinct from the recent lab/field Ps II light (Lavorel and Etienne 1977). Many similar measurements of in vivo fluorescence intensity and its kinds of fluorescence spectroscopy studies of fluo- subsequent correlation to integrated steady-state rescence kinetics have been useful in examining photosynthetic rates determined by standard "C or 02 how excitation energy distribution among the two procedures. Thus, the Kautsky effect may offer little photosystems are affected by changes in endogenous/ insight into why supposedly similar fluorescence environmental variables. intensity measurements of phytoplankton are often completely out of phase with each other. For instance, instantaneous measurements by Kiefer (1973b) of FLUORESCENCE EMISS1ON/EXCITATION SPECTRA diatom populations fluorescence ran antiparallel to Fluorescence emission and excitation spectra, photosynthesis, while 30-s and 2-min measurements respectively, , provide information on the pres ■mce of F in similar mixed diatom populations were parallel of different fluorescence species of chl—proteins and antiparallel, respectively, to photosynthesis and biliproteins (carotenoid fluorescence is not (Prézelin and Ley 1980). Lab studies on the dino- detectable in vivo or in vitro) and on the direction and flagellate G. polyedra indicated all phases in the first efficiency of excitation energy transfer among these 90 s of the fluorescence induction curve were parallel pigment—protein complexes (cf. Goedheer 1972; to integrated 15-min measurements of photosynthetic Govindjee et al. 1979). In fluorescence emission rates (Sweeney et al. 1979). It appears that before measurements, fluorescence intensity (F) is shown to fluorescence induction data can be used reliably in the vary as a function of excitation wavelength ( X, ). The prediction of photosynthetic activity, , it will be neces- resulting fluorescence emission maxima (iv) are sary to (1) characterize fluorescence induction cha- characteristic of discrete fluorescing pigment com- racteristics and their variability in a wider range of plexes present in the sample (Fig. 35). By monitoring phytoplankton species (to date, G. polyedra is the only individual fluorescence emission maximum (X1) as a major marine phytoplankton species to be partially function of excitation wavelength (Xe ), spectral characterized), and (2) to improve techniques to components can be identified that best transfer their resolve and integrate fluorescence and photosynthesis absorbed light energy to the fluorescing pigment changes over a short period of time (however, see component. In this manner, the direction of excitation Samuelsson et al. 1978). energy flow between various pigment components can The addition of DCMU abolishes the IDP phase be mapped. When the absorption intensities of the of the fluorescence induction curve and often increases spectral components also are known, then transfer the overall F signal. Changes in the position and shape efficiencies also can be calculated. In this way, the of this phase have been shown to depend on the relative uses of fluorescence excitation spectra are similar to Ps II/Ps I rates of excitation (Munday and Govindjee those of oxygen action spectra, with the exception that 1969) and on the presence of HC0Ï in chloro- fluorescence measurements can examine the activity plasts (Stemler and Govindjee 1974). For instance, of a single or a mix of pigment components while cells of G. polyedra grown on a light—dark (LD) oxygen measurements are restricted to measuring the cycle had the same pigmentation but a distinct composite activity of all the components affecting Ps fluorescence induction curve from similar cells grown II. Finally, , studying the same fluorescent system under continuous light (LL) (Fig. 36). Among the under different conditions can provide information on changes was a shorter time needed to reach "P" in how changes in various stimuli alter the relative LL cells (3 s compared with 10 s), which suggested concentration, activity, , and direction of energy a higher ratio of Ps II/Ps I activity than LD cells, transfer flow between different fluorescence species due to either a more efficient use of absorbed quanta and supposedly influence associated photosynthetic in the initial transformations of Ps II or to a slower activities . interaction with Ps I (Govindjee et al. 1979). In higher plants and green algae, the room- The slow fluorescence phase "SMT" appears temperature emission spectrum is essentially a red- to be a "thermal" stage not linked directly to the shifted chl a fluorescence spectrum where accessory photochemistry of photosynthesis. However, it is pigments (chl b and carotenoids) contribute to chl a quite sensitive to conformational state changes of emission but their own fluorescence cannot be the thylakoid membrane and can provide some useful detected. The same appears to be true for the chl a- insights into photosynthetic regulation phenomena. chl c—carotenoid system of brown algae, diatoms, For instance, cations like Mg' induce conformational and dinoflagellates (Goedheer 1972). However, the changes in the membrane, which decrease the rate phycobilins present in red and blue-green algae of excitation spillover from Ps II to Ps I, visualized fluoresce in vivo, even though their transfer effi- in the fluorescence induction curve as an increase ciencies to chl a approached 100%. This difference in the S level. Likewise, the S level increases with can be used to detect even small amounts of phyco- preillumination with Ps I light but decreases with bilin-containing organisms in field samples (i.e.

36 Prézelin and Ley 1980). The fluorescence intensities et al. 1979). This information was used to construct of emitting pigments are usually strongly increased a working model for the possible assignment of as temperatures are lowered to liquid nitrogen tempe- absorption (A) and emission (F) maxima to various ratures (77 K), especially at the longer wavelengths chl a—protein complexes known to be present in associated with Ps I. The resultant band sharpening dinoflagellages (cf. Prézelin and Alberte 1978; Boczar and band shifts of emission bands at low temperature et al. 1980; Prézelin and Boczar 1981) (Fig. 36). It is make fluorescence excitation studies easier and are not known, but tempting to suggest, that a similar believed to result from reduced interactions between approach might be useful to interpret the organization different excited states and/or altered rates of energy and activity of photosynthetic pigment in most chl a- transfer between the various chl—protein complexes. chl c—carotenoid containing plants. Based on ap- Three- or four-peaked low-temperature emission proximate Stoke's shifts of 5-15 nm, with possible curves have been described for higher plants and red, absorption peaks used as subscripts and their possible green, and blue-green algae (cf. Papageorgiou 1975; emission peakes as superscripts, the following chl a Lavorel and Etienne 1977) and interpreted as follows spectral forms must be present in G. polyecira: (Fig. 35). Since chl a fluorescence at 685 and 695 Cil a (antenna chl a in peridinin—chl a—protein nm in the emission curve is sensitized mainly by complexes); Chi a : (antenna chl a, possibly wavelengths absorbed by chl h in higher plants and in chl a—cil c protein complexes; cil c is non- phycobilins in red and blue-green algae (determined fluorescent because it transfers energy with 100% from excitation spectrum analysis), these fluores- efficiency to this chl a invivo); Chl OW" (asso- cence maxima are assigned to two different chl a- ciated with Chl a 685 close to reaction center II); proteins closely associated with Ps II. Likewise, a and Chl aï,81;IM (associated with chl a 695, close broad emission band extending from 710 to 735 nm to reaction center I). Working with this model, it becomes highly fluorescent at low temperature with has been possible to follow a change in light energy peak emissions centered around 730 nm (F730). It distribution between the two photosystems when is more sensitized by chl a than by accessory pig- the light period is lengthened from 12 to 24 h (Go- ments and, therefore, is assigned to a chl a—protein vindjee et al. 1979) and to show little or no change complex closely associated with Ps I. It also has been in light energy distribution between the two photo- possible through fluorescence emission/excitation systems over the day when cells were cultured on studies to document changes in the excitation energy a 12:12 light—dark cycle of moderate light intensity distribution within the two photosystems induced (Sweeney et al. 1979). by changes in external stimuli of light color, tempe- rature, redox state, salinitY, etc. (Papageorgiou 1975; Lavorel and Etienne 1977; Ley 1981). Conclusion The situation is not the same in the chl a- chl c—carotenoid system of diatoms, dinoflagellates , Information on the molecular processes of and brown algae (Goedheer 1972; Govindjee et al. photosynthesis and the architecture of the photosyn- 1979). Only two of the low temperature emission thetic apparatus is accumulating at a rapid rate. The bands (F690-695 and F705-715) are evident in majority of research effort to date has been directed diatoms and browns (Goedheer 1972), while dinofla- toward understanding green plant photosynthesis. gellates exhibit only a single major emission peak at Where major algal groups have been studies , the 685-688 nm with a secondary band at 670 nm fundamental photosynthetic light-process appears the (Govindjee et al. 1979). Additional analysis of derived same, but with sufficient variation in details to prompt difference and ratio fluorescence spectra at low tempe- and encourage further detailed analyses of their photo- rature does suggest that long wavelength components chemical and fluorescence characteristics. From this of Ps I analogous to those of higher plants are present newer understanding of photosynthetic light pro- in dinoflagellates. However, they are not susceptible cesses have come suggestions for new probes of to low temperature uncoupling in the same manner as physiological state and/or photosynthetic rate (i.e. those of higher plants (Govindjee et al. 1979). PSU number or fluorescence intensity indices) which In spite of the simplicity of the emission spectra have been applied with varying success to phyto- presented for G. polyedra, a careful examination plankton productivity studies. Future studies into of the absorption spectra, emission spectra, difference the molecular basis of photosynthetic processes in emission spectra, excitation spectra of cil a fluo- phytoplankton promise new insights into how the rescence, and difference and ratio excitation spectra photosynthetic apparatus is organized and functions, together revealed that the dinoflagellate contained as well as perceives external stimuli, and then uses at least four chl a complexes (in addition to those these cues to effect a change in photosynthesis and in the two reaction centers which were so weakly perhaps other cellular processes regulating growth fluorescent they could not be observed) (Govindjee phenomena at the whole plant level.

37 Acknowledgments ANDERSON, J. M. 1975. The molecular organization of chloroplast thylakoids. Biochim. Biophys. Acta 416: 191-235. Dr R. Trench, Dr J. P. Thomber, and Dr B. M. 1980. Chlorophyll-protein complexes of higher Sweeney, as well as B. Boczar, are gratefully ac- plant thylakoids: distribution, stoichiometry and knowledged for their editorial assistance in reviewing organization in the photosynthetic unit. FEBS Lett. the manuscript content. Special appreciation is given 117: 327-331. to Deborah Rupp for her excellent secretarial assis- ANDERSON, J. M., AND J. BARRETT. 1979. Chlorophyll- tance. Financial Support was provided by funds protein complexes of brown algae: P700 reaction centre p. 81-96. In Chlo- from USDA research grant PL95-224 (BBP) and and light-harvesting complexes, rophyll organization and energy transfer in photo- NSF research grant OCE 78-13919 (BBP). synthesis. CIBA Found. Symp. 61, Excerpta Medica. ANDERSON, J. M., J. C. WALDRON, AND S. W. THORNE. 1978. Chlorophyll-protein complexes of spinach and Abbreviations barley thylakoids. Spectral characteristics of six com- plexes resolved by an improved electrophoretic proce- dure. FEBS Lett. 92: 227-233. APC allophycocyanin ARMOND, P. A., AND H. A. MOONEY. 1977. Correlation chl chlorophyll of photosynthetic unit size and density with photo- PC phycocyanin synthetic capacity. Carnegie Inst. Washington Yearb.

PCP peridinin-chl a - protein 77: 234-237. PE phycoerythrin ARNTZEN, C. J., P. A. ARMOND, J. M. BRIANTAIS, P-I photosynthesis-irradiance J. J. BURKE, AND W. P. NOVITZKY. 1977. Dynamic of the chlo- PS photosystem interactions among structural components roplast membrane, p. 316-337. In Chlorophyll-pro- photosynthetic unit PSU teins, reaction centers, and photosynthetic membranes. PQ plastoquinone Brookhaven Natl. Symp. 28. LHC light-harvesting component AvRoN, M. 1975. The electron transport chain in chlo- a light-limited slope roplasts, p. 374-388. In Govindjee [ed.] Bionergetics I = 1/2 Prn„, half-saturation constant for photo- of photosynthesis. Academic Press, New York. synthesis or the light level at which photo- 1981. Proton movement and transmembrane synthesis is half saturated electrochemical potential. 5th Int. Cong. Photosyn. 131 photosynthetic preformance or in situ rate of BADER, K. P., AND G.H. SCHMID. 1980. Dependence of II char- photosynthesis fluorescence behavior and other photosystern acteristics on the nature of the nitrogen source for the fluorescence intensity filamentous cyanobacterium. Abst. 5th lot. Cong. X wavelength Photosynthesis. Haliki, Greece. p. 33. BARBER, J. 1979. Energy transfer and its dependence on membrane properties, p. 283-304. In Chlorophyll organization and energy transfer in photosynthesis. References CIBA Found. Symp. 61, Excerpta Medica. BARRETT, J., AND J. M. ANDERSON. 1977. Thylakoid ALBERTE, R. S. 1981. Dynamics of the photosynthetic membrane fragments with different chlorophyll a, apparatus: prospects for genetic engineering and crop c, and fucoxanthin compositions isolated from chl improvement. Can. J. Genet. Cytol. (In press) the brown seaweed, Ecklonia mdiata . Plant Sci. Lett. ALBERTE, R. S., E. L. FISCUS, AND J. P. THORNBER. 9: 275-283. 1971. Water stress effects on the content and organi- 1980. The P-700 chlorophyll a-protein complex zation of chlorophyll in mesophyll and bundle sheath and two major light-harvesting complexes of Acro- chloroplasts of maize. Plant Physiol. 59: 351-353. carpi(' pan iculata and other brown seaweeds. Biochim. ALBERTE, R. S., ET AL. 1974. Comparison and activity Biophys. Acta 49(3): 309-323. of the photosynthetic apparatus in temperature-sensitive BEARDEN, A. J., AND R. MALKIN. 1975. Primary photo- mutants of higher plants. Proc. Natl. Acad. Sci. chemical reactions in chloroplast photosynthesis. USA 71: 2414-2418. Rev. Biophys. 7: 131-177. ALBERTE, R. S., J. D. HESKETH, AND J. A. KIRBY. BENNETT, A., AND H. W. SIEGELMAN. 1978. Bile pig- I976a. Comparisons of photosynthetic activity and ments of plants. In D. Dolphin [cd.] The . lamellar characteristics of virescent and normal grean Academic Press, New York. peanut leaves. Z. Pflanzenphysiol. 77: 152-159. BERKALOFF, C., J. C. DUVAL, H. JUPIN, F. CIIRISSO- ALBERTE, R. S., P. R. MCLURE, AND J. P. THORNBER. VERGIS, AND L. CARON. 1980. Spectroscopic studies 1976b. Photosynthesis in trees. Plant Physiol. 58: of isolated pigment protein complexes of some brown 341-344. algae thylakoids. Abst. 5th Int. Cong. Photosynthesis. AMESZ, J., AND L.N.M. DUYSENS. 1977. Primary and Haliki, Greece. p. 60. Associated Reactions of System II, p. 149-186. In BISALPUTRA, T. 1974. Plastids, p. 124-160. In W. D. P. Scientific Publication J. Barber [ed.] Primary processes Stewart [ed.] Algal physiology and biochemistry. Univ. of photosynthesis. ElsevieiCo., Amsterdam. California Press, Los Angeles. 38 BJORKMAN, 0. 1972. Photosynthetic adaptation to con- CROFTS, A. R., AND P. M. WOOD. 1978. Photosynthetic trasting light climates. Carnegie Inst. Plant Biol. p. 82. electron-transport chains of plants and bacteria and 1974. Thermal stability of the photosynthetic their role as proton pumps, p. 175-245. In D. P. apparatus in intact leaves. Carnegie Inst. Plant. Biol. Sanadi and L. P. Vernon [ed.] Current topics in p. 748-751. bioenergetics, Vol. 7. Photosynthesis: Part A. Aca- BOARDMAN, N. K., J. M. ANDERSON, AND D. J. GOOD- demic Press, New York. CHILD. 1978. Chlorophyll-protein complexes and CULLEN, J. J., AND E. H. RENGER. 1979. Continuous structure of nature and developing chloroplasts, p. measurement of the physiological state of natural 35-109. In D. Rao Sanadi and L. P. Vernon [ed.] phytoplankton populations as inferred from in vivo Current topics in bioenergetics. Academic Press, New fluorescence response. Mar. Biol. 53: 13-20. York, NY. DOUGHERTY, R. C., H. H. STRAIN, W. A. SVEC, R. A. BOCZAR, B. B. PRÉZELIN, J. MARKWELL, AND J. P. UPHAUS, AND J. J. KATZ. 1966. Structure of chlo- THORNBER. 1979. Evidence for the isolation of a rophyll c. J. Am. Chem. Soc. 88: 5037-5038. chlorophyll a--protein complex from a EMERSON, R., AND W. ARNOLD. 1932. The photo- dinotlagellate. Abst. Photochem. Photobiol. Asilomar, chemical reaction in photosynthesis. J. Gen. Physiol. CA. p. 80. 16: 191-205. 1980. A chlorophyll c-containing pigment-protein FAJER, J., C. K. CHANG, M. S. DAVIS, I. FUJITA , AND complex from the marine dinoflagellate, Glettodinium T. L. NETZEL. 1979. Primary donors and acceptors sp. FEBS Lett. 120: 243-247. of plant photosynthesis: in vitro studies. Abst. 7th BOLTON, J. R. 1977. Photosystem I photoreactions, p. Annu. Meet. Am. Soc. Photobiol. p. 83. 187-202. In J. Barber [ed.] Primary processes of FORK, D.C. 1977. Photosynthesis, p. 329-370. In K. C. photosynthesis. Elsevier Scientific Publications Co., Smith [ed.] The science of photobiology. Plenum/ Amsterdam. Rosette, New York, NY. BOURDU, R., AND J. L. PRIOUL. 1974. Réponses photo- GANTT, E. 1980. Structure and function of phycolilisomes: synthétiques de type adaptatif aux climats lumineux light harvesting pigment complexes in red and blue- de croissance: étude théorétique, applications et diag- green algae. Int. Rev. Cytol. 66: 45-80. nostic précoce. Physiol. Veg. 12: 35-51. GANTT, E., AND S. F. CONTI. 1966. BROOKS, J. E. 1964. Acclimation to light intensity in two location in algae. J. Cell Biol. 29: 423-434. species of marine phytoplankton, Dunaliella tertiolecta GANTT, E., C. A. LIPSCHULTZ, AND B. A. ZILINSKAS. Butcher and Skeletonetna (Greve) Cleve. M.S. Thesis. 1976. Further evidence for a phycobilisome model Univ. Southern California. 30 p. from selective dissociation, fluorescence emission, BROOKS, C., AND E. GANTT. 1973. Comparison of immunoprecipitation and electron microscopy. phycoerythrins (542, 566 nm) from cryptophycean Biochim. Biophys. Acta 430: 375-381. algae. Arch. Mikrobiol. 88: 193-204. 1977. Phycobilisomes in relation to the thylakoid BROWN, J. S., R. S. ALBERTE, AND J.P. THORNBER. membranes, p. 347-357. In Chlorophyll-proteins, 1974. Comparative studies on the occurrence and reaction centers, and photosynthetic membranes. spectral composition of chlorophyll-protein complexes Brookhaven Symp. 28. in a wide variety of plant material, p. 1951-1962. GIMMLER, H. 1977. Photophosphorylation in vivo, p. In 3rd Int. Congr. Photosyn. Res. 448-472. In A. Trebst and M. Avron [ed.] Photo- BROWN, J.S., R. A. GASANOV, AND C. S. FRENCH. synthesis I: photosynthetic electron transport and 1972. A comparative study of the forms of chlorophyll photophosphorylation. Springer-Verlag, Berlin. and photochemical activity of system 1 and system 2 GLAZER, A. N. 1977. Structure and molecular organization fractions form spinach and Dunaliela. Carnegie Inst. of the photosynthetic accessory pigments of cyano- Yearb. 72: 351-359. bacteria and red algae. Mol. Cell. Biochem. 18: BUTLER, W. L. 1977. Chlorophyll fluorescence: a probe 125-140. for electron transfer and energy transfer, p. 149-167. GLAZER, A. N., AND D. A. BRYANT. 1975. Allophy- In A. Trebst and M. Avron [ed.] Photosynthetic cocyanin B (X,,„,) 671, 618 nm, a new cyanobacterial electron transport and photophosphorylation. Springer- phycobiliprotein. Arch. Microbiol. 104: 15-22. Verlag, Berlin. GOEDHEER, J. C. 1972. Fluorescence in relation to photo- CHANG, S., AND R. TRENCH. 1981. Variation Of Sytn- synthesis. Annu. Rev. Plant Physiol. 23: 87-112. biodin juin ( = Gymnodittium) microadriaticum GOLDBECK, J. H., S. LIEN, AND A. SAN PIETRO. 1977. (Freudenthal): analysis of peridinin-chlorophyll Electron transport in chloroplasts, p. 94-116. In A. a-protein complexes. Proc. R. Soc. London (Sub- Trebst and M. Avron [ed.] Photosynthetic electron mitted) transport and photophosphorylation. Springer-Verlag, CHAPMAN, D. J. 1973. Biliproteins and bile pigments, Berlin. p. 162-185. In N. G. Carr and B. A. Whitton [cd.] GOODWIN, T. W. [ed.] 1976. Distribution of carotenoids, The biology of blue-green algae. Univ. California p. 225-262. In Chemistry and biochemistry of plant Press, Berkeley. pigments. Academic Press, London. CIBA Foundation Symposium 61. 1979. Chlorophyll organ- GOVINDJEE, D. WONG, B. B. PRÉZELIN, AND B. M. ization and energy transfer in photosynthesis. Excerpta SWEENEY. 1979. Chlorophyll a fluorescence of Medica: 374 p. Gonyaulax polyedra grown on a light-dark cycle after CLAYTON, R. K. 1970. Light and living matter, Vol. 1. transfer to constant light. Photochem. Photobiol. 30: The physical part. McGraw-Hill, New York, 148 p. 405-411.

39 GOVINDJEE, AND L. YANG. 1966. Structure of the red JUNGE, W. 1977. Physical aspects of light harvesting, fluorescence band in chloroplasts. J. Gen. Physiol. electron transport and electrochemical potential gen- 49: 763-780. eration in photosynthesis of green plants, p. 59-93. HAEFINEL, W. 1976. The ratio of the two light reactions In A. Trebst and M. Avron [ed.] Photosynthesis I. and their coupling in chloroplasts. Biochim. Biophys. Photosynthetic electron transport and photophospho- Acta 423: 499-509. rylation. Springer-Verlag, Berlin. HAIDAK, D., C. MATTHEWS, AND B. M. SWEENEY. KATz, J. J., L. L. SHIPMAN, AND J. R. NORRIS. 1979. 1966. Pigment protein complex from Gonyaulax. Structure and function in photoreaction-centre chlo- Science (Washington, D.C.) 152: 212-213. rophyll I, p. 1-40. In Chlorophyll organization and HALL, D. O., AND K. K. RAO. 1977. Photosynthesis. energy transfer in photosynthesis. CIBA Found. Symp. Edward Arnold Publications, London. 71 p. 61. Excerpta Medica. HARDING, L. W., B. W. MEESON, B. B. PRÉZELIN, KAWAMURA, M., M. MINIURO, AND Y. FUJITA. 1979. AND B. M. SWEENEY. 1981. Diel periodicity of Quantitative relationship between two reaction centers photosynthesis in marine phytoplankton. Mar. Biol. in the photosynthetic system of blue-green algae. 61: 95-105. Plant Cell Physiol. 20: 697-705. HARRIMAN, A., AND J. BARBER. 1979. Photosynthetic KE, B. 1978. The primary electron acceptors in green water-spilling process and artificial chemical systems, plant photosystem I and photosynthetic bacteria, p. p. 243-280. In J. Barber [cd.] Photosynthesis in 75-138. In D. R. Sanadi and L. P. Vernon [ed.] regulation to model systems. Elsevier Scientific Publi- Current topics in bioenergetics, Vol. 7. Photosyn- cations Co., Amsterdam. thesis: Part A. Academic Press, New York. HAXO, F. T. 1960. The wavelength dependence of photo KELLEY, G. J., E. LATZKO, AND M. GIBBS. 1976. synthesis and the role of accessory pigments, p. 339- Regulatory aspects of photosynthetic carbon metabo- 360. In M. B. Allen [ed.] Comparative biochemistry lism. Annu. Rev. Plant Physiol. 27: 181-205. of photoreactive systems. Academic Press, New York. KIEFER, D. A. 1973a. Chlorophyll a fluorescence in HAXO, F. T., J. H. KYCIA, G. F. SOMERS, A. BENNETT, marine centric diatoms: responses of chloroplasts to AND H. W. SIEGELMAN. 1976. Peridinin-chlorophyll light and nutrient stress. Mar. Biol. 23: 39-46. a proteins of the dinoflagellate Amphidinium carterae 1973b. Fluorescence properties of natural phyto- (Plymouth 450). Plant Physiol. 57: 297-303. plankton populations. Mar. Biol. 22: 263-269.

HAYDEN, D. B., AND W. G. HOPKINS. 1977. A second KIRK, J. T. O., AND R. A. E. TILNEY - BASSETT. 1978. distinct chlorophyll a-protein complex in maize The plastids. Elsevier N. Holland, Amsterdam. 960 p. mesophyll chloroplasts. Can. J. Bot. 55: 2525-2539. KNAFF, D. B., AND R. MALKIN. 1978. The primary HEANY, S. I. 1978. Some observations on the use of the reaction of chloroplast photosystem II, P. 139-174. in vivo fluorescence technique to determine chlorophyll In D. R. Sanadi and L. P. Vernon [ed.] Current a in natural populations and cultures of freshwater topics in bioenergetics, Vol. 7. Photosynthesis: Part phytoplankton. Freshwater Biol. 8: 115-126. A. Academic Press, New York. HERMAN, E. M., AND B. M. SWEENEY. 1975. Circadian KNox , R. S. 1977. Photosynthetic efficiency and excitation rhythm of chloroplast ultrastructure in Gonyaulax transfer and trapping, p. 55-98. In J. Barber [cd.] polyedra, concentric organization around a central of photosynthesis. Elsevier Scien- Primary processes cluster of ribosomes. J. Ultrastruc, Res. 50: 347-354. tific Publications Co., Amsterdam. HIYAMA, T., AND B. KE. 1972. Difference spectra and KOB, B., B. FORBUSH, AND M. MCGLOIN. 1970. Co- extinction coefficients of .moo. Biochim. Biophys. operation of charges in photosynthetic 02 evolution. Acta 267: 160-171. I. A linear four step mechanism. Photochern. Photo- HOMANN, P. H. 1979. The light dependence quenching biol. 11: 457-475. of chloroplast fluorescence by cofactors of cyclic KREY, A., AND GOVINDIEE. 1966. Fluorescence studies electron flow in photosystem I. Photochem. Photobiol. on a red alga, Porphyridium cruentuni. Biochim. 29: 815-822. Biophys. Acta 120: 1-18. ISLER, O. [ed.] 1971. Carotenoids. Halsted Press, New KRINSKY, N. I. 1979. Carotenoid pigments: multiple York. 932 p. mechanisms for coping with the stress of photosen- JAGENDORF, A. T. 1977. Photophosphorylation, p. 307- sitized oxidations, p. 163-187. In M. Shilo [ed.] 337. In A. Trebst and M. Avron [ed.] Photosynthesis Strategies of microbial life in extreme environment, I: Photosynthetic electron transport and photophos- Vol. 13. Gustav Fisher Inc. New York. phorylation. Springer-Verlag, Berlin. KULANDAI VELU, G., AND H. DANIELL. 1980. Dichlo- JEFFREY, S. W. 1969. Properties of two spectrally dif- rophenyl dimethylurea (DCMU) induced increase in ferent components in chlorophyll c preparations. chlorophyll a fluorescence intensity. An index of Biochem. Biophys. Acta 177: 456-467. photosynthetic oxygen evolution in leaves, chloro- JOLIOT, P. 1981. Photosystem II and oxygen evolution. plasts, and algae. Physiol. Plant. Pathol. 48: 385- 5th Int. Cong. Photosyn. Halikidi, Greece. (In press) 388. JORGENSEN, E. G. 1969. The adaptation of plankton algae. LAVOREL, J. 1975. Luminescence, p. 223-3 17. in Go- IV. Light adaptation in different algal species. Physiol. vindjee [ed.] Bionergetics of photosynthesis. Academic Plant. 22: 1307-1315. Press, New York. 1970. The adaptation of plankton algae. V. LAVOREL, J., AND A.-L. ETIENNE. 1977. In vivo chlo- Variation in the photosynthetic characteristics of rophyll fluorescence, p. 203-268. In J. Barber [cd.] Skeletonema costantm cells grown at low light inten- of photosynthesis. Elsevier Scien- Primary processes sity. Physiol. Plant. 23: 11-17. tific Publications Co., Amsterdam.

40 LEFORT-TRAN, M., G. COHEN-BAZIRE, AND M. MILLS, J. D., A. TELFER, AND J. BARBER. 1976. Cation POUPHILE. 1973. Les membranes photosynthétiques control of chlorophyll a fluorescence yield in chlo- des algues à biliprotéines observées après cryodéca- roplasts. Location of cation sensitive sites. Biochim. page. J. Ultrastruct. Res. 44: 199-209. Biophys. Acta 440: 495-505. LEY, A. C. 1980. The distribution of absolved light MITCHELL, P. 1974. A chemiosmotic molecular mecha- energy for algal photosynthesis, p. 59-82. In P. G. nism for protontranslocating adenosine triphospha- Falkowski [cd.] Primary productivity in the sea. tases. FEBS Lett. 43: 189-194. Plenum Press, New York, NY. MOHANTY, P., AND GOVINDJEE. 1974. The slow LEY, A. C., W. L. BUTLER, D. A. BRYANT, AND A. N. decline and the subsequent rise of chlorophyll GLAZER. 1977. Isolation and function of allophyco- fluorescence transients in intact algal cells. Plant cyanin B of Porphyridium cruentum. Plant Physiol. Biochem. J. 2: 78-106. 59: 974-980. Moss, G. P., AND B. C. L. WEEDON. 1976. Chemistry LOFTUS, M. E., AND H. H. SELIGER. 1975. Some limi- of the Carotenoids, p. 149-224. In T. W. Goodwin tations of the in vivo fluorescence technique. Chesa- [cd.] Chemistry and biochemistry of plant pigments. peake Sci. 16: 79-92. Academic Press, London. LONEGRAN, T. A., AND M. L. SARGENT. 1978. Regu- MUHLETHALER, K. 1980. Introduction to structure and lation of the photosynthesis rhythm in Euglena gracilis. function of the photosynthesis apparatus, p. 503- Plant Physiol. 61: 150-153. 520. In A. Trebst and M. Avron [cd.] Photosynthetic MACOLL, R., AND D. S. BERNS. 1978. Energy transfer electron studies on Cryptomonad biliproteins. Photochem. transport and photophosphorylation. Springer- Verlag, Berlin. Photobiol. 27: 343-350. MALKIN, S. 1977. Delayed luminescence, p. 349-432. MUNDAY, J. C. JR., AND GOVINDJEE. 1969. Light- In J. Barber [cd.] Primary processes of photosynthesis, induced changes in the fluorescence yield of chloro- Vol. 2. Elsevier Scientific Publications Co., Am- phyll a in vivo. III. The dip and the peak in the sterdam. fluorescence. Biophys. J. 9: 1-21. MALKIN, S., D. C. FORK, AND P. A. ARMOND. 1977. MURATA, N. 1969. Control of excitation transfer in photo- Probing photosynthetic unit sizes of leaves by fluo- synthesis. II. Magnesium ion-dependent distribution rescence-induction measurements. Carnegie Inst. of excitation energy between two pigment systems Washington Yearb. 77: 237-247. in spinach chloroplasts. Biochim. Biophys. Acta 189: MANN, J. E., AND J. MYERS. 1968. On pigments, growth, 171-181. and photosynthesis of Phaeodactylum tricornutum. MURATA, N., M. NISHIMURA, AND A. TAKAMIYA. J. Phycol. 4: 349-355. 1966. Fluorescence of chlorophyll in photosynthetic MARGULIES, M. M. 1970. Changes in absorbance spectrum systems. II. Induction of fluorescence in isolated of the diatom Phaeodactylunz tricornutum upon spinach chloroplasts. Biochim. Biophys. Acta 126: modification of protein structure. J. Phycol. 6: 160- 234-243. 164. MYERS, J., AND J. GRAHAM. 1971. The photosynthetic MARKWELL, J. P., C. C. MILES, R. T. BOGGS, AND unit in Chlorella measured by repetitive short flashes. J. P. THORNBER. 1979. Solubilization of chloroplast Plant Physiol. 48: 282-286. membranes by zwitterionic detergents. FEBS Lett. O'CARRA, P., AND C. O'HEOCHA. 1976. Algal bilipro- 99: 11-14. teins and phycobilins, p. 328-376. In T. W. Goodwin MARK WELL, LP., J. P. THORNBER, AND R. T. BOGGS. [cd.] Chemistry and biochemistry of plant pigments. 1979. Evidence that in higher plant chloroplasts all Academic Press, London. the chlorophyll exists as chlorophyll-protein com- °QUIST, G. 1974. Light-induced changes in pigment plexes. Proc. Natl. Acad. Sci. USA 76: 1233-1235. composition of photosynthetic lamellae and cell-free MARK WELL, J. P., J. P. THORNBER, AND M. P. SKRDLA. extracts obtained from the blue-green alga Atzacystis 1980. Effect of detergents on the reliability of a nidulans. Physiol. Plant. 30: 38-44. chemical assay for P700. Biochim. Biophys. Acta PAPAGEORGIOU, G. 1975. Chlorophyll fluorescence: 591: 391-399. an intrinsic probe of photosynthesis, p. 319-371. In MAssFlo, T. V., AND B. Kos. 1971. Detection and iso- Govindjee [cd.] Bioenergetics of photosynthesis. lation of P700, p. 515-522. In A. San Peitro [cd.] Academic Press, New York. in enzymology, Vol. XXIII, Photosynthesis: Methods PATTERSON, D. T., J. A. BUNCE, R. S. ALBERTE, AND Part A. Academic Press, New York. E. VAN VOLKENBURGH. 1977. Photosynthesis in MEESON, B., S. CHANG, AND B. M. SWEENEY. 1981. relation to leaf characteristics of cotton from controlled Characterization of the peridinin-chlorophyll a-pro- and field environment. Plant Physiol. 59: 384-387. tein (PCP) in the marine dinoflagellate Ceratium PERRY, M. J., M. C. LARSEN, AND R. S. ALBERTE. furca. (In prep.) 1981. Photoadaptation in marine phytoplankton: MEESON, B. W., AND B. M. SWEENEY. 1981. Adap- Response of the photosynthetic unit. Mar. Biol. 62: tation to irradiance: photoadaptive mechanisms and 91-101. growth rates for Ceratiwn furca and Gonyaulax PRÉZELIN, B. B. 1976. The role of peridinin-chlorophyll polyedra at two growth temperatures. (In prep.) a-proteins in the photosynthetic light adaptation of MILES, C. D., J. P. MARK WELL, AND J. P. THORNBER. the marine dinoflagellate, Glenodinium sp. Planta 1979. Effect of nuclear mutation in maize on photo- 130: 225-233. synthetic activity and content of chlorophyll-protein 1982. Growth dynamics in Gonyaulax polyedra. complexes. Plant Physiol. 64: 690-694. I. Light dependent patterns of aging in photosynthesis

41 and cell metabolism. Mar, Biol. (Submitted) SAMUELSSON, G., AND G. OQUIST. 1977. A method PRÉZELIN, B. B., AND R. S. ALBERTE. 1978. Photo- for studying photosynthetic capacities of unicellular synthetic characteristics and organization of chloro- algae based on in vivo chlorophyll fluorescence. phyll in marine dinoflagellates. Proc. Natl. Acad. Physiol. Plant. 40: 315-319. Sci. USA 75: 1801-1804. SAMUELSSON , G., G. OQUIST, AND P. HALLDAL. 1978. PRÉZELIN, B. B., AND B. BOCZAR. 1982. Light-har- The variable chlorophyll a fluorescence as a measure vesting chlorophyll-protein complexes of dinoflagel- of photosynthetic capacity in algae. Mitt. Int. Ver. lates. Symposium on pigment-protein complexes in Theor. Angew. Limnol. 21: 207-215. photosynthesis. Abst. Photochem. Photobiol. Colo- SANE, P. V., D. J. GOODCHILD, AND R. B. PARK. 1970. rado Springs, Col. p. 96-97. Characterization of chloroplast photosystern 1 and 2 1981. Chlorophyll-protein complexes from the separated by a non-detergent method. Biochim. photosynthetic apparatus of dinoflagellates. 5th Int. Biophys. Acta 216: 162-178. Cong. Photosyn. Halikidi, Greece. (In press) SATOH, K. 1980. F-695 emission from the purified photo- PRÉZELIN, B. B., AND F. T. HAXO. 1976. Purification system II chlorophyll a-protein complex. FEBS Lett. and characterization of peridinin-chlorophyll a-pro- 110: 53-56. teins from the marine dinoflagellates Glenodinium SATOH, K., AND W. L. BUTLER. 1978. Low temperature sp. and Gonyaulax polyedra. Planta 128: 133-141. spectral properties of subchloroplast fractions purified PRÉZELIN, B. B., AND A. C. LEY. 1980. Photosynthesis from spinach. Plant Physiol. 61: 373-379. and chlorophyll a fluorescence rhythms of marine SAUER, K. 1975. Primary events and the trappings of phytoplankton. Mar. Biol. 55: 295-307. energy, p. 115-181. In Govindjee [cd.] Bioenergetics Press, New York. PRÉZELIN, B. B., A. C. LEY, AND F. T. HAXO. 1976. of photosynthesis. Academic Effects of growth irradiance on the photosynthetic 1981. Charge separation in the light reactions Cong. Photosyn. Halikidi, action spectra of the marine dinollagellate, Glenodi- of photosynthesis 5th Int. niunz sp. Planta 130: 251-256. Greece. (In press) SEELEY, G. R. 1966. The structure and chemistry of PRÉZELIN, B. B., AND H. A. MATLICK. 1980. Time- functional groups, p. 67-110. In L. P. Vernon and course of photoadaptation in the photosyntheses- G. R. Seeley [cd.] The chlorophylls , physical, chemical irradiance relationship of a dinoflagellate exhibiting and biological properties. Academic Press, New photosynthetic periodicity. Mar. Biol. 58: 85-96. York. PRÉZELIN , B. B., B. W. MEESON , AND B. M. SWEENEY. SENGER, H. 1970a. Characterization of a synchronous 1977. Characterization of photosynthetic rhythms in culture of Scenedesmus obliques: its potential photo- marine dinoflagellates. I. Pigmentation, photosyn- synthetic capacity and its photosynthetic quotient thetic capacity and respiration. Plant Physiol. 60: during the life cycle. Planta 92: 243-266. 384-387. 1970b. Quantum yield and variable behavior PRÉZEL1N, B. B., AND B. M. SWEENEY. 1977. Char- of the two photosystems of the photosynthetic appa- acterization of photosynthetic rhythms in marine ratus during the life cycle of Scenedesmus obliques dinoflagellates. II. Photosynthesis-irradiance curves in synchronous culture. Planta 92: 327-346. and in vivo chlorophyll a fluorescence. Plant Physiol. SHIOZAWA, J. A., R. S. ALBERTE, AND J. P. THORNBER. 60: 388-392. 1974. The P700 chlorophyll a-protein. Isolation and 1978. Photoadaptation of photosynthesis in some characteristics of the complex in higher plants. Gonyaulax polyedra. Mar. Biol. 48: 27-35. Arch. Biochem, Biophys. 165: 388-397. 1979. Photoadaptation of photosynthesis in two SIEGELMAN, H. W., J. H. KYCIA, AND F. T. HAXO. bloom-forming dinoflagellates, p. 101-106. In C. 1977. Peridinin-chlorophyll a-proteins of dino- Ventsch [cd.] Proc. 2nd Int. Conf. Toxic Dinoflagellate flagellate algae. Brookhaven Nat!. Symp. 28: 162-169. Blooms. SLOVACECK, R. E., AND P. J. HANNAN. 1977. In vivo RAMUS, J., S. I. BEALE, D. MAUZERALL, AND L. fluorescence determinations of 197 phytoplankton HOWARD. 1976a. Changes in photosynthetic pigment chlorophyll a. Limnol. Oceanogr. 22: 919-925. concentration in seaweeds as a function of water depth. SONG, P.-S., P. KOKA, B. B. PRÉZELIN, AND F. T. Mar. Biol. 37: 223-230. flAxo. 1976. Molecular topology of the photosynthetic 1976b. Correlation of changes in pigment content light-harvesting pigment complex, peridinin-chlo- with photosynthetic capacity of seaweeds as a func- rophyll a-protein, from marine dinoflagellates. tion of water depth. Mar. Biol. 37: 231-238. Biochemistry 15: 4422-4427. RAVEN, J. A., AND J. BEARDALL. 1981. Respiration and SONG, P. S., E. B. WALKER, J. JUNG, R. A. AUERBACH. photorespiration. Can. Bull. Fish. Aquat. Sci. 210: G. W. ROBINSON, AND B. B. PRÉZEL1N. 1980. 55-82. Primary processes of phytobiological receptors. ln REGER, B. J., AND R. W. KRAUSS. 1970. The photo- New horizons in biological chemistry. Acad. Publ. synthetic response to a shift in the chlorophyll a to Cent. Tokyo, Japan. ratio of More/la. Plant Physiol. 46: SONNEVELD, A., H. RADEMAKER, AND L. N. M. 568-575. DUYSENS. 1979. Chlorophyll a fluorescence as a REY, F. 1978. Some results on the application of the monitor of nanosecond reduction of the photooxidized in vivo chlorophyll fluorescence method to marine primary donor P-680f of photosystem II. Biochim. primary productivity studies, p. 36-47. in Symposium Biophys. Acta 548: 536-551. on fluorescence and luminescence. Dep. Plant Physiol., SPECTOR, M., AND G. D. WINGET. 1980. Purification Univ. Umea, Sweden. of a manganese-containing protein involved in photo-

42 synthetic oxygen evolution and its use in reconstituting THORNBER, J. P., R. S. ALBERTE, F. A. BUNTER, J. A. an active membrane. Proc. Natl. Acad. Sci. USA SHIOZAWA, AND K. S. KAN. 1977. The organization 77: 957-959. of chlorophyll in the plant photosynthetic unit, p. STAEHELIN, L. A. 1976. Reversible particle movements 132-148. J. M. Olson and G. Hind [ed.] Chlorophyll- associated with unstacking and restacking of chloro- proteins, reaction centers, and photosynthetic mem- plast membranes in vitro. J. Cell Biol. 71: 136-158. branes. Brookhaven Symp. Biol, 28. STAEHELIN, L. A., P. A. ARMOND, AND K. R. MILLER. THORNBER, J. P., AND J. BARBER. 1979. Photosynthetic 1977. Chloroplast membrane organization at the pigments and models for their organization in vivo, supramolecular level and its functional implications, p. 27-70. In J. Barber [ed.] Photosynthesis in relation p. 278-315. 10 Chlorophyll-proteins, reaction centers, to model systems. Topics in photosynthesis. Vol. 3. and photosynthetic membranes. Brookhaven Symp. Elsevier Scientific Publications CO. Biol. 28. THORNBER, J. P., J. P. MARKWELL, AND S. REINMAN. STAEHELIN, L. A., AND C. J. ARNTZEN. 1979. Effects 1979. Plant chlorophyll-protein complexes: recent of ions and gravity forces on the supramolecular organ- advances. Photochem. Photobiol. Yrly. Rev. 29: ization and excitation energy distribution in chloroplast 1205-1216. membranes, p. 147-175. In Chlorophyll organization TREBST, A. 1974. Energy conservation in photosynthetic and energy transfer in photosynthesis. CIBA Found. electron transport of chloroplasts. Annu. Rev. Plant Symp. 61, Excerpta Medica. Physiol. 25: 423-458. STAEHELIN, L. A., D. P. CARTER, AND A. MCDONNEL. TRENCH, R. K., AND G. S. RONZIO. 1978. Aspects 1980. Adhesion between chloroplast membranes: of the relation between Cyanophora paradoxa (Kor- experimental manipulation and incorporation of the schikoff) and its endosymbiotic cyanelles Cyanocyta adhesion factor into artificial membranes, p. 1-179. korschikoffiatm (Hall & Claus). II. The photosynthetic In B. Gilula [ed.] Membrane-membrane interactions. pigments. Proc. R. Soc. London Ser. B 202: 445-462. Raven Press, New York. VELTHUYS, B. R. 1980. Mechanisms of energy frow in SWEENEY, B. M. 1981. Chloroplast membranes of photosystem II and toward photosystem I. Annu. Gaunyaulax polyedra (Pyrrophyta), a photosynthetic Rev. Plant Physiol. 31: 545-567. dinoflagellate with peridinin-chlorophyll-protein, VIERLING, E., AND R. S. ALBERTE. 1980. Functional studied by the freeze-fracture technique. J. Phycol. organization and plasticity of the photosynthetic unit (In press) of the cyanobacteriumAnacystis nidulans. Physiol. Plant. SWEENEY, B. M., B. B. PRÉZELIN, D. WONG, AND 50: 93-98. GOVINDJEE. 1979. Chlorophyll a fluorescence of VINCENT, W. F. 1980. Mechanisms of rapid photosyn- Gonyaulax polyedra grown on a light-dark cycle and thetic adaptation in natural phytoplankton commu- after transfer to constant light. Photochem. Photobiol. nities. II. Capacity for non-cyclic electron transport. 30: 405-411. J. Phycol. 16: 368-577. TANADA, T. 1951. The photosynthetic efficiency of caro- WILLIAMS, W. P. 1977. The two photosystems and their tenoid pigments in Navieula minima. Am. J. Bot. interactions, p. 99-145. In J. Barber [ed.] Primary 38: 276-290. processes of photosynthesis. Vol. 2. Elsevier Scientific TELFER, A., J. NICHOLSON, AND J. Barber. 1976. Cation Publications, Amsterdam, Holland. control of chloroplast structure and chlorophyll a fluo- WITT, H. T. 1979. Energy conversion in the functional rescence yield and its relevance to the intact chloro- membrane of photosynthesis. Analysis by light pulse plast. FEBS Lett. 65: 77-83. and electric pulse methods. The central role of the TERBORGH, J., AND G. C. McLEop. 1967. The photo- electric field. Biochim. Biophys. Acta 505: 355-427. synthetic rhythm of Acetabulana cremdata. I. Con- 1979. Charge separation in photosynthesis, p. tinuous measurements of oxygen exchange in altern- 303-330. In H. Gerischer and J. J. Katz [cd.] Light- ating light-dark regimes and in constant light at induced charge separation in biology and chemistry. different intensities. Biol. Bull. 133: 659-669. H. Gerischer and Dahlem Konferenzen, Berlin. TERRI, J. A., D. T. PATTERSON, R. S. ALBERTE, AND WOOD, P. M. 1977. The roles of c-type cytochromes in R. M. CASTELBERRY. 1977. Changes in the photo- algal photosynthesis. Extraction from algae of a cyto- synthetic apparatus of maize in response to simulated chrome similar to higher plant cytochrome f. Ent.. J. natural temperature fluctuations. Plant Physiol. 60: Biochem. 72: 605-612. 370-373. WRAIGHT, C. A., AND A. R. CROFTS. 1970. Energy THIMANN, K. V. [ed.] 1980. Senescence in plants. CRC dependent quenching of chlorophyll a fluorescence Press, Boca Raton, FL. 238 p. in isolated chloroplasts. Eur. J. Biochem. 17: 319-327. THORNBER, J. P., AND R. S. ALBERTE. 1977. The organ- ZILINSKAS, B. A., B. K. ZIMMERMAN, AND E. GANTT. ization of chlorophyll in vivo, p. 574-582. In A. 1978. Allophycocyanin forms isolated from Nostoc Trebst and M. Avron [cd.] Photosynthesis I: photo- sp. phycobilisomes. Photochem. Photobiol. 27: 587- synthetic electron transport and photophosphorylation. 595. Springer-Verlag, Berlin.

43 Dark Reactions of Photosynthesis

BRUNO P. KREMER1 Universitiit zu Kôln, Seminar fiir Didaktik der Biologie D-5000 Kôln 41, West Germany

Introduction involving oxidative metabolic processes that demand an oxidizing environment, allowing for final con- All life on earth ultimately depends on assimi- version of all carbon compounds to the most oxidized lation of carbon dioxide as the main carbon source form ( = CO 2 ). Photosynthesis restores the balance for production of organic material; the energy for by generating oxygen and by fixation of CO 2 , re- this endergonic process is provided by light origi- sulting in more reduced carbon being made avail- nating from nuclear fusions iii the sun. Collectively, able to catabolism. Thus, photosynthesis supports the reactions for this conversion of absorbed photons all forms of life on earth by a rather unique complex into stable chemical energy are termed photosynthe- of reactions that can utilize and conserve radiant sis. Usually, it is approached as a three-phase process energy from outside. comprising (i) the absorption of light quanta and Photosynthetic organisms capable of these retention of light energy, (ii) the conversion of light reactions include all plants equipped with chloro- energy into chemical potential, and (iii) the inter- phyll a together with at least one major accessory conversion, stabilization, and storage of the chemical pigment. Among them are vascular plants and potential. While the complex of light absorption, mosses, the wide variety of blue-green, red, green, energy transduction, photochemical events, and and brown pigmented algae, the lichens, as well finally the concomitant formation of products such as a number of photosynthesizing bacteria belonging as reducing (NADPH) and energy equivalents (ATP) to the Thiorhodaceae (purple sulfur bacteria), Athior- represent the light reactions of photosynthesis (see hodaceae (purple nonsulfur bacteria), and Chloro- Prézelin 1981), further steps involving the fixation and bacteriaceae (green sulfur bacteria). Photosynthesiz- reduction of inorganic carbon at the expense of ing organisms sensu lato are also aquatic invertebrate NADPH and ATP are the so-called dark reactions of animals that contain unicellular algae or even algal photosynthesis . organelles as functional endosymbiotic units (cf. Light is not directly required for the fixation Trench 1979). and reduction of CO.. During light—dark transients, however, the biochemical conversion steps involved Basis of Autotrophy: Reductive Pentose rapidly decline due to steady exhaustion of the assimi- Phosphate Cycle latory power provided by the light reactions. On the other hand, the individual conversions can be CARBON SOURCE studied in vitro with isolated systems, when NADPH and ATP are added to the test system instead of The initial steps among the dark reactions of being derived directly from irradiated thylalcoids. photosynthesis involve uptake, incorporation, and In addition, there are some further, presumably fixation of inorganic carbon by carboxylation of a regulatory, effects of the light reactions on individual preformed substrate, ribulose-1, 5-bisphosphate. reaction steps in the biochemical reactions of pho- Inorganic carbon is available to terrestrial plants as tosynthesis other than those requiring ATP and atmospheric carbon dioxide. Under normal conditions NADPH. They will not be considered further in this (20°C, 1 bar (1 bar = 100 kPa)), the atmosphere context. contains at least 0.03% CO 2 (already in 1980 about Photosynthesis is the basic process that provides 360 ppm), which is equivalent to — 15 gmol living organisms with the organic material needed Marine plants live in an environment that contains for growth and maintenance, but, in turn, plant inorganic carbon either as dissolved CO 2 or as H. CO3 cells or their products also yield the foodstuff for as well as HCC:q and COI- . The total carbon content all other components of the biosphere. All organisms in seawater is a function of pH, temperature, and salin- have to consume energy by means of degradation ity (see Skirrow 1975). In the normal pH range (between 7.8 and 8.2) HCO 1-• is by far the predom- ' Correspondence to: Dr B. P. Kremer, Andreasstrasse imant form, comprising more than 90% of the total 51, D-5300 Bonn 2 (FRG). inorganic carbon present. The average content

44 amounts to about 2.2-2.5 mmol • L - '. Some aquatic indicative of the nature of the primary carboxylation plants take up unhydrated CO 2 , while others utilize event of the photosynthetic pathway. The resulting the bicarbonate (HCO7) (Jolliffe and Tregunna '"C values allow discrimination between C, and 1970). Those operating with CO., rely on only about C, plants (see discussion below). 10 iLL • L- ' as a carbon source, whereas those incor- porating HCO usually have more than 2 mmol • L- ' available. REACTIONS OF REDUCTIVE PENTOSE PHOSPHATE CYCLE The bicarbonate concentration of seawater is not likely to limit the rate of photosynthesis in The basic biochemical pathway that provides plankton or seaweeds; it is usually far more than fixation of inorganic carbon and converts it to highly saturating (cf. Harris 1978). It must be considered, reduced organic compounds is the reductive pentose however, that the diffusion constant of CO., in the phosphate cycle (RPP cycle below). The reaction atmosphere is about 0.16 cm' • s - ', whereas it is sequence originally proposed (see Bassham and 2 x 10-3 cm' • s- ' in seawater. This ratio of about Calvin 1957) has required no essential additions or 10' in the diffusion velocities compensates for any changes since its discovery, and is obviously a uni- possible advantage that a bicarbonate-utilizing marine versally operating metabolic pathway, as may be plant might have over a terrestrial species, with determined from experimental findings by a wide regard to concentrations of inorganic carbon in the variety of photosynthesizing organisms. This brief respective environments (cf. Raven 1970). consideration of the reactions involved follows the The carboxylating enzyme initiating photosyn- more comprehensive treatment by Bassham (1979), thetic carbon incorporation, ribulose-1,5-bisphos- which should be consulted for additional information. phate carboxylase, has been found to use unhydrated Three operational phases in series can be dis- COo rather than HCO;i" (cf. Raven 1970). Those tinguished, comprising the carboxylation, the re- aquatic species relying on bicarbonate must, there- duction, and the cyclic regeneration of the starting fore, be able to convert HCO to unhydrated CO.,. material. The appropriate enzyme system, carbonic anhydrase Carboxylation — The enzyme ribulose-1,5- (= carbonate dehydratase), catalyzes this conversion bisphosphate carboxylase ( = RuBP Case, EC effectively. It is present in various representatives 4.1.1.39) catalyzes the condensation of CO 2 supplied of algae (Graham and Smillie 1976), and it appears by inter-/intracellular diffusion with the Co atom reasonable to ascribe to it a significant role in photo- of ribulose-1,5-bisphosphate (RuBP). There is con- synthesis. Nonetheless, various aspects of the function vincing experimental evidence that the transition and importance of this peculiar enzyme remain Cli intermediate, resulting from the initial carboxyla- elusive (cf. Lamb 1977). tion step, The CO., of the atmosphere is a mixture of is an enzyme-bound, 2-carboxy-3-ketoribi- tol-1,5-bisphosphate (Sj6din and Vestermark 1973). the stable isotopes I2C Ili0 Hie , 13C••Hu-• HO, and This short-lived intermediate is then hydrolytically 12 C 'HO ' 80, but the isotope ' 2C in total contributes split to give two molecules of 3-phosphoglycerate 98.89% to atmospheric COo . Most photosynthesizing (PGA). PGA is a C3 acid and the first stable, analyt- plants have been found to prefer this lightest of tile ically traceable product of the carbon isotopes. Usually, the ratio '"C/ '"C as photosynthetic car- boxylation step. determined by mass spectroscopy is expressed as It was one of the earliest photosyn- thetic intermediates identified by kinetic tracer the SHC (%0) value. Most plant 6 '"C values are studies using "CO.,. Cells were allowed to reach negative, indicating that the proportions of the iso- steady-state photosynthesis topes in organic compounds arising from photosyn- in ''CO2 so that all pools thesis do not exactly reflect the relative amounts of the presumably cycling intermediates were satu- present in the abiotic atmosphere, but that living rated with radiocarbon. When the further incubation organisms are somewhat depleted in '"C. was interrupted by darkening the cells, the immediate In aquatic environments, the equilibrium reac- products of the photosynthetic light reactions (ATP tions of CO., complicate the situation. As bicarbonate and NADPH) were cut off, while carboxylation could already has a HC value which is somewhat lower continue. This resulted in the simultaneous disap- than that of atmospheric CO2 , algae utilizing HCO pearance of the CO., acceptor molecules (RuBP) and the appearance of the compounds derived from are expected to show a comparable shift. In fact, the carboxylation (PGA) (Fig. 1). such shifts have been found for a variety of marine The C atom from CO., fixed to the Co atom algae in the range of the isotope fractionation between of RuBP furnishes the carboxyl group of one HCO and CO., (Smith and Epstein 1971; Black PGA. This was demonstrated by chemical degradation of and Bender 1976). PGA, which allowed the further fate of the labeled Carbon isotope fractionation between ' 2 C and carbon atom to be traced through the subsequent HC (and, moreover, between ' 2 C and NC) is also reactions of the RPP cycle.

45 Light Light off on

100 4110 800 2000 TIME[s]

FIG. 1. Changes in concentrations of 3-phosphoglycerate (PGA) and ribulose-I,5- bisphosphate (RuBP) during light—dark transients (after Bassham and Calvin 1957).

Because the first product of the initiation phase of These conversions include the following individual the cycle is a C3 acid, this type of carbon fixation is reactions, and are shown schematically in Fig. 2. usually termed C3 photosynthesis. 1) Two molecules of GAP are converted to dihydroxyacetone phosphate (DHAP) with triose Reduction of PGA —During the reduction phosphate isomerase. phase of photosynthetic carbon assimilation, PGA 2) The enzyme aldolase is able to condense 3-phosphate (GAP) is converted into glyceraldehyde two molecules of triose phosphate to give fructose- (triose phosphate). representing a C3 carbohydrate 1,6-bisphosphate (FBP) in a reversible reaction. This conversion is strictly endergonic, requires 3) FBP is subjected to dephosphorylation by coupling with an exergonic reaction, and occurs in fructose bisphosphatase yielding fructose-6-phos- two subsequent steps. First, PGA is phosphorylated phate (F6P). GAP originating from the reduction to give the acylphosphate glycerate-1,3-bisphos- phase of the cycle and F6P provided by conversions phate (PPGA) at the expense of ATP as the donor 1-3 form the substrates of the subsequent rearrange- of the phosphate group. The reaction is catalyzed ments which involve some changes in the chain by phosphoglycerate kinase. length of the sugar phosphates participating in the implies the reduction The second conversion sequence. of PPGA to GAP by a NADPH-specific triose phos- 4) By action of the enzyme transketolase a phate dehydrogenase. This is a typical and specific C.› unit is transferred from F6P to GAP leaving a photosynthetic enzyme located exclusively in the C 1 sugar phosphate, erythrose-4-phosphate (E4P), photosynthesizing chloroplast. It differs considerably and simultaneously yielding a C5 sugar phosphate, from the cytoplasmic NADH-linked enzyme par- xylulose-5-phosphate (Xu5P). This reaction is re- in glycolysis. The overall reduction of PGA ticipating versible. It provides one of the three pentose phos- to GAP is energetically somewhat unfavorable. phates required for the regeneration of RuBP.

Regeneration of the CO., acceptor — The third 5) The aldose phosphate E4P then condenses phase in the dark reactions of photosynthesis, fol- with the triose phosphate DHAP in a reaction again lowing carboxylation and reduction of the first oc- mediated by aldolase (cf. step 2 of the regenerating curring photosynthate, includes a set of conversions phase) to form a C7 compound, sedoheptulose-1,7- designed to regenerate the specific acceptor molecule bisphosphate (SBP). Similar to RuBP, this compound of CO,, RuBP. For this purpose, the triose phos- is restricted to the photosynthetic pathway of carbon phate undergoes a series of isomerizations, conden- metabolism and does not occur in further metabolic sations, and molecular rearrangements by which five sequences. molecules of GAP finally provide six molecules of 6) The ketose phosphate SBP is dephospho- a pentose phosphate (ribulose-5-phosphate, R5P). rylated to sedoheptulose-7-phosphate (S7P) with

46

RuBP HC-OR5P HC-OH He- 04:9 1-12Ç-Cle H2C-OH HÇ-OH CO CO - 00C-C-OH 6.0 HÇ-OH Ô-0 HO-CH Hzç-De HC-OH HC-OH HÔ-OH H2C-0-e H2C-oe côo- coo- coo- coo- coo- COO- 1-12e-oe HOH HC-OH HÔ-OH HÔ-OH HC-OH HÔ-OH PGA H2c-oe Xu5P o H-Pge Fee) HP-P-eH2C-Pe H26-9® 1-P,k-00 OH 6 ATP HOW OH S7P cdo-ID 001D c60-(0 côo-ID c60-e coolD HO HÔ-OH Hc-OH HÔ-01-1 HÔ-OH HÇ-OH H 6-01-1 PPGA HO® H26-ge H2c-Per-12c-9(9 H2t-pe H0 H5-0 0 He-0-e 6 NADPH HÇ=0 1-1C-0 HC:0 HC;0 HC-.0 HO \. HO/ OH HÇ-OH HÇ-OH HC-OH HÔ-OH HÔ-OH HÔ-OH GIP

HO À SBP H2C-Oe I-Lp-oeHece Hp-oe Rpoe,H,c-oe H2Ç-OH HÇ 0 C.0 HC-OH HO-ÔH HÔ-OH }-12Ç-OH F-12ç-OH HÔ-OH H2C-0{0- C, 0 ÔO NET GAIN H2C-0-e 71/ E4P H2C-0-eH2C-0-e DHAP Xu5P

e-O-CH2 0 yH -OH e0-CH20 1e0 -12) HQ O H 0 (,,,., OH nut HO HO F61' HO

FIG. 2. Reactions of the reductive pentose phosphate cycle. Upon one turn of the cycle, three molecules of CO2 are fixed and a net gain of one molecule of the C, com- pound glyceraldehyde phosphate is achieved. All further reactions are concerned with the cyclic regeneration of the CO, acceptor. For abbreviations see text. Enzymes involved in the conversions and molecular rearrangements: 1, ribulose-1,5-bisphosphate carboxylase; 2, phosphoglycerate kinase; 3, triosephosphate dehy- drogenase; 4, triosephosphate isomerase; 5, aldolase; 6, fructose-1,6-bisphosphatase; 7, transketolase; 8, sedoheptulose-1,7-bisphosphatase; 9, phos- phoketopentose epimerase; 10, phosphoketopentose isomerase; 11, ribulose-5-phosphate kinase. catalysis by sedoheptulose-1,7-bisphosphatase, an the dark reactions of photosynthesis is thus very enzyme that is obviously distinct from the phospha- near 90%. However, if the actual, physiological in tase mediating step 3 of this cycle. situ concentrations of the individual metabolic inter- 7) Finally, a further conversion catalyzed by mediates of the cycle are considered, a more realistic transketolase occurs by which the C I and C., atoms approach to the energetics and the overall efficiency of S7P are again transferred to GAP, resulting in can be achieved. From such calculations, which will the formation of two different pentose phosphates not be detailed further here, the theoretical àG (Xu5P and ribose-5-phosphate; R5P). Both com- values are corrected to physiological A.Gs values. pounds are interconvertible by simple isomerization On this certainly more relistic and relevant back- and represent the remaining C, compounds needed ground, the energy input of 6 mol of ATP and of for the regeneration of RuBP. 9 mol of ATP amounts to 427 kcal. The free energy As in Fig. 2, three molecules of RuBP are finally converted to heat to keep the entire cycle carboxylated by RuBPCase to form six molecules of running is àGs = —73.4 kcal. This would estab- PGA and, on reduction, six molecules of GAP. lish an efficiency in the range of 80% which, however, Five of these GAP molecules are consumed for the is still rather respectable (Bassham and Krause 1969). regeneration of three molecules of RuBP, thus com- pleting the RPP cycle. The remaining sixth GAP MAPPING THE RPP CYCLE; represents the net gain of fixed organic carbon per METHODS AND DEVELOPMENTS turn of the cycle resulting from the initial input of Several methodological advances initiated the three molecules of CO,. research on photosynthetic carbon assimilation, The further fate of the net gain GAP involves which finally led to the formulation of the RPP cycle conversions similar to those formulated as the initial as given above. The most important progress was steps of the regenerating phase of the RPP cycle. probably the introduction of the radioisotope tracer They provide net photosynthate in the form of hexose technique. Since then it became possible to follow phosphates (giving rise to accumulated free hexoses the path of carbon from the carboxylation to the and related compounds) as well as the biosynthesis reduction of the triose phosphate and the cyclic of polymeric storage carbohydrates. There are several regeneration of the acceptor molecule. In addition, branching points where intermediates of the RPP this acquisition required further developments, cycle can leave the photosynthetic pathway sensu including two-dimensional chromatography and auto- stricto to be subsequently fed into other anabolic radiography for the characterization and identification sequences for the biosynthesis of cellular constituents. of small amounts of compounds involved. Only by such methods was it found that PGA, DHAP, FBP, and GAP were among the products formed during ENERGETICS AND STOICHIOMETRY the first seconds of exposure to 'CO— Chemical For each mole of CO2 entering the carboxylation degradation of such early photosynthates revealed by RuBPCase, I mol of ATP is consumed for the that ' 1 C was located mostly in the carboxyl oe PGA formation of RuBP from Ru5P or Xu5P and R5P, or the C3 /C I atoms of the hexose phosphates, thus respectively (see Fig. 2). Two further moles of ATP strongly suggesting that the C2 acid once synthesized are required for the activation of 2 mol of PGA (arising is rapidly converted to C i; carbohydrates. Time- from each mole of carboxylated RuBP) to PPGA. The dependent changes in proporional "C labeling as subsequent reduction step strictly needs 2 mol of obtained from kinetic studies of carbon flow con- NADPH. As a complete turn of the RPP cycle involves firmed this view. The methods originally applied uptake and fixation of three molecules of CO2 to give a have meanwhile been supplemented and improved net gain of one GAP, a total of 9 mol of ATP and 6 mol in many respects. They are still an indispensable tool of NADPH are the minimal investment required. for the investigation of algal primary metabolism The 9 mol of ATP represent an energy equiv- (see Kremer 1978). alent of 68.8 kcal, the 6 mol of NADPH a total of Another fundamental approach to the dark 315.5 kcal. The energy input into one turn of the reactions of photosynthesis was the characterization RPP cycle is thus, theoretically, in the range of of the individual enzymes involved. Quayle et al. 384.3 kcal. Energy recovered from the end product (1954) first demon trated the formation of PGA of the reduction phase, GAP, amounts to about in vitro upon carboxylation, when radioactive CO., 350.4 kcal/mol. Substracting the energy stored in and RuBP were added to an enzyme preparation GAP from that expended by consumption of ATP obtained from Chlorella cells. Comprehensive work and NADPH, a difference of AG ' = —33.9 kcal has since been conducted on the properties of is obtained, representing about 10% of the original RuBPCase, which now appears to be one of the input which is obviously wasted to drive the RPP most important plant enzymes. During approximately cycle. The efficiency of carbon reduction during the same period that the RPP cycle was formulated 48 on the basis of radiotracer kinetic studies, an oxi- (1957) found that in photosynthesizing sugarcane dative pentose phosphate cycle was discovered (see leaves the first occurring ''C-labeled photosynthates Horecker et al. 1953), and it soon became evident were preponderingly C 3 dicarboxylic acids such as that some conversions formulated for this pathway aspartate and malate rather than PGA, the traditional were the exact reverse of the reactions of the RPP primary product of photosynthetic CO, fixation. Just cycle, thus providing further biochemical evidence one decade after the elucidation of the universally for the consolidation of the concept. operating RPP cycle, Hatch and Slack (1966) pre- A third discovery should be mentioned. Arnon sented a biochemical reaction scheme to account (1952) demonstrated the activity of an enzyme able for the particular pattern of ' 4C labeling among the to use NADPH as a coenzyme and to catalyze the early photosynthates observed in sugarcane. This reduction of PPGA to GAP. This enzyme, glyceral- was the onset of the C wave. dehyde phosphate dehydrogenase, provides the C photosynthesis has hitherto mostly been asso- fundamental reduction step from the activated C3 ciated with green vascular plants exhibiting a spe- acid to the level of a carbohydrate. This reaction cialized type of leaf anatomy. In contrast with the utilizes the reducing power of NADPH and thus majority of angiosperms whose leaf tissue is usually shows where and how an important product generated organized into the two distinct cell types (the palisade by the photosynthetic light reactions is consumed. layer and the spongy parenchyma), some other species The light and dark reactions of photosynthesis are have a so-called Kranz leaf anatomy. In such leaves intimately connected at this particular site of the the veins are much closer together, and each is RPP cycle, which, in addition, represents the site surrounded by a particular layer of bundle sheath of an immediate on—off regulation of the entire cells that contain large numbers of chloroplasts. The cycling pathway. bundle sheath layer is embedded in mesophyll cells with only rather small air spaces between them. Additional Mechanisms of The complex reactions of C photosynthesis are Carbon Fixation compartmentated between bundle sheath cells (BSC) and mesophyll cells (MC). The entire context is schematically represented in Fig. 3. The basic re- Experiments using the tracer technique to follow action takes place in the MC and involves the p- patterns and kinetics of 'C labeling provide evidence carboxylation of PEP to oxaloacetate (OAA) by that fixation of CO, by the RPP cycle is not the phosphoenolpyruvate carboxylase (PEPCase). sole or an exclusive reaction by which carbon is Although it has originally, and then repeatedly, been fixed. There are further mechanisms for assimilation proposed that this enzyme is associated with MC of inorganic carbon that should briefly be discussed chloroplasts, it is now widely accepted that it is in this connection. Such reactions are best associated actually located in the cytoplasm (Coombs 1979). with the term "C metabolism," which is, as will OAA is a rather unstable reaction product and will be discussed below, not necessarily synonymous, usually not be found after extraction and chroma- or even identical with C photosynthesis. Nonetheless, tographic isolation. It is immediately converted to these reactions (often) show certain relationships to the either malate by malate dehydrogenase (NADPH- RPP cycle and may, therefore, also be regarded as linked) and/or to aspartate by OAA—asparate amino- components of the photosynthetic dark reactions. transferase. Whether prevailingly malate or aspartate A common feature of the reactions to be con- is formed from OAA depends on the species inves- sidered here is the substrate to be carboxylated. It tigated. is a C3 compound, usually phosphoenolpyruvate Accordingly, malate and/or aspartate is ex- (PEP), which is condensed with CO, to form a ported from MC and C dicarboxylic acid. In this particular connection, transferred to BSC. If malate is the C dicarboxylic two enzymatic reactions have to be considered: phos- acid to be transferred, it is decarboxylated in the phoenolpyruvate carboxylase and phosphoenol- cytoplasm by malic enzyme, simultaneously regenerating pyruvate carboxykinase. Both enzymes introduce the NADPH previously consumed for a second carboxyl group ( = 0-carboxyl) into the its synthesis. Transfer of malate from C3 acid PEP by a reaction generally termed 0-carbo- MC to BSC thus implies the transport of reduction xylation. equivalent. The C3 compound resulting from de- carboxylation of malate is pyruvate. It is returned P-CARBOXYLATION VIA to MC for restoring the pool of PEP. PHOSPHOENOLPYRUVATE CARBOXYLASE If aspartate is transferred from MC to BSC, it is first reconverted to OAA and then decarboxylated A first indication of the existance of a further by NAD-linked malic enzyme. In this case, pyruvate carboxylating mechanism involved in photosynthetic or alanine (formed from pyruvate on transamination carbon fixation was obtained when Kortschak et al. to prevent increasing concentration of toxic NH;,

49 C°2

Malate PyrtNate çoo- coo- COO- PEP CO2 OAA 1-1C-01-1 HC-01-1 è=13 + CO2 çoo- Coo- - d-I2 91-12 6k3 Ô-o Ô00- COO- RPP O ço JP CYCLE bH2 cH2 000 - COO' COO boo- --E12KI-1-1 -CH C.0 + CO2 CI-12 CI-12 3 à1-1 boo- boo- GAP Aspartate STARCH

CHLOROPLAST C00. - 900- H2N-CH • H2N-CH i pp d1-13 CH3 ÇOO- Alanine C4D AMP ATP Cl-13 ADP ATP CHLOROPLAST MESOPHYLL CELL BUNDLE SHEATH CELL

FIG. 3. Reactions and conversions in mesophyll and bundle sheath cells of C I plants. For abbreviations see text. Enzymes involved: 1, phosphoenolpyruvate carboxylase; 2, malate dehydrogenase; 3, oxaloacetate-aspartate aminotransferase; 4, NADH-linked malic enzyme (after reduction of OAA to malate by malate dehydrogenase); 5, NADPH-linked malic enzyme; 6, pyruvate-alanine aminotransferase; 7, pyruvate dikinase. In some plants reaction 4 involves decarboxylation of OAA by phosphoenolpyruvate carboxykinase to form PEP and CO, The scheme does not include any figure on the participation of mitochondria, although the NADH-linked malic enzyme is clearly of mitochondrial origin.

in BSC) is reimported into MC. In some C plants, than does RuBPCase (K,,, in the range of 400 pM), aspartate is decarboxylated by phosphoenolpyruvate C , plants are able to absorb CO., from much lower carboxykinase in BSC and PEP is restored to MC. concentrations than C3 plants. Thus, the C i mech- Pyruvate contributed by BSC is regenerated to anism concentrates effectively the intracellular CO2 PEP in MC by an enzyme, pyruvate dikinase, which and keeps photosynthetic carbon fixation running, requires ATP and inorganic phosphate (Pi ) to produce even if activity of RuBPCase would gradually de- PEP, AMP, and pyrophosphate (PP; ). PP, is hydro- crease due to unfavorable external conditions. C lyzed into two P, by a pyrophosphatase, and the photosynthesis, therefore, results in notably lower AMP present reacts with another molecule of ATP CO., compensation points (about 5 pi, ) than to form two molecules of ADP. This powerfully the 50 p,L •1,- ' determined for average C 3 plants. exothermic reaction converting, in sum, two ATP Though C plants usually show higher photo- into two ADP, runs strongly in the direction of PEP synthetic rates and remarkable productivity, they synthesis and, therefore, greatly facilitates the initial are not necessarily more efficient than Ca plants carboxylation step mediated by PEPCase. relying exclusively on the RPP cycle. In fact, C , The transfer of aspartate or malate, or in some plants consume more assimilatory power for the species both, allows classification of the C plants. fixation of each molecule of CO2 . In addition to The basic function of these transferred substrates three ATP and two NADPH as the usual investment is to serve as a CO., source in the BSC on decarboxy- into the RPP cycle (C 3 photosynthesis in BSC), lation. The so-formed CO., is immediately fixed in the C4 mechanism preceding the C3 pathway requires BSC chloroplasts via RPP cycle by the reactions as two further ATP for the biochemical operation of detailed above (C 3 photosynthesis). the MC, i.e. continuous supply of PEP. Therefore, The physiological and biochemical significance C, plants can simply make better use of the excess of C photosynthesis is manifold. Plants possessing light available in their environment to drive the C the C, machinery are capable of much higher rates machinery for the concentration of CO 2 , and to of photosynthesis. As PEP Case exhibits a much maintain a higher level of inorganic carbon at the higher affinity for CO., (K„, ((.02 , about 20-70 44) local environment of RuBPCase in the BSC, where

50 it can be more rapidly fixed. The C, pathway is by 13-CARBOXYLATION VIA nomeans an alternative to the RPP cycle, as it PHOSPHOENOLPYRUVATE CARBOXYKINASE virtually results in no net reduction of the primarily fixed CO.,. There are several exclusive criteria such as specialized leaf anatomy, low CO., compensation Crassulacean acid metabolism — Certain points, low photorespiration, less extreme '2C/ 1"C succulent plants possessing voluminous central carbon isotope fractionation, primary CO., fixation vacuoles and a population of chloroplasts in the same into C, dicarboxylic acids, and a particular set of cells increase their acid content in the night by ac- enzymes that must be fulfilled for unequivocal dis- cumulating malate in the vacuole(s), which, how- tinguishing of C 1 photosynthesis in any plant spe- ever, disappears during the subsequent day. Thus, cies. Unfortunately, in some cases only one of these the malate content of the vacuoles undergoes a distinct basic features has been used to establish C 1 photo- diurnal rhythm, usually accompanied by an inverse synthesis. This highly specialized pathway of carbon rhythm of polyglucan content. This type of metab- (pre-)fixation has been claimed for such diverse olism is termed Crassulacean acid metabolism (CAM) photosynthesizing plants as blue-green algae (Mohler because it has first been discovered in members of 1974), seaweeds in general (Joshi and Karekar this angiosperm family. CAM involves the synthesis 1973), as well as a marine grass (Benedict and Scott of malate by /3-carboxylation mediated through 1976). These reports have caused considerable con- PEPCase at night, and the degradation of this C, acid fusion in the (reviewing) literature and certainly during daytime to liberate CO., for fixation via RPP require some further remarks. cycle (Kluge and Ting 1978). CAM in many regards The conclusion that the C, pathway of photo- ressembles C, photosynthesis. The underlying reac- synthesis occurs in the blue-green alga Anacystis tions are basically the same. Again, PEPCase is con- nidulans was based on the observation of considerable cerned with harvesting CO, by carboxylation of PEP "C labeling of aspartate. Furthermore, this species into OAA, which is subsequently reduced to malate by was found to show activity of PEPCase. When ' 1 C- a NADH-linked malate dehydrogenase. Carboxyla- labeled exogenous aspartate was supplied, this tion and reduction take place in the cytoplasm. In the compound was directly converted into other amino light, malate is released from the vacuoles into the acids rather than into intermediates of the RPP cycle cytoplasm and is there decarboxylated to yield free (Dôhler 1974). There is no doubt that aquatic plants CO., and a G, residue. One group of CAM plants including algae incorporate appreciable amounts of performs an oxidative decarboxylation by malic en- photosynthetically fixed "C into C, compounds such zyme (NADP- or NAD- linked). Others show high as aspartate and malate. This labeling pattern simply activities of phosphoenol- pyruvate carboxykinase. In requires 0-carboxylation of a G, acid, but as such such species malate is first oxidized to OAA and then is far away from, and not necessarily identical with, split into CO, and PEP. CO 2 is reintroduced into C, photosynthesis. photosynthesis. In a variety of seaweeds, including Ulm lactuca Although CAM enables certain plant species and some further representatives of the major classes, to cover their CO., demand by nocturnal CO, fixation, the occurrence of C, photosynthesis was concluded it does not provide an alternative to the RPP cycle on the basis of measurements of PEPCase activity either. In essence, the G, cycle, involving carboxy- (Joshi and Karekar 1973). However, in all marine lation via RuBPCase, is the central event and abso- algae so far investigated in detail, this enzyme is lutely indispensible for net carbon assimilation. As present in negligibly low amounts, insufficient to in C, photosynthesis, the /3-carboxylation represents account for a CO.,-concentrating mechanism. Apart some kind of a CO 2 supercharger. The implication from that, PGA is the earliest "C-labeled photo- of this preset reaction step solves the problem of synthate in seaweeds containing usually more than how to effect maximum carbon gain with a minimum 80% of total "C assimilated during short-term photo- of water loss. Behind closed stomata, CO 2 is recycled synthesis (< 5-10 s) (Kremer and Willenbrink 1972; from decarboxylation of malate and fed into the RPP Kremer and Ktippers 1977). cycle, while inversely the stored photosynthate Undoubtedly, extreme caution is required when provides the source for the generation of sufficient single characteristics are used to establish the presence amounts of the primary CO., acceptor PEP. of C photosynthesis. In essence, if critically eval- In CAM plants, /3-carboxylation of PEP and uated, so far there is no experimental evidence carboxylation of RuBP are not spatially (as in C strongly suggesting the occurrence of C, photosyn- plants: MC vs. BSC), but temporally separated. thesis in unicellular algae or in seaweeds. It has not CAM does not confer high rates of photosynthesis. (yet) been demonstrated that radiocarbon label in In this regard, it is remarkably inefficient, but it the C position (0-carboxyl) of. malate and/or permits the continuation of photosynthesis under aspartate is transferred quantitatively to the C, posi- extremely unfavorable environmental conditions. tion of PGA through decarboxylation and subsequent

51 RuBP H2O-0-e) O=0 3-PGA 2-PGA PEP OAA Asp HC-OH •C00- .000- * COO- •çoo- 'coo- HC-OH I-1-011 —0-> EI-0-(9 >0'0 O 1-9\* H26-0-e H2C-0{0 — HP-OH– — CH2 CH2 CH2 *Ù00- *000- *CO2 *002

FIG. 4. 0-Carboxylation in the light. Reactions provide "C labeling in a- and 0-carboxyl of aspartate. For abbreviations see text.

refixation, as it is in C, plants as well as CAM plants. became evident that maximum 'C labeling of the Moreover, it is rather unlikely to expect that such C dicarboxylic acids in the light is achieved dis- a specialized metabolic apparatus, which certainly tinctly later than in PGA. Furthermore, chemical provides distinct advantages under peculiar ecological degradation of [ N C]aspartate labeled in the light conditions as given in arid or xeric terrestrial envi- showed that ''C is located in both C, and C, atoms. ronments, has been evolved independently in phyto- This suggests that the substrate of p-carboxylation plankton or seaweeds. In the environment of these already contained the 'C labeling in the a-carboxyl plants there is no actual need to possess biochemical (C, atom) when it was subjected to a further carboxy- mechanisms for the concentration of intracellular lation step by PEPCKase. Hence, the Ca acid PEP CO2 . is derived from photosynthetically labeled PGA, As appreciable ''C labeling in aspartate and and it must be concluded from the ''C labeling radioactive malate is observed even after short-term photosyn- pattern in the early photosynthates that of the RPP cycle thesis in marine algae, p-carboxylation, to account carbon flows from an intermediate for this labeling pattern, must be regarded as a carbon (PGA) to compounds usually associated with C in fixation mechanism supplementing carboxylation for metabolism (aspartate and malate), but, at least initiation of the RPP cycle. However, the metabolic the representatives of brown macroalgae, under no dark, the reactions and conversions involved should better be circumstances in the opposite sense. In the associated with the presumably less ambiguous term substrate of PEPCKase, PEP, is provided by gly- C, metabolism (Benedict 1978). colytic catabolism of reserve carbohydrate. Thus, p-carboxylation can proceed in the light and in the This phenomenon has attracted much attention dark with comparable efficiently (Kremer 1981a, b) during the last decade. Pioneering work was under- (see Fig. 4). taken by Akagawa et al. (1972) who showed that Findings from marine phytoplankton, particu- the enzyme performing p-carboxylation in seaweeds larly diatoms, also strongly suggest that /3-carbox- is a phosphoenolpyruvate carboxykinase (PEPCKase). ylation must be regarded as a notable source of Originally, this enzyme was understood to be re- carbon in addition to CO., fixation provided by the sponsible for dark fixation (light-independent fixation) RPP cycle. As has been shown in experiments with of CO.,. Carbon uptake and fixation into organic Phaeorlactyluin cornuntin and Skeletonenza costatuni, compounds during darkness is a special feature of designed to analyze different parameters of carbon macrophytic brown algae. In representatives of the input such as long- and short-terni patterns of "C Laminariales the rates of this nonphotosynthetic CO 2 assimilation, as well as sensitivity of photosynthesis to incorporation may account for about 30% of the environmental oxygen, a remarkable diversity in CO., assimilation rates achieved under photosynthetic fixation is found in marine algae (see Beardall and conditions at the same temperature and carbon supply. Morris 1975; Glover et al. 1975; Beardall et al. 1976; Particularly in young, differentiating frond areas Mukerji et al. 1978). According to this work, the p-carboxylation due to high activity of PEPCKase carboxylation of C3 compounds to C dicarboxylic is significantly eihanced as compared with mature, acids is to be regarded as an important component of differentiated frond tissue (Kremer and Küppers photosynthetic carbon metabolism supplementing the 1977; Küppers and Kremer 1978). universally occurring reactions of the RPP cycle. "Dark" carbon fixation yielding ''C labeling Because diatoms and dinoflagellates exhibit consi- of aspartate and malate upon /3-carboxylation of PEP derable rates of carbon fixation in the dark due to by PEPCKase is obviously not restricted to dark continuation of /3-carboxylation, their overall carbon periods, but also occurs under photosynthetic con- strategy might be strictly comparable to that of the ditions. From kinetic tracer studies using "C it brown seaweeds. 52 Some discussion has arisen concerning the phyceae). I. Effect of nitrogen deficiency and light nature of the enzyme mediating p-earboxylation in intensity. J. Phycol. 11: 424-429. phytoplankton species. Although Mukerji and Morris GRAHAM, D., AND R. M. SMILLIE. 1976. Carbonate (1976) report the extraction of a PEPCase from dehydratase in marine organisms of the Great Barrier Phaeodactylum cornutum , Holdsworth and Bruck Reef. Aust. J. Plant Physiol. 3: 113-119. (1977) and Kremer and Berks (1978) presented results HARRIS, G. P. 1978. Photosynthesis, productivity and strongly suggesting the participation of a PEPCKase. growth. The physiological ecology of phytoplankton. Arch. Hydrobiol. Suppl. 10: 1-171. This qualification is now accepted (see Davies 1979). HATCH, M. D., AND C. R. SLACK. Whether the carbon 1966. Photosynthesis flow from RPP cycle interme- by sugarcane leaves. Biochem. J. 101: 103-111. diates to C 1 compounds during /3-carboxyl ation in HOLDSWORTH, E. S., AND K. BRUCK. 1977. Enzymes the light is also similar to that evaluated for members concerned with f3-carboxylation in marine phyto- of the Phaeophyceae awaits future investigation. plankter. Purification and properties of phosphoenol- pyruvate carboxykinase. Arch. Biochem. Biophys. 182: 87-94. HORECKER, B. L., P. Z. SMYRNIOTIS, AND H. KLENOW. 1953. Characterization of a transketolase from spinach References leaves. J. Biol. Chem. 218: 769-783. JOLLIFFE, E. A., AND E. B. TREGUNNA. 1970. Studies AKAGAWA, H., T. IKAWA, AND K. NISIZAWA. 1972. on HCO ion uptake during photosynthesis in benthic The enzyme system for the entrance of CO2 in the marine algae. Phycologia 9: 293-303. dark fixation of brown algae. Plant Cell Physiol. 13: JOSHI, G. V., AND M. D. KAREKAR. 1973. Pathway 999-1016. of ''CO fixation2 in marine algae. Proc. Indiana Acad. ARNON, D. E. 1952. Glyceraldehyde phosphate dehy- Sci. B 39: 489-493. drogenase of green plants. Science (Washington, KLUGE, M., AND I. P. TING. 1978. Crassulacean acid D.C.) 116: 635-637. metabolism. Ecological studies. Vol. 30. Springer BASSHAM, J. A. 1979. The reductive pentose phosphate Verlag, Berlin, Heidelberg, New York. p. 1-180. cycle and its regulation, p. 9-30. In M. Gibbs and KORTSCHAK, H.P., C. E. HARTT, AND G. O. BURR. E. Latzko [ed.] Encyclopedia of plant physiology. 1965. Carbon dioxide fixation in sugarcane leaves. New Series. Vol. 6. Springer Verlag, Berlin, Hei- Plant Physiol. 40: 209-213. delberg, New York. KREMER, B. P. 1978. Determination of photosynthetic BASSHAM, J. A., AND M. CALVIN. 1957. The path of rates and '''C photoassimilatory products of brown carbon in photosynthesis. Prentice Hall Inc. p. 1-107. seaweeds, p. 269-283. In J. A. Hellebust and J. S. BASSHAM, J. A., AND G. H. KRAUSE. 1969. Photosyn- Craigie [cd.] Handbook of phycological methods. thesis: energetics and related topics. Biochim. Bio- Vol. II. Cambridge University Press. phys. Acta 189: 207-221. 1981a. Metabolic implications of non-photo- BEARDALL, J., AND I. MORRIS. 1975. Effects of envi- synthetic carbon assimilation in brown seaweeds. ronmental factors on photosynthesis patterns in Phycologia 20. (In press) Phaeodactylum tricornutum (Bacillariophyceae). II. 198 lb. Aspects of carbon metabolism in marine Effect of oxygen. J. Phycol. 11: 430-434. macroalgae. Oceanogr. Mar. Biol. Annu. Rev. 19: BEARDALL, J., D. MUKERJI, H. E. GLOVER, AND I. 41-94. MORRIS. 1976. The path of carbon in photosynthesis KREMER, B. P., AND R. BERKS. 1978. Photosynthesis by marine phytoplankton. J. Phycol. 12: 409-417. and carbon metabolism in marine and freshwater BENEDICT, C. R. 1978. Nature of obligate photoauto- diatoms. Z. Pflanzenphysiol. 87: 149-167. trophy. Annu. Rev. Plant Physiol. 29: 67-93. KREMER, B. P., AND U. KÜPPERS. 1977. Carboxylating BENEDICT, C. R., AND J. R. SCOTT. 1976. Photosynthetic enzymes and pathway of photosynthetic carbon assimi- carbon metabolism of a marine grass. Plant Physiol. lation in different marine algae - evidence for the 57: 876-880. C I pathway? Planta 133: 191-196. BLACK, C. C., AND M. M. BENDER. 1976. '3C values KREMER, B. P., AND J. WILLENBRINK. 1972. CO2 - in marine organisms from the Great Barrier Reef. Fixierung und Stofftransport in benthischen marinen Aust. J. Plant Physiol. 3: 25-32. Algen. I. Zur Kinetik der "CO2-Assimilation bei COOMBS, J. 1979. Enzymes of C I metabolism, p. 251-262. Laminaria saccharina. Planta 103: 55-64. In M. Gibbs and E. Latzko [ed.] Encyclopedia of KÜPPERS, U., AND B. P. KREMER. 1978. Longitudinal plant physiology. New Series. Vol. 6. Springer profiles of CO2 fixation capacities in marine macro- Verlag, Berlin, Heidelberg, New York. algae. Plant Physiol. 62: 49-53. DAVIES, D. D. 1979. The central role of phosphoenol- Lame, J. E. 1977. Plant carbonic anhydrase. Life Sci. pyruvate in plant metabolism. Annu. Rev. Plant 20: 393-406 , Physiol. 30: 131-158. MUKERJI, D., H. E. GLOVER, AND I. MORRIS. 1978. DÜHLER, G. 1974. C i-Weg der Photosynthese in der Diversity in the mechanisms of carbon dioxide fixation Blaualge Anacystis nidulans. Planta 118: 259-269. in Dunaliella tertiolecta (Chlorophyceae). J. Phycol. GLOVER, H., J. BEARDALL, AND I. MORRIS. 1975. 14: 137-142. Effects of environmental factors on photosynthetis MUKEIUI, D., AND I. MORRIS. 1976. Photosynthetic patterns in Phaeodactylum tricornutum (Baccillario- carboxylating enzymes in Phaeodactyluin tricornu-

53 nun. Assay methods and properties. Mar. Biol. 36: of carboxylated ribulose-1,5-diphosphate. Biochim. 199-206. Biophys. Acta 297: 165-173. PRÉZELIN, B. B. 1981. Light reactions in photosyn- SKIRROW, G. 1975. The dissolved gases — carbon thesis. Can. Bull. Fish. Aquat. Sci. 210: 1-43. dioxide, p. 1-192. In J. P. Hiley and G. Skirrow QUAYLE, J. R., R. C. FULLER, A. A. BENSON, AND [cd.] Chemical oceanography. Academic Press: New M. CALVIN. 1954. Enzymatic carboxylation of York. ribulose diphosphate. J. Am. Chem. Soc. 76: 3610– SMITH, B. N., AND S. EPSTEIN. 1971. Two categories 3611. of '"CP2 C ratios for higher plants. Plant Physiol. RAVEN, J. A. 1970. Exogenous inorganic carbon sources 47: 380-384. in plant photosynthesis. Biol. Rev. 45: 167-221. TRENCH, R. K. 1979. The cell biology of plant–animal SidDIN, B., AND A. VESTERMARK. 1973. The enzymatic symbiosis. Annu. Rev. Plant Physiol. 30: 485-531. formation of a compound with the expected properties

54 Respiration and Photorespiration

JOHN A. RAVEN AND JOHN BEARDALL Department of Biological Sciences, University of Dundee, Dundee DDI 4HN, Scotland, U.K.

Introduction from the 02 uptake component of the RuBPc-o and PCOC reactions, as they are both strictly light A number of processes in phytoplankters run dependent. The oxygenase function of the primary counter to the main energy-dependent process of carboxylase of photosynthesis (RuBPc-o) is com- photosynthesis. Photosynthesis involves the genera- petitive with the carboxylase function, and is 02 tion of "high-energy intermediates" (reductant, consuming. The phosphoglycolate produced in this proton gradient, ATP) with the evolution of O., and, reaction is converted to glycolate which can be ultimately, the consumption of the "high energy excreted, or metabolized via the PCOC, with intermediates" in the generation of reduced organic further uptake of 02 and evolution of CO.,. The final compounds from CO.,. The processes which run respiratory process, and the one that is most remote counter to this reductive, energy-storing trend of from photosynthesis, is that complex of reactions photosynthesis are included under the headings of subsumed under the heading of "dark respiration"; "photorespiration" and "dark respiration," and as its name implies this process can occur in the involve the uptake of 02 in the oxidation of some dark but may also be important in the light. The reduced compound and (often) the evolution of CO2 . dark respiratory reactions have a more clearly under- A simplified scheme which relates these various stood function than do the photorespiratory reactions, "respiratory" processes to photosynthesis is given in that they supply the NADPH, ATP, and C skeletons in Fig. 1. The respiratory process closest to the required for growth and maintenance processes in photochemistry of photosynthesis (and often con- all aerobic organisms. sidered as a part of photosynthesis) is the re-oxidation The approach which we shall adopt in analyzing of redox intermediates of the "light reactions" of the role of these respiratory processes in the life of photosynthesis (Mehler reaction = pseudocyclic phytoplankton organisms will be a quantitative one. electron flow = 02 -linked noncyclic electron flow). We shall see that there are considerable difficulties Experimentally this process is difficult to distinguish in quantifying the various O., uptake and CO., evo-

Radiationless loss Fluorescence

Absorbed 2 H 0 0 „11 2

processes Thylakoid Me hier AT P Re>ict a nt Ç 2 H20 Reaction

CO2 922 Photorespirat r on

glyc RuBPo/ Reductant A TP C skeletons PCRC pGA Pglyc PCOC co CO 2 Dark 2 Ru BP respiration { Carbohydrate

FIG. I. Outline of the pathways of energy, carbon, and oxygen in photosynthesis, photorespiration, dark respiration, and growth of a phytoplankter. No attempt is made to represent stoichiometries. (Abbreviations: Glyc, glycolate; PCOC, photorespiratory carbon oxidation cycle; PCRC, photosynthetic carbon reduction cycle; PGA, 3-phosphoglycerate; Pglyc, phosphoglycolate; RuBP, ribulose bisphosphate; RuBPo, ribulose bisphosphate oxygenase.

55 lution processes, particularly under natural conditions Mehler Reaction and in the light when the opposing photosynthetic (Pseudocyclic Electron Flow) and respiratory fluxes must be distinguished. Having quantified the metabolic fluxes of a, and CO2 on Once the problems of extracellular diffusion the basis of biomass or growth rate, we can then barriers to 0 2 exchange had been appreciated (see explore the relationship of the processes to growth Jackson and Volk 1970), experiments on tracer 0 2 and maintenance; for this analysis three criteria will exchange in microalgae consistently showed a light- be used. One is the impact of the process on the stimulated 0 2 uptake; this process occurred in all growth efficiency in terms of running costs; using of the algal classes tested (Bacillariophyceae: Bunt the commonly limiting resource of light as an exam- 1965; Bunt et al. 1966; Chlorophyceae: Brown and ple, we can ask what influence the process has on Weis 1959; Hoch et al. 1963; Fock et al. 1980; the quantum requirement for growth. Pseudocyclic Radmer and 011inger 1980; Chrysophyceae: Weis electron flow strictly coupled to ATP synthesis, and and Brown 1959; Euglenophyceae: Gerster and generation of NADPH and ATP in tightly coupled Tournier 1977; Cyanophyceae (): dark respiration, are examples of the generation of Hoch et al. 1963; Owens and Hoch 1963; Lex et al. compounds essential for growth and maintenance 1972; Radmer and Kok 1976). This light-stimulated involving a certain minimum quantum input per unit 02 uptake is underestimated by the tracer 0 2 tech- of growth; "uncoupled" operation of these pathways, nique to the extent that 0 2 is recycled between the and most manifestations of RuBPo and the PCOC 02 evolution and 02 uptake sites without becoming represent a use of quanta which is not essential for available to the sensor (mass spectrometer inlet) growth under all circumstances. A second criterion employed; this underestimate is the sanie in absolute in natural selection is the efficiency with which terms for 0 2 uptake and 02 evolution, but thefrac- capital resources are deployed: we may ask what tional underestimation is larger for the process of investment of potentially limiting resources (light, smaller magnitude, i.e. 02 uptake. N, P, etc.) is involved in producing the metabolic This light-stimulated 0 2 uptake in photosynthe- machinery for the various pathways of respiration, tically competent algae depends on the functioning and what would be the investment for such alternative of photosystem II: this is shown by studies of action pathways as may serve the same end (e.g. photo- spectra, photosynthetic mutants, and the effects of phosphorylation rather than oxidative phosphory- photosynthetic inhibitors (Govindjee et al. 1963; lation as a source of ATP in the light for processes Hock et al. 1963; Owens and Hoch 1963; Rafter other than the photosynthetic production of carbo- and 011inger 1980). The involvement of photosystem hydrates). A third consideration is that of safety and one is likely, although studies of the "Emerson stability; do the respiratory processes act as stabilizing Enhancement Effect" (the classical method of dem- or as destabilizing processes with respect to (for onstrating the involvement of two photosystems in example) the dissipation of light energy absorbed plant photosynthesis) show negative rather than in excess of what can be used in growth under a positive effects (Govindjee et al. 1963; Owens and certain set of conditions? The Mehler reaction could Hoch 1963). These steady-state effects can be inter- act to dissipate excess energy which would otherwise preted in terms of interactions of cyclic and pseudo- generate triplet chlorophyll, but it can also serve cyclic redox processes (see Glidewell and Raven to generate toxic O., derivatives (0;7, H 2 02 ). 1975, 1976) and an involvement of photosystem Recent reviews of the topics covered in this one in the light-dependent 0 2 uptake; this view is article include Jackson and Volk (1970), Raven supported by net 02 exchange phenomena under (1976a), and Foyer and Hall (1980) (Mehler reaction transient conditions (French and Fork 1960), and by (Mehler and Brown 1952) and pseudocyclic photo- analogy with studies on subcellular systems from phosphorylation), Merrett and Lord (1973), Tolbert higher plants (Elstner and Konze 1976; Foyer and (1974), Andrews and Lorimer (1978), Lorimer et Hall 1980). al. (1978a), Heber and Krause (1980), and Raven In general the tracer 0 2 uptake in the light (1980a, b) (glycolate metabolism, RuB Po, PCOC, exceeds that attributable to the sum of dark respiration photoinhibition, inorganic C concentration mecha- and of RuBPo/PCOC activity. As will be seen later nism), and ap Rees (1974), Lloyd (1974a, b), Lloyd (see Table 1) the 0 2 uptake capacity of dark respi- and Turner (1980), and Raven (1971, I972a, b, ration is relatively limited (especially in cyanobacteria 1976a, b) (dark respiration). Where important com- and nonmotile eukaryotes), so even if this process parative considerations are involved, or data on algae were proceeding at its maximum rate in the light are lacking, the metabolism of other organisms will (cf. the blue light stimulation of algal respiration: be discussed. Heterotrophy (Droop 1974; Nielson Sargent and Taylor 1965) it could not account for and Lewin 1974) and the role of respiratory incapacity the observed 02 uptake rate in the light. The con- in obligate autotrophy (Whittenbury and Kelly 1977) tribution of RuBPo/PCOC to this light-stimulated will not be discussed in detail. 02 uptake is limited by the relatively high exogenous 56 TABLE 1. Comparison of some properties of the major 02 -consuming processes in phytoplankton organisms (data from Sargent and Taylor 1972; Radmer et al. 1978; Raven and Glidewell 1978; Palmer 1979; Badger 1980; Fock et al. 1981).

Oxygen-consuming reaction

Mehler reaction (pseudocyclic Cytochrome Alternate Property electron flow) RuBPo oxidase oxidase

V max V„,„, for gross 0.15 of Typically 0.1 As for cytochrome (catalytic oxygen evolution for RuBPc or of V„,„, of gross oxidase, capacity at gross 02 02 evolution but not additive substrate evolution (range 0.02-0.3) with it saturation)

btA4 100-150 500-1000 2 6

In vivo effect Absolute Absolute Variable Variable of full noncyclic requirement requirement (inhibits or (inhibits or chain" stimulates) stimulates)

Inhibition by HCN No Yes Yes No

Inhibition by substituted hydroxamic acids (No) (No) No Yes

In organisms using H2 0 as redox donor to photosynthesis rather than those adapted to H2 S or H2.

CO, levels employed in many experiments (particu- ('2 fLM) (Sargent and Taylor 1972; see Table 1). larly the earlier ones), and by the occurrence of a Thus we may attribute the great majority of the low- "CO, concentrating mechanism" (see below) in affinity 02 uptake to a Mehler reaction in vivo. algae grown at low CO, concentrations; these two The low affinity found for these algae is consistent factors together mean that RuBPo activity is con- with the 00 affinity in Hydrodictyon africanum siderably suppressed by the competitive inhibitor for phosphate uptake energized ultimately by ATP CO2. Thus a substantial fraction of the observed from pseudocyclic photophosphorylation (Raven light-stimulated 02 uptake is most reasonably attri- and Glidewell 1975a), but much higher affinities buted to the Mehler reaction. are found in isolated higher plant chloroplasts (e.g. The extent of the Mehler reaction under natural Asada and Nakano 1978) for the Mehler reaction. conditions is probably substantially greater than that A major discrepancy exists between measure- reported in many of the papers cited above as ments conducted in different laboratories with respect the half-saturation constant for 00 is quite high and to the magnitude of the Mehler reaction relative to (for technical reasons) many of the earlier experiments gross photosynthetic 02 evolution. Radmer and used 02 concentrations well below the for 02 associates (Radmer and Kok 1976; Radmer et al. in the Mehler reaction. Recent estimates of the 1978; Radmer and 011inger 1980) find that there for 00 uptake in algal cells (using tracer 02 ) gave is a constant gross 00 evolution with varying CO2 values of 8-11% and 12.5% 02 in the gas phase fixation rates, with any decrease in electron flow for Scenedesmus obliquus and Chlamydomonas to CO2 caused by limiting CO, concentration, or reinhardtii, respectively, corresponding to 100-150 by inhibition of the PCRC, being compensated aiiM in solution at 25°C (Radmer et al. 1978; Fock quantitatively by increased 02 uptake. The 02 uptake et al. 1981). These values were obtained under in the light varies from about 0.1 of gross 02 evolution conditions in which RuBPo activity was only suf- when the PCRC is operating at its maximum rate ficient to account for a small fraction of the 00 to 1.0 of gross 02 evolution at the CO2 compensation uptake, and the K+for 02 in the light is much higher point or in the presence of PCRC inhibitors. By than that for dark respiration in algae via either the contrast, other recent work (Fock et al. 1981), in alternate oxidase (— 6 1.2,M) or cytochrome oxidase agreement with earlier results, shows that 02 uptake

57 in the light does not completely compensate for generated in noncyclic photophosphorylation by the decreased PCRC activity as an electron sink, i.e. external pyridine nucleotide dehydrogenase of mito- gross 02 evolution falls as the rate of CO., fixation chondria (Fig. 5) with a P/e., of 2 in oxidative falls, and while 02 uptake may increase at low CO 2 phosphorylation added to the P/e., of 1.33 in non- fixation rates it does not quantitatively compensate cyclic photophosphorylation to give a quantum yield for the reduced availability of CO., as electron of 0.83 does not seem to have been realized in nature. acceptor. The inefficiency of pseudocyclic ATP generation The reasons for these discrepancies are not may seem to militate against its occurrence during clear; despite the large quantitative variation in the growth of algae under light-limiting conditions where magnitude of the Mehler reaction estimated by it is often difficult to balance ATP and reductant different workers, it is possible to discuss the pos- requirements for growth with quantal input even with sible roles on this process under the headings pro- the most efficient mechanisms of energy conversion posed in the introduction. (Raven 1976a; cf. Pirt et al. 1980). However, there Dealing first with running costs, there is evi- is considerable evidence that the Mehler reaction dence that the Mehler reaction is at least facul- does occur under light-limiting conditions (e.g. al. 1981) to an extent tatively coupled to ATP synthesis in vivo (see Raven Hoch et al. 1963; Fock et and Glidewell 1975a; Raven 1976a; Gimmler 1977). which could (granted complete coupling) supply It is thus a possible contender (with cyclic photo- much of the "extra" ATP needed for growth. phosphorylation and oxidative phosphoiylation) for Turning to considerations of capital costs for the the role of supplier of any "extra" ATP used in Mehler reaction, we may again compare its economics photosynthetic production of reduced carbon, and (as a means of generating "extra" ATP) with that other ATP-requiring growth and maintenance pro- of cyclic and oxidative phosphorylation. Because it cesses in illuminated photosynthetic cells. By "extra" has a quantum yield which may be only half of that ATP is meant ATP in addition to that supplied by of cyclic photophosphorylation, the generation of noncyclic photophosphorylation during the gener- a certain quantity of ATP per unit time by cyclic ation of the reductant for CO 2 fixation and for the photophosphorylation requires only half the invest- reduction of N0,7• and SO1 . Current opinion (cf. ment in antenna chlorophyll, and in reaction centers, Raven 1976a; Rosa 1979) seems to favor a P/e2 than does pseudocyclic photophosphorylation. Again, ratio in noncyclic (and pseudocyclic) photophospho- cyclic photophosphorylation appears to be superior rylation of less than 1.5 (Kaplan et al. 1980b), so under light-limiting conditions. The other alternative that "extra" ATP is needed even for carbohydrate for the generation of "extra" ATP, i.e. oxidative production in "pure PCRC" photosynthesis when phosphorylation following complete photosynthesis, 3 ATP are needed per 4 electrons (2 NADPH) used seems even more inefficient with respect to capital in CO2 reduction. If pseudocyclic ATP synthesis investment than does pseudocyclic photophos- has a maximum quantum yield of 0.33 ATP/quantum phorylation, as all the mitochondrial machinery is (1.33 ATP/e2 , 2 quanta per electron), we can needed as well as the thylakoid and stroma apparatus, compare its efficiency with that of cyclic photo- thus doubling the cost of a given rate of ATP synthesis phosphorylation and oxidative phosphorylation. If the in terms of energy, , C, or N invested (Raven 1976a, in cyclic electron flow is 4, and 1 quantum b, 1980a, b). However, as pointed out by Raven absorbed by photosystem one moved I electron, an (1976a), if the ATP-using as well as the ATP- 1-1-1- /ATP of 3 in the ATP synthetase will give a producing investments are considered, oxidative maximum quantum yield of 0.67 ATP/quantum, in phosphorylation is less costly, as the ATP-using accord with in vivo data (Raven 1976a, 1980a, b; cf. machinery can, in principle, be used over a full Olsen et al. 1980). Oxidative phosphorylation yields 24-h light—dark cycle when oxidative phosphoryla- (see below) 5.60 ATP per C (at the redox level of tion is the ATP source whereas the direct use of carbohydrate) oxidized, so with the minimum quan- photophosphorylative ATP means that the ATP-con- tum requirement for CO, fixation of 8.5 quanta/CO., suming machinery is standing idle in the dark (cf. (3 ATP and 2 NADPH generated by a mixture of Foy et al. 1976; Foy and Smith 1980; van Liere et al. noncyclic and cyclic photophosphorylation), the 1979). maximum quantum yield is 0.66 ATP/quantum. The final point for consideration is that of safety; Thus, granted optimal efficiency in coupling of ATP does the process under consideration produce toxic synthesis to redox reactions and in diverting excita- compounds which some other process is required tion energy to the most needy photoreaction (Ried and to detoxify, or is the process itself a detoxifying Reinhardt 1980), it is clear that cyclic photophos- procedure? For the Mehler reaction the answer seems phorylation and oxidative phosphorylation are each to be "both"; it is involved with the potentially about twice as efficient as pseudocyclic photophos- dangerous outcome of having more excitation energy phorylation. The option of reoxidizing the reductant supplied to the photochemical apparatus than can

58 organisms (on either a growth rate or a biomass inter alia be used in the energy-requiring processes of growth basis), and their detoxification involves superoxide dismutase and catalase (Halliwell 1974; and maintenance, and it can act as a sink for possibly Elstner and Konze 1976; Foyer and Hall 1980). dangerous accumulations, of excited states of chloro- A final aspect of the functioning of the Mehler phyll at the expense of generating toxic, partly H2 02). The "energy-sink" role of the Mehler reaction reaction is the possibility that, in addition to cata- would, ideally, only come into play at irradiances lyzing photophosphorylation, it can also have a in excess of those at which light is limiting the regulatory role in "redox poising" and in permitting reduction of the essentials for growth and main- the optimal activity of the cyclic and noncyclic tenance (when natural selection would favor strictly photophosphorylation pathways (see Ziem-Hanck coupled electron flow in the Mehler reaction); at and Heber 1980, for an up-to-date discussion of this such supersaturating irradiances uncoupled pseudo- hypothesis). cyclic electron flow could act to dissipate excess excitation (in concert with enhanced excretion of Glycolate Synthesis and Metabolism: organic products of photosynthesis related to an Ribulose Bisphosphate Oxygenase excess of photosynthetic rate over growth rate: Fogg and the Photorespiratory Carbon 1975). The finding that the Mehler reaction increases Oxidation Cycle in parallel with CO2 fixation as the light intensity increases up to that required to saturate photosyn- A major potential source of light-dependent thesis, and then continues to increase (in parallel 02 uptake (and, to a lesser extent, CO, evolution) with an increment of gross 02 evolution) as irradiance in illuminated photosynthetic cells is the synthesis is increased to values higher than those needed to and metabolism of glycolate (see Fig. 1). It is be- saturate CO, fixation, is consistent with such an lieved (see Lorimer and Andrews 1973; Andrews "energy-dissipating" role (Brown and Weis 1969; and Lorimer 1978; Somerville and Ogren 1979; Weis and Brown 1959; Hoch et al. 1963; Fock et al. Christen and Gasser 1980; Raven and Glidewell 1980). The findings of Radmer and co-workers 1981) that the major pathway of glycolate synthesis concerning the quantitative replacement of CO, in autotrophic organisms is via the oxygenase activity fixation by 02 uptake in the Mehler reaction when- (RuBPo) of the major carboxylase, ribulose bis- ever the PCRC cannot operate at its maximal rate phosphate carboxylase-oxygenase (RuBPo, EC would constitute an effective energy sink under 4.1.1.39). This enzyme, in addition to catalyzing low-CO, conditions, and accordingly one would the carboxylase activity (RuBPc-o equation (1)): not expect photoinhibition to be manifested at irra- diances similar to those required to saturate CO, (1) Ribulose-1,5-bisphosphate + CO 2 + H2 0 fixation (under CO2 -saturating conditions), regard- —› 2 (3-phosphoglyceric acid) less of the CO2 concentration. A similar argument applies to the "induction phase" of CO, fixation at also catalyzes the oxygenase activity (RuBPo: equa- a dark—light transition, when the high Mehler reaction tion (2)): rate could (if uncoupled) act as an energy sink in addition to supplying (if coupled) the "priming" (2) Ribulose-1,5-bisphosphate + 0 2 > ATP needed during induction (Radmer and Kok 3-phosphoglyceric acid + 1976). However, in cases in which this quantitative 2-phosphoglycolic acid. replacement of CO 2 uptake by 02 uptake at low CO, concentrations does not occur, photoinhibition A report (Branden 1978) that the reactions shown occurs at low CO, concentrations at an irradiance in equations (1) and (2) are catalyzed by different which does not photoinhibit at higher CO, con- enzymes has not been confirmed (McCurry et al. centrations (Fock et al. 1981), suggesting that the 1978). Mehler reaction is an imperfect energy sink under The presence of RuBPo activity in phytoplankton these conditions. organisms has been shown in two ways. One is the The other aspect of safety with respect to the demonstration of the stoichiometry shown in equa- Mehler reaction is its role in generating toxic 0 2 tion (2) when extracted RuBPc-o from microalgae radicals and H2 02 ; indeed, H2O, production has is incubated (after appropriate activation: Lorimer been used as a (qualitative) indicator of Mehler et al. 1977) with RuBP. This procedure has demon- reaction activity (Patterson and Myers 1973; cf. strated RuBPo activity in RuBPc-o from members Radmer and Kok 1976). The generation of these of the Bacillariophyceae (Beardall and Morris 1975), toxic intermediates of the reduction of 02 to H2O Chlorophyceae (Berry and Bowes 1973; Lord and is probably much more quantitatively significant Brown 1975; Berry et al. 1976; Nelson and Surzycki in photosynthetic organisms than in other aerobic 1976a, b), Cyanophyceae (Okabe et al. 1979; Badger 59 not, as yet, available 1980), and Euglenophyceae (McFadden et al. 1979). evidence for this pathway is In the Cyanophyceae the enzymes In vivo demonstration of this activity can be inves- for these algae. in Fig. 2 are present, albeit tigated by the use of "0 2 : the reaction shown in of the pathway shown and Stewart 1973); enzymic equation (2) incorporates one 0 atom from 0 2 into at low activity (Codd inhibitor evidence, however, favors a major the carboxyl group of phosphoglycolate (and, after and pathway shown in Fig. 3 (Codd and the action of phosphoglycolate phosphatase, of role for the Stewart 1973, 1974; Grodzinski and Colman 1975). glycolate), the other 0 atom appearing in H2 0 (Lorimer et al. 1973). Demonstration of a substantial A number of comments about the pathways enrichment of the carboxyl group of glycolate in shown in Fig. 2 and 3 must be made briefly before 18 0 during glycolate synthesis by illuminated algae the quantitative significance of glycolate synthesis in the presence of 1802 has been shown for members and metabolism in relation to the running costs, of the Chlorophyceae (Gerster and Tournier 1977; capital costs, and safety of algal growth are con- Lorimer et al. 1978b; Fock et al. personal commu- sidered. nication) and Euglenophyceae (Dimon and Gerster 1) The enzyme which catalyzes the oxidation 1976; Gerster and Tournier 1977). This in vivo of glycolate to glyoxylate is, in most algae, a gly- demonstration of the role of RuBPo in glycolate colate dehydrogenase rather than the glycolate oxi- synthesis fails to prove that RuBPo is the unique dase found in bryophytes and tracheophytes and source of glycolate in microalgae for two reasons. their green algal (charophyte) ancestors. Glycolate One is that the enrichment of 0 in the carboxyl dehydrogenase is the characteristic glycolate-oxi- group of glycolate is never as high as it is in the dizing enzyme of the Bacillariophyceae (Paul and "Oo supplied; this can be accounted for by dilution Volcani 1974, 1975; Paul et al. 1975), Chlorophyceae of the 1802 at the site of RuBPo activity by photo- (Frederick et al. 1973; Gruber et al. 1974; Stewart synthetic 1602 (Dimon and Gerster 1976; Lorimer and Mattox 1975; Floyd and Salisbury 1977; cf. et al. 1978b). The other problem is that it is implicitly Codd and Schmid 1972; Bullock et al. 1979), Cyano- assumed that RuBPo is the only mechanism of gly- phyceae (Codd and Stewart 1973, 1974; Codd and colate synthesis which leads to "0 incorporation Sellal 1978), and Euglenophyceae (Codd and Merrett from 1802 ; while no other in vitro pathway shows 1971a, b). Glycolate dehydrogenase is a membrane- substantial 180 enrichment in glycolate (Lorimer associated enzyme. In the prokaryotic cyanobacteria et al. 1978b), Gerster and Tournier (1977) showed it is associated with the thylakoid membranes (Codd that 180 from "O., is still found in glycolate syn- and Sellai 1978), while in the eukaryotes it is asso- thesized by Chlorelln and Euglena in the presence ciated with the inner mitochondrial membrane of 1 mM KCN which might be expected to completely (Stabenau 1974a, b; Paul et al. 1975; Paul and Vol- inhibit both the carboxylase and oxygenase activities cani 1976a), although there is also glycolate dehy- of RuBPc-o (see Glidewell and Raven 1975, 1976; drogenase activity associated with microbodies in cf. Vennesland and Jetschmann 1976). Euglena (Collins and Merrett 1975). The oxidation The data reviewed above are consistent with a of glycolate via the dehydrogenase can lead to ATP large fraction of glycolate synthesis being a result synthesis via the "sites" between UQ and 02 (Collins of RuBPo activity in microalgae. Thus we may and Merrett 1975; Paul and Volcani 1975, 1976a; expect an (as yet unquantified) contribution of Paul et al. 1975). The magnitude of this ATP syn- RuBPo to the total 02 uptake in illuminated algae. thesis for the energetics of algae is indicated in Further 02 uptake, and some CO2 evolution, can Table 2. The Charophyceae sensu lato have glycolate result from the further metabolism of glycolate (Fig. oxidase in microbodies; this direct coupling to 02 1, 2, and 3). Enzymic and in vivo labeling data precludes coupled ATP synthesis (Stewart and Mattox support a role for the pathway (photosynthetic carbon 1975; Stabenau 1975, 1980). oxidation cycle or PCOC) shown in Fig. 2 in gly- 2) It is likely that there is some "leakage" colate metabolism in members of the Bacillario- from the pathways shown in Fig. 2 and via addi- phyceae (Paul and Volcani 1976b; Burns 1977; tional CO , release as some glyoxylate is oxidized Coughlan 1977), Chlorophyceae (Pritchard et al. (by glycolate oxidase/dehydrogenase, or by EL 02 ) 1961, 1963; Lord and Merrett 1970; Tolbert 1974; to CO2 plus formate, or to CO., alone (Halliwell Burns 1977) and Euglenophyceae (Codd and Merrett 1978; Grodzinski 1979). 1971a, b). In some members of the Chlorophyceae 3) The glycine to serine step (Fig. 2) occurs (Badour and Waygood 1971a, b, 1972) there is in the mitochondria of eukaryotes; the NADH gen- enzymic and tracer evidence for the pathway shown erated in this step is shown in Fig. 2 as being oxidized in Fig. 3, involving glyoxylate carboligase and by the mitochondrial electron transport pathway with tartronic semialdehyde. The enzymes of this path- ATP generation. The NADH generated in this way way are also present in members of the Bacillario- probably cannot be used to reassimilate the NH 3 phyceae (Paul and Volcani 1976b), although tracer also released in the glycine to serine step within

60 20 AT P

20 RuBP 1 02 2 glyoxylate 7-> 2 glycine 1 02,>< NADH—-1 2 0 2(-NH2) 18 RuBP 2 RuBP ATP) yi 2 Ed' 1 CO2 3 ATP 1,-18 CO2 2 2 glycolate 36 PGA 2 PGA 2 Pglycolate PCOC 11-NH2) 1 serine 39 1 PGA

1 glycerate <-.1 HO - pyruvate ATP 39 ATP + 39 NADPH NADPH 17 reduced

FIG. 2. Integrated PCRC/PCOC cycle for a RuBPc/RuBPo ratio of 9. With the in vitro RuBPc-o kinetics measured for Anabaena variabilis by Badger (1980), a RuBPc/ruBPo ratio of 9 could be achieved at 25°C (using equation (3)) if the "CO, concentrating mechanism" can maintain a steady-state intracellular CO, concentration of 67 it/14; this is consistent with the data of Kaplan et al. (1980a). Per turn of the integrated cycle, there is a net fixation of 17 CO, (18 gross CO, fixed by RuBPc, I CO, produced in the PCOC), 17 0 2 are evolved (20.5 gross 0 2 evolved in reductant generation for PCRC/PCOC, in 3.5 0, taken up in RuBPo/PCOC); 17 reduced C at the level of carbohydrate are produced. The energetics of this cycle, and of variants on it, are given in Table 2. 20 AT P

20 Ru5P 20 RuBP — 2 RuBP 10 2

2 Pglycolate-->.2 glycolate ‘ , >2 glyoxylate 18 RuBP CO2 2 ,PGA PCRC (4 ATP) 36 PGA

39 PGA giycerate< tartronic semi- aldehyde r \ ATP NAOPH 39 Triose P tartronic semialdehyde 17 reduced 39 ATP + COOH 39 NADPH HCOH NCO

Flo. 3. Integrated PCRC/tartronic semialdehyde cycle for a RuBPc/RuBPo ratio of 9. The role of the "CO, con- centrating mechanism" in bringing about a RuBPc/RuBPo of 9 is discussed in the caption of Fig. 2. Per turn of the integrated cycle, there is a net fixation of 17 CO, (18 gross CO, fixed by RuBPc, I CO, produced in the tartronic semi- aldehyde cycle); 17 02 are evolved (20 gross 0 2 evolved in reductant generation for the PCRC/tartronic semialdehyde cycle, 3 02 taken up in RuBPo and glycolate oxidation); 17 reduced C at the level of carbohydrate are produced. The energetics of this cycle, and of variants on it, are given in Table 2. the mitochondria as the glutamate dehydrogenase as reductant for glutamate synthetase. The feasability (at least in higher plants) in the mitochondria is very of using the NADH generated in the glycine to poor at glutamate synthesis (Hartmann and Ehmke serine step in the mitochondria in generating plastid 1980). It is more likely that ammonia assimilation reductant, via dicarboxylate shuttles at the mito- occurs in the plastids, using the glutamine syn- chondrial inner membrane and the inner chloroplast thetase (GS) glutamate synthetase (GOGAT) pathway envelope membrane, is questionable on thermo- (Keys et al. 1978) and photoproduced ferredoxin dynamic grounds (cf. Krebs and Veech 1969; Woo

61 and Osmond 1976; Moore et al. 1977; Wiskich may be disadvantageous. However, there seems to 1977; Heber and Walker 1979). Few data are available be no good reason why tartronic semialdehyde for microalgae in this respect (cf. Syrett 1981). should lealc more than its isomer, hydroxypyruvate, 4) The activity of the enzymes of the pathway which is an intermediate of the glycine-serine path- of Fig. 2 are, in several members of the Chloro- way (Fig. 2). phyceae and Euglena, much higher when the algae Having considered the biochemical potential are grown under "natural" concentrations of CO., of microalgae to synthesize and to metabolize gly- (close to air equilibrium) rather than in air supple- colate by pathways which consume 0 2 and generate mented with 0.5-5% CO., (Pritchard et al. 1961, CO, we now turn to a consideration of the extent 1963; Nelson and Tolbert 1969; Lord and Merrett to which these pathways function in these organisms. 1971; Stabenau 1977). The derepression of these Starting with RuBPc-o, it is possible to predict the enzymes is paralleled by increases in the activity of ratio of RuBPc to RuBPo activity at a given ratio carbonic anhydrase and of catalase (Nelson et al. of the concentrations of the mutually competitive 1969; Graham and Reed 1971; Reed and Graham substrates, 02 and CO2 , using a relationship derived 1977; Stabenau 1977; cf. Kaplan et al. 1980a). by Laing et al. (1974), equation (3): 5) The operation of the pathway shown in Fig. 2 in eukaryotes involves the cooperation of a number vo _ Vo Ke [02 of organelles. In bryophytes and tracheophytes the (3) y,. - V,. K„ [CO2] conversion of phosphoglycolate to glycolate, and the phosphorylation of glycerate, occurs in the where v„ and v„ are the achieved rates of RuBPo and chloroplasts; glycolate is converted to glycine, and RuBPc (measured in terms of specific reaction rates, serine is converted to glycerate, in the microbodies s - ', or other suitable units) at the 0 2 and CO 2 (peroxisomes); the glycine to serine conversion occurs prevailing at the site of RuBPc-o, concentrations in the mitochondria (Tolbert 1974). This distribution respectively, V„ and V,. are the maximal rates of of the reactions is also found in those (charophyte) RuBPo (at saturating 0 2 ) and RuBPc (at saturating green algae which are the probable ancestors of the CO2 ), respectively, and K,. and K„ are the values of higher plants and which have glycolate oxidase for the mutually competitive alternative rather than glycolate dehydrogenase (Stabenau substrates CO., and 02 , respectively. 1975, 1980), although some of the microbody en- Application of equation (3) to a number of C3 zymes (e.g. hydroxypyruvate reductase and catalase, higher plants for which the relevant data (net CO 2 but not glycolate oxidase) are also found in an exchange rates as a function of leaf intercellular organelle distinct from the microbody. In the euka- space CO2 and 02 concentrations, and the in vitro ryotic algae which possess glycolate dehydrogenase kinetic characteristics of RuBPc-o) are available the entire sequence from glycolate to glycerate shows that much of the gas exchange characteristics probably occurs in the mitochondria (Stabenau can be explained in terms of the kinetics of RuBPc-o, 1974a, b; Paul and Volcani 1975, 1976b; Paul et al. assuming that the phosphoglycolate generated by 1975), although some of the hydroxypuruvate re- RuBPo is metabolized as shown in Fig. 2 (Laing ductase activity seems to be in the soluble fraction et al. 1974; Lorimer et al. 1978a; Farquhar et al. of cell extracts (Stabenau 1975, 1980). 1980; Raven and Glidewell 1981). Among the quan- In prokaryotes (Cyanophyceae) the pathway tities predicted are the net CO., fixation rate in air, shown in Fig. 2 or, more significantly, that in Fig. 3, the stimulation of net CO 2 fixation at air levels of probably has glycolate dehydrogenase as the only CO2 by reducing 0 2 to 1% or less (by increasing membrane associated enzyme, the rest being free in gross photosynthesis ye when the competing the cytosol. In the eukaryotic microalgae the location is removed, and by decreasing CO., loss in the PCOC of the glyoxylate to glycerate portion of the tartronic which is running at a much reduced rate corresponding semialdehyde pathway (Fig. 3) is not clear. to a much reduced v„ and hence much slower gly- 6) The tartronic semialdehyde pathway (Fig. 3) colate synthesis), and the magnitude of the CO., seems to have considerable economy of capital in- compensation concentration (the CO 2 concentration vestment compared with the glycine-serine pathway. at which no net CO., exchange occurs and (ignoring (Fig. 2); it is not clear what selective advantage dark respiration) v„ = 2v,.). the more cumbersome glycine-serine pathway The only case in microalgae in which such a might have which causes its dominance, at least good fit between RuBPc-o kinetics in vitro and gas in the higher plants. Tartronic semialdehyde has exchange and glycolate excretion data obtained in been characterized as an extracellular product in vivo has been obtained is that of Chlainydonhanas some algae (Badour and Waygood 1971b), and the reinhardtii grown at high CO2 concentrations (Berry occurrence of a "leaky" intermediate (in addition to and Bowes 1973). Ogawara et al. (1980) found that glycolate and glyoxylate: Stewart and Codd 1980) the Oo inhibition of growth in Chlorella vulgaris

62 grown in the light with 1.0-2.4% CO, and 65% RuBPo activity and enhancement of RuBPc activity 02 could be explained in terms of the competition is achieved by the operation of a "CO2 pump" between RuBPc and RuBPo (equation (3)) using based on an auxilliary carboxylation of PEP and a the kinetic constants for higher plant RuBPc-o; C acid cycle combined with specialized leaf anat- however, as we shall see, the kinetics of the algal omy. Although enhanced p-carboxylation has been enzyme are not identical with those of the higher suggested in some instances (Beardall et al. 1976; plant enzyme. Appleby et al. 1980), the algae which show this When we consider algae grown at "normal" "C g -like" physiology generally have clear-cut C gg CO., concentrations (i.e. a concentration in solution the primary biochemistry, i.e. RuBPc-o catalyzes close to that in equilibrium with air, and thus relevant conversion of inorganic to organic C (Raven and to the vast majority of limnological and oceano- Glidewell 1978; Coleman and Colman 1980b; Raven graphic situations) a very different picture emerges. 1980a). We feel that the most plausible explanation One striking observation is that, although the King, of the "C 3 gas exchange but Cgg biochemistry" para- for the algal enzyme is higher than that for the higher dox lies in the occurrence of a "CO, concentrating plant enzyme, the Ki ggg) , in vivo is lower in air- mechanism" in many air-adapted algae. This is based grown algae than in higher plants; this difference on the active influx of some inorganic C species between the IQ- values in vivo and in vitro in air- at some membrane between the bulk medium which grown algae cannot be explained in terms of a large supplies the inorganic C and the chloroplast stroma excess of the enzyme in the air-grown algae (Whit- (eukaryotes) or the cytosol (cyanophytes) in which tingham 1952; Bidwell 1977; Hogetsu and Miyachi RuBPc-o fixes CO 2 . 1977, 1979; Lloyd et al. 1977; Badger et al. 1978; Direct evidence for such a " CO.,-concentrating Findenegg and Fischer 1978; Badger 1980; Coleman mechanism" has been obtained for Chlainydomonas and Colman 1980a; Shelp and Canvin 1980a, b; reinhardtii (Badger et al. 1977, 1978, 1980; Spalding Beardall and Raven 1981). and Ogren 1980), Anabaena variabilis (Badger et Another discrepancy between the in vivo and in al. 1978; Kaplan et al. 1980a), Dunaliella sauna vitro activities of the algal RuBPc-o in air-grown (Zenvirt et al. 1980), and Chlorella emersonii (Bear- cells lies in the effect of changing 02 concentrations dall and Raven 1981). All of these experiments are from 21 to 1% or less; the stimulation of net photo- based on estimations of the intracellular free inorganic synthesis is much smaller than is the case for C3 C pool in photosynthesizing algal cells using a silicone terrestrial plants (Brown and Tregunna 1967; Bunt oil centrifugation technique, and they all show that 1971; Beardall and Morris 1975; Beardall et al. air-grown cells have a greater ratio of intracellular 1976; Bidwell 1977; Lloyd et al. 1977; Coleman to extracellular inorganic C than can be accounted and Colman 1980a; Shelp and Canvin 1980a, b). for by diffusive equilibration of CO, between medium Furthermore, the extent of CO, loss to CO2-free and cells, taking the (measured) pH of the two air in the light is smaller than would be expected compartments into account. The extent of accumu- from equation (3) and the operation of the PCOC, lation in Anabaena variabilis is much higher than provided techniques are used which do not lead to in the chlorophytes; this may reflect the absence of artifacts due to varying specific activities of the carbonic anhydrase in A. variabilis (Kaplan et al. substrate for CO., production in tracer experiments 1980a) because if HCO7 is the transported inorganic (Bidwell 1977; Findenegg and Fischer 1978; Cole- carbon species, a high intracellular concentration is man and Colman 1980a). Finally, the CO, com- required to give a sufficiently high rate of (uncata- pensation concentration is much lower than equation lyzed) conversion to CO., to explain the observed (3) would predict, especially if the occurrence of rate of CO 2 fixation. In the chlorophytes with their dark respiration in the light is taken into account. relatively high carbonic anhydrase activity the Equation (3) applied to the in vitro kinetic data of HCO—0O2 system is closer to equilibrium, and Badger (1980) gives an estimated CO, compensation less intracellular HCOI is needed in the steady state concentration (v g. 2v0 , with PCOC) of some to maintain the CO2 level. It thus appears that the 4 gliM for Anabaena variabilis in 21% 02 at 25°C, actively maintained inorganic carbon pool is an while the measured CO 2 compensation concentra- essential intermediate in "C 1 -like" photosynthesis tions in air-grown algae are generally 0.1-1.0 guM in the air-grown algae and can account for the in vivo at pH 7 (Egle and Schenk 1952; Brown and Tre- activities of RuBPc and RuBPo in terms of the in vivo gunna 1967; Bidwell 1977; Lloyd et al. 1977; CO., and 02 concentrations, and the in vitro RuBPc-o Findenegg and Fischer 1978; Birmingham and kinetics (despite uncertainties as to the barrier behind Colman 1979; Shelp and Canvin 1980a, b). which the accumulation occurs in eukaryote cells). These characteristics of air-grown algae re- The occurrence of this inorganic carbon pool to which semble those of C 1 higher plants (see Raven and "leaked" CO., is rapidly returned (see below) is Glidewell 1978), where substantial suppression of likely to complicate the interpretation of "pulse-

63 TABLE 2. Energetics of photosynthesis: Required input of reductant and ATP from "light reactions" per C fixed into useful compounds under various conditions of operation of the "CO2 concentration mechanism" and of pathways of glycolate metabolism. The operation of the "CO, concentrating mechanism" is assumed to require 3 ATP per net inorganic C transported; the kinetics of RuBP-co are assumed to be those described by Badger (1980) for Anabaena variabilis.

Operation of "CO2 Gas Pathway of concentrating phase glycolate metabolism mechanism" (H)/C fixed ATP/C fixed

(1) CO2 /02 sufficient Not produced No 4 3 to suppress RuBPo (2) Air Not metabolized Yes; [CO 2], 67 btilic 5.4 7.1 (3)" Air PCOC/glyc dh, Yes; [CO21; 67 1.1M 4.8 6.1 GS-GOGAT, NADH ox"' (4)" Air PCOC/glyc ox, Yes; [CO 2] 1 67 itM 4.8 6.4 GS-GOGAT, NADH ox"' (5)" Air PCOC/glyc dh, Yes; [CO 2], 67 /..tAl 4.7 6.3 GS-GOGAT, NADH exp"' (6)" Air Tartronic semi- Yes; [CO2], 67 I.LM 4.7 6.3 aldehyde/glyc dh (7) Air Tartronic semi- Yes; [CO 2], 67 le/ 4.7 6.5 aldehyde/glyc ox (8)" Air PCOC/glyc dh, No 12.5 4.5 GS-GOGAT, NADH ox" (9) Air PCOC/glyc ox, No 11.3 8.5 GS-GOGAT, NADH exp"' (10) Air Not metabolized No No net CO2 fixation

"The rates of oxidative phosphorylation required for condition (8) are too high to be supported by the respiratory capacity of most phototrophic cells (see Table 1). The other rates of oxidative phosphorylation (conditions (3), (4), (5), (6)) are within the capacity of most phototrophic cells (Table 1). "NADH ox means NADH produced in glycine to serine conversion oxidized with coupled ATP synthesis; NADH exp means this NADH is exported and used to support other reductive synthesis. e The [CO2 ] to give a ve /v,, ratio of 9. "GS-GOGAT = glutamine synthetase—glutamine oxoglutarate aminotransferase. chase" experiments designed to see if /3-carboxyla- (1980a) and Beardall and Raven (1981) suggest tion, and decarboxylation of C, dicarboxylic acids, that the "C00 concentrating mechanism" can be is an obligate reaction in net CO 2 fixation in photo- manifested regardless of whether CO2 or FICO synthesis by certain phytoplankters (see Raven is the species crossing the outer permeability barrier 1980a). (outer membrane of Cyanophyceae, plasmalemma of eukaryotes). Further, experiments with the lipid- Our inability to extract functional chloroplasts soluble cation tetraphenyl phosphonium (TPP + ) from the eukaryotic algae which exhibit the "C00 show that the inside-negative electrical potential concentrating mechanism" means that the inorganic difference across some membrane becomes more C specie- s which is actively transported, as well as negative when cells in which the "CO2 concentrating the membrane at which the transport occurs, is not mechanism" is de-repressed are supplied with the clear. Even in the Cyanophyceae the situation is inorganic C substrate for this transport mechanism; complicated by the occurrence of the typical gram- this has been shown for Chlorella einersonii (Bear- negative outer membrane which is not involved in dall and Raven 1981) and for Anabaelia variabilis active transport but which may influence experiments (A. Kaplan personal communication). The most designed to determine whether C 00 or HCO,i is economical explanation of these findings involves the species transported across the plasmalemma a primary, electrogenic influx of Hoeq at the plas- (Beardall and Raven 1981). Data reviewed by Raven malemma of the cyanophyte (A. Kaplan personal

64 communication) and at the (?) inner membrane Turning to capital costs, the CO2 -grown alga of the chloroplast envelope of the chlorophyte (Bear- has minimal ancillary catalysts for carbon assimi- dall and Raven 1981). lation — the carbonic anhydrase, PCOC and "CO, We may conclude that most of the apparent concentrating mechanism" activities are minimal C 1-like photosynthetic characteristics which are (see above, and Badger et al. 1977, 1978; Kaplan exhibited to a greater or lesser extent by air-grown et al. 1980a; Beardall and Raven 1981). Under low microalgae can be accounted for by the de-repression CO2 conditions these various accessory pathways of the "CO2 concentrating mechanism" with con- are de-repressed (thus increasing the capital costs sequent suppression of RuBPo activity (see Table 2). per unit C fixed) without any saving in terms of This has been elegantly demonstrated by Fock et al. lowered levels of PCRC enzymes and, in particular, (1981) for Chlanydomonas reinhardtii using ' 802 : the very energy — and N — expensive RuBPc-o air-grown cells show very little incorporation of (Raven 1977, 1980a; Reed and Graham 1977; Badger 18 0 into glycolate, glycine, or serine, and the sub- et al. 1977, 1978; Hogetsu and Miyachi 1979; Kaplan stantial light-stimulated ' 802 uptake is very largely et al. 1980a). This makes the hypothesis of Brown attributable to the Mehler reaction with a contribution (1978) for terrestrial C plants (i.e. that the energetic from dark respiration (see above). Cells grown in and, particularly, the N cost of producing the ancil- high CO2 show substantial 'HO labeling of glycolate lary C I machinery is more than outweighed by the when exposed to 18 02 -labeled air; these high CO., savings in the quantity of RuBPc and PCOC enzymes grown cells had not had time to de-repress their needed for unit C fixation) less readily applicable CO2 concentrating mechanism (H. P. Fock personal to microalgae with "C 1-like" physiology. How the communication). capital costs of the "CO., accumulating mechanism" To conclude our consideration of glycolate compare with those of the additional PCOC enzymes metabolism, the PCOC and RuBPc in relation to which would be required in the absence of this photorespiration in algae, we may consider the effect mechanism awaits further elucidation of the mech- of these pathways on running costs, capital costs, anism of "CO., accumulation." and safety in microalgae. Finally we turn to safety. RuBPo per se is not Dealing first with running costs, it is clear that a good thing for a plant to possess in a high 02 the energy input per unit reduced C produced is higher the lethality of a Chlamy- environment — witness when RuPBo is operative than when it is not (Table domonas mutation which has a higher than wild-type 2); this is related to the energy costs of glycolate ratio of RuBPo to RuBPc when attempts are made synthesis and of "scavenging" the glycolate back to to grow the organism at normal 02 levels in the light some more useful compound via the PCOC (where (Nelson and Surzycki 1976a, b), and of the (higher present) and, a fortiori, to the costs of glycolate plant) Arabidopsis mutants lacking phosphoglycolate synthesis when glycolate is excreted rather than phosphatase or serine-glyoxylate aminotransferase scavenged. These energy costs are fairly readily (Somerville and Ogren 1979, 1980) grown in 21% quantified (Table 2); what is less readily quantified 02 and with less than 1% CO2 . Justification for is the energy cost of the "CO 2 concentrating mecha- RuBPo activity in terms of the synthesis of glycine nism" which (by favoring RuBPc over RuBPo) can and serine in the PCRC is difficult in view of alter- spare much of the cost of glycolate synthesis and native pathways to these amino acids (Tolbert 1974). metabolism in air-grown algae. Raven (1980a) has The major "use" for RuBPo, and the associated computed an ATP requirement of 3/net CO 2 fixed PCRC, seems to be as a means of energy dissipation for active transport of inorganic C in a microalgal when the rate of light absorption exceeds the rate cell, allowing for leakage through a barrier with at which energy can be used in C fixation by the Pc,), = 10 -" cm • ; this is close to the P( (), deter- PCRC, i.e. at low CO2 /02 ratios. Despite the rela- mined by compartmental analysis in Dunaliella tively small fraction of light energy absorbed by a sauna (D. Zenvirt and A. Kaplan personal commu- cell exposed to full sunlight which can be processed nication). Experimental determinations of the quan- via the RuBPo-PCOC/RuBPc-PCRC pathways tum requirements for net CO, fixation in high — at the CO 2 compensation concentration, there is and low — CO.,-grown microalgae do not show very good empirical evidence of the efficacy of the path- large differences (Shelp and Canvin 1980a; A. Kaplan way in C, land plants (Powles 1979; Heber and personal communication). Further work is needed Krause 1980). The extent to which microalgae with to quantify the running costs of the "CO, concen- a functional "CO2 concentrating mechanism" can trating mechanism' vis a vis that of the RuBPo (with dissipate excess light energy at the CO2 compensation or without the PCOC) which would be incurred in concentration depends inter alia on the degree to the absence of the "CO., concentrating mechanism" which RuBPo is suppressed, and the "leakiness" under similar (air-equilibrium) CO., and 0 2 concen- of the inorganic C pump; Fock et al. (1981) find trations. that photoinhibition rapidly sets in when air-grown

65 Stored Carbohydrate Chlamydonionas is maintained under photosynthesis- the CO2 compensation CO 2 saturating irradiances at NADPH concentration. Sayre and Homann (1979) suggest Hexose P ATP*., that hydrogenase-mediated H., evolution may be a Ru5P significant alternative route for the dissipation of Triose in microalgae with an effi- NADH•r-.4..,ATP phenolics; excess excitation energy Light pent0Ses cient "CO., concentrating mechanism"; if such a PGA mechanism is present in the strain of Chlainydo- EMP pathway nionas used by Fock et al. (1981) it is not very PEP effective! In conclusion we concur with Andrews and Lorimer (1978) in regarding RuBPo as a neces-

Pyruvate sary concomitant of CO., fixation by RuBPc-o in an air-equilibrated solution, and view the various .\ 1C2°C2H] ATP for suppres- Aspartate lactlyt CoA --> Fatt Y acids, mechanisms which have been described and related Terpenoids \ sing RuBPo activity (C ,-like metabolism) or dealing amino- acids: pyrimidines with the glycolate produced (C 2-like metabolism) ATP< 2 mediate as representing the best selective compromise for a °2 given organism between the various considerations lumorate TCAC isocitrate 02 of minimizing running costs and capital costs per ATP< r's succinate ----)>ATP unit growth or per unit maintenance which is con- 2 -oxoglutarate °2ATP L-..\. sistent with the safe operation of the metabolic ma- succinyl glutamate chinery. CoA and related amino-acids

Dark Respiration

Porphyrins INTRODUCTION FIG. 4. The major "dark" respiratory processes in aerobic "Dark respiration" subsumes a heterogeneous phototrophs, together with their major products. collection of metabolic processes. The major com- NOTES: (1) The oxidative pentose phosphate pathway ponents are shown in Fig. 1 and 4. The "core" of provides NADPH for reductive biosynthesis (and, in some the glycolytic (EMP) obligate photolithotrophs with an incomplete TCAC, carbohydrate catabolism is NADPH for oxidative phosphorylation); its role in gen- pathway; running partly in parallel with the glycolytic erating C 4 and C5 biosynthetic C skeletons can occur via pathway is the oxidative pentose phosphate pathway, the regenerative part of the cycle working alone (i.e. while the TCAC (Krebs cycle) is in series with without the oxidative steps, or the reductive steps of the glycolysis. These various processes make three major PCRC) (Raven 1972a, b; Raven 1976a). (2) The glycolytic contributions to the economy of the cell, as follows pathway is probably the major pathway from sugar to (Davies et al. 1964; Beevers 1970; ap Rees 1974). pyruvate in phytoplankters other than the cyanobacteria, 1) By providing ATP for biochemical and biophy- where a low activity of phosphofructokinase may mean sical growth and maintenance processes. The ATP an increased role for the oxidative PPP. (3) The TCAC, substrate-level phosphorylations when complete (as is probably the case in most phyto- is generated in plankters, including all facultative heterotrophs and some in the glycolytic and TCAC pathways, and, in obligate photolithotrophs), has an important role in gener- much larger quantities in the aerobic cell, by ating reductant for oxidative phosphorylation as well as membrane-associated oxidative phosphorylation in generation of C skeletons for biosynthesis. When the using reductant generated by dehydrogenases cycle is incomplete (as in some obligate photolithotrophs) acting on organic substrates. the net generation of reductant is related stoichiometrically to the generation of biosynthetic C skeletons (reductant generated in right-hand limb minus reductant consumed (C + C3 ) as shown (Appleby et al. 1980) or by a glyoxylate in the left-hand limb). The relative operation of these limbs cycle; this latter could overcome the restriction on reductant in producing a cell of a given composition depends on the supply from organic acid metabolism imposed by an in- pathway by which porphyrins are synthesized: the "clas- complete TCAC (Whittenbury and Kelly 1977) as well sical" route involves the use of succinate and glycine to as permitting total cell synthesis from 2C compounds. generate 8-aminolaevulinic acid, while the "C;" route (5) Recent work suggests that the "classical" oxidative makes use of 2-oxoglutarate; the latter pathway predomi- PPP as portrayed in textbooks is wrong, and that a more nates in most 02 -evolvers, particularly for chlorophyll complex variant involving octuloses is used (Mujaji 1980). synthesis (Beale 1978; Troxler and Offner 1979). (4) (6) The CO, production related to C skeleton synthesis The biosynthetic use of the TCAC requires a net input of for growth is some 0.1-0.15 CO, per net C assimilated C acids, achieved either by anaplerotic CO ? fixation (Raven 1972a, b; I976a, b).

66 2) By providing reductant, mainly as NADPH from outside ( rest of the oxidative PPP, for reductive biosynthesis, cell )

e.g. the reduction of nitrate and nitrite. NADH • H 3) As the "core" of metabolism, yielding essential • C skeletons for biosynthesis. NAD' A El, Of these products, (1) and (2) are susceptible to being supplemented, or even replaced, by the direct use of the light (thylakoid) reactions of photo- synthesis (ATP, NADPH/reduced ferredoxin: Raven f Exogenous NADH dh 1971, 1972a, b, 1976a, b). This option is, of course, only open to photosynthetically competent cells in 2 H the light; nonphotosynthetic cells growing hetero- 2H . cyts trophically must obtain their ATP and reductant from the respiratory processes. Certain of the C skeletons mentioned under (3) are unique to "respi- ratory" processes, and cannot be generated in "photo- synthetic" reactions. This applies particularly to the lower portion of glycolysis and to the TCAC (except for such (nonalgal) phototrophs as the Chlo- robineae which may use a reversed, reductive TCAC in photosynthetic C fixation (Benedict 1978)).

LOCATION, STOICHIOMETRY, AND VARIABILITY OF THE PATHWAYS FIG. 5. The mitochondrial H+-transporting redox chain. The chain represented here has a proton-translocating The location within the photosynthetic euka- "loop" associated with the (endogenous) NADH to UQ ryotic cell of "dark" respiratory reactions has been step (Lawford and Garland 1972), a "proton-motive Q discussed by Raven (1976a). Summarizing and cycle" (Mitchell 1976), and a "conformational" (as opposed bringing this information up to date, the entire EMP to a redox-loop) H+ transport site associated with cyto- sequence is present in the cytosol; by comparison chrome oxidase (Wikstrom and Krab 1979; cf. Moyle and with higher plants the hexose phosphate-3PGA Mitchell 1978a). These mechanisms give a stoichiometry portion of the EMP pathway is probably also present of 8 H+ transported per 2e moving from NADH to 0 2 , in the plastids of green (chlorophyte and charophyte) as found by Brand et al. (1978) (cf. Moyle and Mitchell 1978b). Most of the evidence as to mechanisms and stoi- algae (Stitt and ap Rees 1979). This is probably chiometry has been obtained with mammalian and fungal related to the storage of starch in these plastids; mitochondria; the specifically "plant" features have been no other major taxon of algae, even the chlorophyll best investigated on higher plant and fungal mitochondria b containing Euglenophyceae, stores polysaccharide (e.g. the NADHdh for exogenous NADH, and the alternate in the plastids, and such plastids may well lack those oxidase; Palmer 1979). The "alternate oxidase" represents glycolytic enzymes not common to photosynthesis. a "physiological uncoupling" which has a very different The same may be true of the oxidative PPP; the mechanism from that found in vertebrates (Nicholls 1979), enzymes (dehydrogenases) specific to this pathway but which has a very similar end result and control mech- are absent from the plastids of Euglena, but are anism (Sharpless and Buetow 1970; Nicholls 1979; Van- derleyden et al. 1980a, b). (dh = dehydrogenase) probably present in the plastids of chlorophyte and charophyte algae as well as higher plants (Smillie 1963; Stitt and ap Rees 1979, 1980). The implications The Cyanophyceae have the enzymes of the EMP of the presence or absence of polysaccharide storage pathway (often with low activity of phosphofructo- in the plastids for the location of enzymes of hexose kinase), the oxidati've PPP, and the TCAC (but metabolism should, perhaps, be reconsidered in without 2-oxoglutarate dehydrogenase) in the cytosol relation to osmoregulation and the need for "compa- (Raven 1972a, b; Whittenbury and Kelly 1977). tible solutes" (often identical with the main soluble Oxidative phosphorylation is, of course, associated carbohydrate reserve) to occur within the plastids with membranes. Peschek (1980) has recently shown as well as in the cytosol if the relative volumes of that the cell membrane of Anacystis nidulans cells these compartments are to be maintained at different which had lost, by photobleaching, their intracellular internal osmolarities. The TCAC enzymes are located (thylakoid) membranes, still show respiratory activ- in the mitochondrial matrix, while the catalysts of ity, although with a lower specific activity (on a oxidative phosphorylation are associated with the membrane protein basis) than did the pigmented inner mitochondrial membrane (Lloyd 1974a, b; membrane (cell membrane plus intracellular mem- Lloyd and Turner 1980); see Fig. 5 and 6. brane) of control cells. It is likely that at least some 67 outside inner inside (rest of mitochondrial (matrix) cell) membrane

2 H . 2H * ATP&"

4-• AD_P3-

H2PO4- H 2 P°4- ATP for Endergonic * H processes

A,DP ,- ADP 3-

CA 4 " TP ATP 4-

FIG. 6. The mitochondria' ATP synthetase and ADP, ATP, and P, transport systems. The stoichiometry of the ATP synthetase is 2 H+ per internal ADP + Pi converted to ATP, while the exchange of internal ATP for external ADP + P, consumes another 1 H+ per ATP generated. The overall stoichiometry of H+ transported per exogenous (cytoplasmic) ADP + P1 converted to exogenous (cytoplasmic) ATP is thus 3:1 (Klingenberg 1979). With the 1-1 4-/e2 of 8 in the oxidation of NADH by 02 (Fig. 5), the P/e 2 ratio for NADH oxidation by 02 would be 2.67 rather than the classical value of 3 (Brand et al. 1978; Brand 1979).

of the oxidative phosphorylation activity is normally lysis are oxidized by the "external" NADH dehy- in the thylakoid (Peschek and Schtnetterer 1978; drogenase of the inner mitochondrial membrane). Peschek 1980). The ATP production will be lower if site 1 of oxi- The stoichiometry of generation of the products dative phosphorylation is bypassed in a trade-off of respiration is quite well understood for the oxi- of efficiency of conversion of sugar energy into dative PPP; 2 NADPH are produced per CO., evolved. ATP energy against the rate of ATP synthesis: For use in reductive biosynthesis each NADPH bypassing site 1 can increase the rate of ATP syn- generated by the oxidative PPP has a higher cost thesis per unit of mitochondrial machinery with in absorbed quanta than does an NADPH generated internal NADH as substrate, although it is not clear directly by the photoacts; the generation of the carbo- if this bypass occurs in algae (Lloyd 1974a, b; Ere- hydrate substrate for the oxidative PPP involves cinska et al. 1978; Lloyd and Turner 1980; cf. Odum the input of as many NADPH from photosynthesis and Pinkerton 1955; Wamcke and Slayman 1980). as are later regenerated upon carbohydrate oxidation, ATP production per unit carbohydrate oxidized together with the ATP required for the PCRC (plus will also be decreased if the "alternate oxidase" the "CO2 concentrating mechanism" and/or PCOC) is operative; this pathway is widespread in photo- and storing and mobilizing the carbohydrate product trophs and fungi, and in trypanosomid protozoa, (see Raven I976a). Thus the running costs for and bypasses all but site 1 of oxidative phospho- NADPH generation by the oxidative PPP exceed rylation (Henry and Nyns 1975; Lloyd and Turner those for the direct use of photosynthetic NADPH. 1980; Kirst 1980). This pathway may be useful if The stoichiometry of oxidative phosphorylation operation of the TCAC is required for C skeleton shown in Fig. 5 and 6 is still a matter of some biosynthesis but NADH oxidation in mitochondria debate; if it is accepted, then the ATP generated from is prevented by a suppression of oxidative phospho- the complete oxidation of one molecule of endoge- rylation by photophosphorylation (Marre 1961). nous glucose is some 33.33 ATP (assuming 2 ATP Thus the 33.33 ATP per glucose (corresponding are used in the hexokinase and phosphofructokinase to a P/e2 of 2.8 overall) is an upper limit for the reactions, and that the 2 NADH generated in glyco- efficiency of oxidative phosphorylation in euka- 68 ryotic algae. The relative efficiency of oxidative TABLE 3. Rates of membrane-associated 0 2 uptake phosphorylation and the various forms of photo- required to account for "photorespiratory" glycolate phosphorylation has already been mentioned in metabolism with a RuBPc/RuBPo of 9. relation to the Mehler reaction. In Cyanophyceae (prokaryotes) one might expect a higher P/e., ratio from the H+/e2 and 1-1±/P ratios shown in Fig. 5 02 uptake per 2 and 6, as the energy-requiring adenylate and glycolate metabolized 02 uptake phosphate transport across the inner mitochondrial (i.e. 2 02 metabolized per net C membrane does not occur. However, the (chloropl ast- Pathway by RuBPo) fixed type) coupling factor may have a higher Fl+ /ATP ratio than its mitochondria] equivalent, and the PCOC (glycolate 0.5 0.03 H+/e., ratio may be lower due to a simpler cyto- oxidase, chrome oxidase structure (Garland 1977; Fergusson GS-GOGAT) et al. 1979; Yamanaka and Fujii 1980). PCOC (glycolate 1.5 0.09 At all events in vivo estimates of the P/e2 dh, GS-GOGAT) give values of 2.63-3.08 ratio in cyanobacteria PCOC (glycolate 0 0 (Pelroy and Bassham 1973). oxidase, GDH) PCOC (glycolate 1 0.06 dh, GDH)" CAPACITY OF DARK RESPIRATION Tartronic semi- 0 0 discussion of the role of dark respiratory aldehyde, Any glycolate oxidase processes in the life of microalgae requires an estimate Tartronic semi- 1 0.06 of the capacity of the pathways, i.e. the maximum aldehyde, rates at which they can generate ATP, reductant, glycolate dh and C skeletons for comparison with rates of growth and the energy requirements of growth and main- "GDH = glutamic dehydrogenase. tenance. This is particularly important for photo- trophs, as it is difficult to measure the activity of dark respiratory processes in the light in green rates (Ried et al. 1962; Soeder et al. 1962; Ried cells, and an estimate of the capacity of the pathways et al. 1963; Peschek and Broda 1973; Pelroy et al. of respiration at least puts an upper limit on the 1976). The capacity of the oxidative PPP to generate extent to which they contribute to metabolism in the reductant for biosynthesis can be tested by the addition light — the supply of the ATP, NADPH, and C of NO in the dark to N-deprived cells (Syrett 1955). skeletons which they carry out in the dark, together Finally, the comparison of the ratio of inner mito- with any additional functions in the light (e.g. certain chondrial membrane area to thylakoid area, or of reactions of the PCOC which occur in the membranes ubiquinone (respiratory) to plastoquinone (photo- catalyzing oxidative phosphorylation — see above, synthetic) content of cells, can be used to estimate and Tables 1 and 3). the ratio of oxidative phosphorylation to photosyn- A number of approaches to the estimation of thetic phosphorylation capacity of phototrophic cells, the capacity of dark respiratory processes in algae as the rate of ATP synthesis per unit membrane have been employed. In vitro methods involve the area or per unit quinone is similar in the two energy- measurement of enzyme activities in extracts under coupling membranes (Shimikazi et al. 1978; Raven optimal conditions; the lowest activity reflects the 1980a, b, and unpublished calculations). maximum capacity of the metabolic sequence in Expressing the results of these calculations of vivo (ap Rees 1974). In vivo methods include respiratory capacity in the form (capacity for CO., measurements of the rate of CO 2 and 00 exchange production in dark respiration)/(maximum achieved in the dark under conditions designed to relieve as growth rate), with both quantities expressed in log, many constraints as possible on the rate of respiratory units 11— ' , the results obtained show considerable processes. These include the presence of an uncoupler variation between major taxa of algae. The data for of oxidative phosphorylation to maximize respira- respiratory capacities for the Bacillariophyceae, tory electron transport activity; the addition of some Chlorophyceae and Euglenophyceae compiled by exogenous organic carbon source whose uptake and Raven (1976a, b), and additional data on capacities metabolism requires ATP (a method which can clearly derived from the references in Raven (1976a, b), only work for those algae with substantial capacities Buetow (1968) , and Werner (1977) , together with for transport and metabolism of exogenous sub- those cited in the previous paragraph, suggest that strates), or the addition of exogenous osmotica which the maximum CO., produced/unit of C assimilated stimulate ATP-requiring ion transport and compatible during growth (bir /g„ , where p„.„,,, denotes dark solute metabolism, also maximize electron transport respiratory capacity and p,„ denotes growth) in these

69 three classes can be as low as 0.3-0.5 for phototro- ENERGY REQUIREMENTS FOR GROWTH phically grown cells, while for heterotrophically AND MAINTENANCE IN RELATION TO THE grown cells the ratio is 0.5-1.2. It is thus clear that ENERGY PROVIDED BY DARK RESPIRATION the respiratory capacity associated with a given maximum specific growth rate is lower for photo- The classic exposition of the relationship be- trophs than for otherwise comparable heterotrophs. tween growth rate and the rate of respiration for a This difference in respiratory capacities is exem- heterotroph is (equation (4)): plified by comparisons of the fraction of the (non- (4) Pg. C vacuolar) volume of the cell which is occupied by = mitochondria; this is significantly lower in photo- where 14 is the specific respiration rate (11' ), p,„ trophic than in heterotrophic microalgae, even if the is the specific growth rate (11 - '), p,„ is the specific results are expressed in terms of the volume of cyto- maintenance rate (11 - ' ), and c is the ratio of C lost plastids (whose volume can plasm excluding the as CO., in growth-associated respiration to the C vary considerably between cells grown in the different assimilated into cell material. The values of p„. and trophic regimes) (Raven 1980a, b; Pellegrini 1980). p,„ are obtained from plots of IL,. vs. ,u,„ ; the slope of the graph (ideally a straight line) gives c, while Data on p„..„,, //,/,, for the Cyanophyceae sug- the intercept at 14 = 0 gives p.,,. The value of c gest that the ratio can be substantially lower in should be related to the composition of the organism these organisms that in the three eukaryote classes in that the CO., production should equal the net discussed above. For Anacystis nichtlans the ratio requirement for reductant in converting the organic is probably less than 0.05 (Kratz and Myers 1955; and inorganic substrates into cell material (higher Peschek and Broda 1973; Doolittle and Singer 1974) if nitrate rather than ammonium is the N source, and in the wild type, and is even lower in revertants if acetate rather than glucose is the C source) plus to mutant strain 704 which lack both of the dehy- the CO., production which equals the 02 uptake drogenases of the oxidative PPP (Doolittle and involved in ATP generation in oxidative phospho- Singer 1974). Pelroy et al. (1976), in a study on the rylat ion in order to produce sufficient ATP for growth facultative heterotroph Aphanocapsa 6714 which (Raven 1971, 1972a, b, 1976a, b; Penning de Vries dealt with the phosphorylation capacity under various et al. 1974). Values of c which are higher than the light and dark conditions, showed that oxidative values predicted from cell composition and the phosphorylation had only 0.05 of the capacity of known mechanisms of transport and synthesis in the (noncyclic plus cyclic) photophosphorylation. organism may be attributed to some kind of "slip- By contrast, the Dinophyceae are a class of algae page" or "uncoupling" (Beevers 1970; see below). which seem to be characterized by high rates of In the case of phototrophic growth we have the dark respiration relative to the rate of growth or of problem that, in the light, some of the ATP and net photosynthesis (Moshkind 1961; Dunstan 1973; NADPH required for processes other than the con- Humphrey 1975; Prézelin and Sweeney 1978; Burns version of CO2 into carbohydrate may be produced and Owens 1978). The work of 1977; Falkowski by the light reactions of photosynthesis rather than Prézelin and Sweeney (1978) gives a specific growth by the "dark" respiratory processes and that the rate of 0.0086 h- ' at light saturation in a 12-h light:12- extent to which the "dark" respiratory processes h dark cycle, while the achieved specific respiration occur in the light is not easy to estimate (Ried 1970; rate is 0.0096 , in cultures of Gonyaulax polyedra. Raven 1971, 1972a, b; 1976a, b). In this case the The potential respiratory rate is probably higher than involvement of dark respiration in "growth" (con- the achieved rate (Hochachka and Teal 1964; cf. version of photosynthate into cells) can be estimated Thomas 1955). In phototrophic members of the Dino- by comparing the CO., /C ratio which is required phyceae we may conclude that the capacity for dark for the conversion of photosynthate into cells if all respiration is as high as it is in heterotrophic cells of of the ATP and NADPH were supplied by respiration many- other algae when expressed as a fraction of the with likely values of the CO2/C actually resulting growth rate. from dark respiration. The energy requirements and the energy input must take maintenance into account The computations of respiratory capacity as a (see below); two strategies may be followed in esti- fraction of growth rate that have been performed mating the respiratory contribution. One is to take here generally tend to underestimate the respiratory the measured dark respiration in the dark period (if capacity, as other processes may be limiting the any) of the daily cycle, and to add to it the respiration respiratory rate achieved in vivo; despite this, it is which would have occurred in the light period if the clear that there are variations of at least an order of "dark" respiration rate during the light period was magnitude in this ratio between different classes of identical with the rates measured in the first steady- algae. state CO., evolution found after cessation of long-term

70 illumination (more than an hour). The other strategy is respiration is more than adequate to supply the re- to take, as before, the respiration which was measured quirement for ATP and NADPH for growth, and, during the dark period of the delay cycle, and to add to if this respiration is not subject to great slippage it the respiration which would have occurred in the or uncoupling, the direct use of photoproduced light period if dark respiration had been proceeding at cofactors for growth need not be invoked. its maximum rate (the capacity discussed above). The conclusion that (Dinophyceae aside) the For Chlorella (see Raven 1976a, b) an appro- direct use of photoproduced cofactors is important priate value of c in equation (4) would be 0.46 for in supplying growth processes is reinforced if the growth on ammonium as N source and 0.76 for temporal differences between respiratory supply and growth on nitrate as N source. These figures both growth demand for energy is considered; this is parti- contain a CO2 /C of 0.36 for the generation of 2 ATP cularly the case if very little growth processes occur per C assimilated (P/e., of 2.8, respiratory quotient in the dark period of the diurnal cycle although = 1; see above), and a term for the CO., produced respiration exceeds the maintenance requirement in generating the reductant needed to bring (carbo- (Raven 1976a, b). hydrate plus ammonium plus sulphate) or (carbo- hydrate plus nitrate plus sulphate) to the redox level The other energy-requiring process to which of the final cell material. For ammonium as N source respiration may contribute is maintenance. Even this CO2 /C increment is 0.1, and for nitrate as N when no growth occurs, the maintenance of viability source, 0.4 (Raven 1976a, b). From the discussion and, more particularly, the ability to resume growth in Raven (1976a, b) it is clear that the lower estimate at a rapid rate as soon as the missing resource (e.g. of dark respiration mentioned above (but which is light or some nutrient) is restored, requires a con- still in accord with much experimental data) would tinued energy supply. This is used inter alia for give a "dark" respiratory input of ATP and NADPH resynthesizing unstable macromolecules, and for the considerably lower than that required according to active transport which recovers ions which have a COdC of 0.46 (ammonium) and 0.76 (nitrate). leaked through a membrane (Penning de Vries 1975). Even the higher estimate (assuming that "dark" Maintenance respiration may be computed by the use respiration proceeded at its maximum rate throughout of equation (4); this value for ,u„ may be similar the light period) gives insufficient respiratory ATP to the respiration rate measured after prolonged and NADPH to supply growth requirements for darkness. The alternative method of computing a either ammonium of nitrate as N source (bearing in maintenance requirement is to extrapolate the mind that the capacity for the oxidative PPP in vs. incident irradiance relationship to zero irradiance, eukaryotes is not readily available to supply reductant yielding a maintenance coefficient ge (van Liere and to oxidative phosphorylation (ap Rees 1974). Thus Mur 1979). In general, the measured rate of respi- for Chlorella and for other chlorophyceaeans, for ration in the dark period of a light—dark diurnal members of the Bacillariophyceae and Chryso- cycle is greater than the maintenance coefficients phyceae, and for some Euglena strains, it is likely or 11,1, computed by extrapolation (van Liere et that dark respiration cannot make all the ATP and al. 1979), particularly in the Cyanophyceae, where NADPH required for growth processes during normal varies from 0.001 to 0.004 11 -1 as compared with phototrophic growth (Laws and Caperon 1976; Raven 0.007-0.015 h -1 in the Chlorophyceae. In this respect 1976a, b; Laws and Wong 1978): some of the ATP it is of interest that Doolittle and Singer (1974) and NADPH for growth, as well as that used in the found that revertant strains of Anacystis nidulans conversion of CO2 to carbohydrate, must come mutant 704 had specific respiration rates in the dark from noncyclic, pseudocyclic, and cyclic photophos- of below 0.005 h -1 , yet had as good an ability to phorylation. remain viable in prolonged darkness as did the wild type whose specific respiration rate shortly after This conclusion applies a fortiori to cyano- cessation of illumination was about 0.05 h -1 . The phyceans even allowing for the (possible) greater role of fermentation in maintenance in the revertant flexibility in the use of reductant (NADPH) for strains with negligible activity of the two dehydro- either reductive biosynthesis or for ATP generation genases of the oxidative PPP remains to be explored in oxidative phosphorylation; the respiratory capacity (cf. Peschek and Broda 1973). of cyanobacteria gives a C 04C for respiration of less than 0.1 in fast-growing cells while the required The remainder of this discussion of dark respi- /C if respiration is to supply all of the ATP ration is taken up with possible explanations for and NADPH for growth processes is similar to that variations in specific respiration rates between or- for Chlorella , i.e. 0.4-0.8 depending on N source. ganisms and between the same organism growing For Dinophyceae, however, the CO, /C from dark under different conditions.

71 VARIATIONS IN SPECIFIC RESPIRATION NADPH required per unit C assimilated in growth, RATES: GENOTYPIC AND PHENOTYPIC there are substantial differences in the energy costs EFFECTS AND POSSIBLE EXPLANATIONS of synthesizing, from "photosynthate" and inorganic nutrients, unit C in unit C In the preceding section we saw that there were protein compared with in polysaccharide; the case if considerable differences in specific respiration rate this is particularly nitrate the N source between different phytoplankton organisms which rather than ammonium or urea is (Raven 1972a, b, 1976a, b; Penning de Vries et al. could not be entirely explained by differences in 1974; Raven and Glidewell 1975b). It has been 1.4 according to equation (4) (cf. Laws 1975; Banse suggested (Raven that this may 1976). Three general classes of explanation may be and Glidewell 1975b) be significant in decreased investigated; one appeals to variability in the extent shade adaptation as a to protein would to which the direct use of photoproduced ATP and fraction of the cell dry weight due reductant "subsidizes" the respiratory provision of reduce the "growth" energy requirement (from these cofactors; a second involves differences in cell respiration or more directly from photosynthetic composition which implies different energy re- partial reactions) per unit C assimilated without quirements per unit cell growth as changes in the decreasing gross photosynthesis at low irradiances, made in en- maintenance requirement; the third attributes differ- provided the protein economies were ences zymes (not limiting at low irradiances) rather than in respiratory rate to different degrees of complexes coupling between exergonic (respiratory) in the light-harvesting pigment-protein and ender- seems to be gonic (growth and maintenance) processes. (see Prézelin 1981). This argument applicable and Dealing first with the notion of changes in the to genotypically adapted " sun" "shade" benthic macroalgae (compare Table 1 of extent of the "energy subsidy" from photoproduced Raven Raven 1981), but cofactors which "spares" the requirement to generate et al. 1979, with Table 7 of its validity algae is less readily these cofactors by respiratory processes, the data for phytoplankton demonstrated. Here we are forced to rely on data discussed in the previous section showed that al- from phenotypic adaptation of a given genotype though there is an a priori case for such a subsidy in on the many classes of phytoplankters, the Dinophyceae to different irradiance regimes. The evidence the cell yield no such evidence (assuming efficient coupling activity of a major enzymic contributor to shows of respiratory and energy-requiring processes). This protein level, RuBPc-o (see Raven 1980b), argues for considerable genotypic variation between that its activity per unit cell protein decreases at low algal classes with respect to this subsidy, with Cyano- irradiances for growth in Phaeoclactylian tricornaturn phyceae having the largest subsidy and Dinophyceae (Beardall and Morris 1976) and Scenedesinas the smallest. The extent of phenotypic variation obliquas (Senger and Fleischenhacker 1 978) (cf. in the extent of the subsidy appears to be small, Molloy and Schmidt 1970), and that cell protein also as equation (4) is obeyed with an essentially constant declines at low irradiances (Beardall and Morris other value of c when is altered by changes in light or 1976; Parrot and Slater 1980). However, , in (chemical) nutrient supply; this is exemplified by the cases there is either no change in the fraction of irradiance measurements of Myers and Graham (1961) and Cook cell weight contributed by protein or N as (1961) on light-limited cultures of Chlorella and alters, or the N /C ratio increases at low irradiances Euglena respectively, and by the work of Laws (e.g. Yader 1979; Scott 1980; Tomas 1980). and Caperon (1976) and Laws and Wong (1978) on We have already noted that the thorough inves- nitrate-limited growth of Thalassiosira ,Monochrysis , tigations by Laws and Caperon (1976) and Laws and DunaHenn. In all of these cases, specific respira- and Wong (1978) showed that equation (4) held tion rate is a sufficiently low fraction of the specific with a constant c value for cultures of Thalassiosira, growth rate that there is a requirement for an "energy ltionochrysis , and Danaliella in which growth rate subsidy" for growth and maintenance processes from was varied by nitrate supply although, as the cell photosynthetic partial reactions, and in each case this analyses given in those papers show, the C:N ratio subsidy (as a fraction of the total energy required increases with decreasing (nitrate-limited) growth for processes other than carbohydrate generation in rates (cf. Goldman et al. 1979; Laws and Bannister photosynthesis) is relatively invariant with growth 1980). Further investigation is needed to clarify rate. We may note that these conclusions only hold the energy costs of synthesis under these culture if there are not significant differences in the cell conditions and its influence on the rate of respiration. composition as a function of limitation of growth rate The effects of variation in photoperiod on by lack of light or nutrients; as we shall see, there are specific respiration rate, specific growth rate, and cell changes in cell composition with growth rate. composition are not identical with the effects of varied Turning to the possibility that changes in c and irradiance at constant photoperiod (Foy et al. 1976; p„, (equation (4)) can be related to changes in cell Hobson et al. 1979; Humphrey 1979; Foy and Smith composition, and hence changes in the ATP and 1980). Ultimately these effects should be interpretable 72 in terms of optimal temporal strategies of biosynthesis (Sleigh 1974); if Gonyaulax has a total flagella length in varying photoperiods, as is discussed in this paper of 250 sm , then the maximal ATP consumption per and by Cohen and Parnas (1976), van Liere et al. cell in flagella activity is 2.5. x 10' 7 mol ATP • s'; (1979), and Foy and Smith (1980). with a P/e 2 of 2.8 and a C/cell volume of 0.125 Another compositional difference which might g • cm', the respiration associated with motility is alter the energy cost of growth and maintenance is only 0.000015 , which is only a small fraction of cell structure; this is an extension of the chemical maintenance respiration (at least 0.001 h" ). The total composition differences discussed above. It is pos- energy use by flagella (e.g. in nutrient uptake) is sible that the increased intracellular membrane limited by ATP supply along the axoneme (see Raven content of eukaryotes compared with prokaryotes 1980a). Thus flageller motility is not a major contri- can, in part, explain the larger maintenance costs in butor to the energy requirements of dinoflagellates. the Chlorophyceae compared with the Cyanophyceae. A cell characteristic which has been frequently The maintenance of the volume of the cytosol relative discussed with respect to rates of growth and of dark to that of "non-leaky" organelles (Raven 1980a) respiration in microalgae is cell size (e.g. Laws requires a constant input of energy to avoid "colloid- 1975; Banse 1976). The dependence of "pump and osmotic swelling" of the compartment with the higher leak" maintenance energy requirements on cell size content of macromolecules (Jakobssen 1980). How- (via the ratio of plasmalemma area to cell volume) ever, van Gemerden (1980) points out that similar has been emphasized, although the area of intra- differences in maintenance requirement can occur cellular membranes and the quantity of protein per within the prokaryotes (the Chlorobineae having unit volume is clearly also important in determining values similar to the Cyanophyceae, and the Rho- the maintenance requirement (see above, and Laws dospirillineae resembling the Chlorophytes), although 1975). Empirically, the analysis by Laws (1975) this again may be related to differences in intracellular suggests that there is a significant variation in the membrane arrangement (Raven and Beardall 1981). respiration/growth rate relationship with cell size, Cell structure may be significant in accounting while Banse (1976) in a subsequent review of the for the (usually) higher respiration rates of flagellate data found no significant variation in this ratio with compared with nonflagellate microalgae (Moshkind cell size. In view of the significance of cell size in 1961; Falkowski and Owens 1978). Raven (1976c, the ecology of phytoplankton (Guillard and Kilham 1980a) has pointed out that flagellate cells are effec- 1977), a mechanistic approach to the relationship tively wall-less with respect to osmoregulation even between cell size, specific growth rate, and specific when (e.g. Chlatnydotnonas) they possess a wall, respiration rate would be desirable. as the flagellar membrane is always wall-less. Vol- A further possible explanation of differences in ume regulation in wall-less cells requires a constant the specific respiration rate (and the ratio specific expenditure of energy, regardless of whether the cells respiration rate/specific growth rate) is related to are hypertonic in freshwater and osmoregulate by stress. We have implicitly taken an "unstressed" contractile vacuoles, or are isotonic in seawater organism as our paradigm; stress in general increases but still have to deal with colloid-osmotic swelling the rate of respiration relative to that of growth, if (Raven 1976c; Jakobssen 1980). Walled cells, by only because the decreased growth rate makes main- contrast, have a volume regulation (the cell wall) built tenance a greater fraction of the total respiration in during growth, and consequently may have a (as is the case for low irradiances or restricted nutrient lower maintenance energy requirement for volume availability which we have already considered). Aside and osmotic regulation as far as the whole cell is from this, stress (e.g. temperatures or osmolarities concerned, although turgor maintenance probably which are nonoptimal) can increase c and as well involves energy input in a "pump and leak" system as decrease ,u,„ . It is often difficult to decide if an (Raven 1976c). increased c or js,. results from a "respiratory control" It is not easy to quantify these osmoregulatory response to an increased demand for energy as part energy requirements; it is easier (because the mech- of adaptation or is increased "slippage" as a result anism is better understood) to evaluate the suggestion of damage to the respiratory or energy-consuming (Moshkind 1961; cf. Falkowski and Owens 1978) processes. The data are generally not easy to interpret that the high respiration rates of flagellates and, in (e.g. Guillard and Ryther 1962; Ryther and Guillard particular, of dinoflagellates is related to their mo- 1962). tility. A Gonyaulax polyedra cell 50 gm in diameter Finally, in explanation of variations in c and (Prézelin and Alberte 1978) has a specific respiration it,„ we can fall back on "slippage" (Beevers 1970; rate of up to 0.0096 h" (Prézelin and Sweeney Neijssel and Tempest 1976). This includes not just 1978). Flagella have some 600 ATPase (dynein) the wasteful reoxidation of NADPH by O., rather molecules per micrometre length and a specific than nitrate reductase, the excessive operation of the reaction rate of these ATPases of up to 100 s' alternate oxidase, or the operation of futile cycles

73 which act as ATPases; it can also extend to mis- efficient use is made of unit synthetic machinery matches between the operation of reductant-gener- if the ATP and NADPH are supplied continuously ating and reductant-consuming reactions, or of (as is possible with respiratory energy supply) rather energy-transducing and C-skeleton-generating than intermittently (as is demanded by the direct use reactions, either spatially or temporally, resulting in of photoproduced cofactors) (Raven 1976a, b; see the "wasteful" oxidation of some potentially useful above). Thus the relative advantages of direct versus intermediate. respiratory generation of cofactors are difficult to analyze in terms of capital costs. DARK RESPIRATION - CONCLUSIONS The final consideration is that of safety. Dark An analysis of dark respiration in microalgae respiration involves the reduction of 02, but the in terms of running costs, capital costs, .and safety problem of 02 radical generation is less severe than will serve to summarize our discussion. The yield with photosynthetic processes (e.g. the Mehler (ATP/quantum) of oxidative phosphorylation (cal- reaction); the major superoxide and peroxide gen- culated from the energy costs of carbohydrate syn- erator in respiration is the alternate oxidase (Rich thesis in photosynthesis, and the ATP yield from the et al. 1977). It is difficult to see how dark respiration complete oxidation of carbohydrate is not more than can serve to prevent photoinhibition by consuming 0.66. This value assumes not only complete coupling more excitation energy, as short-term feedback in photosynthetic and oxidative phosphorylation, effects of photosynthate concentration on the rate of but also that the cost of CO, fixation is 3 ATP and photosynthesis are poorly documented; thus addi- 2 NADPH per CO., (see above); this underestimates tional consumption of photosynthate by dark respi- the cost in normal air when the RuBPo-PCOC and/ ration would not necessarily increase the rate of or CO2 concentrating mechanism increase the energy energy consumption by CO 2 fixation (cf. Fogg 1975). required per CO2 fixed. The maximum yield of In any case, the extra respiration would presumably pseudocyclic and cyclic photophosphorylation is be via the alternate oxidase which would itself (like 0.33 and 0.67, respectively, so oxidative phospho- the Mehler reaction) generate toxic radicals. rylation has a similar efficiency to cyclic photo- In wider safety and survival terms, the pre- phosphorylation and a better efficiency than pseudo- sence of a substantial respiratory capacity may be cyclic photophosphorylation. The direct use of photo- useful in pennitting the acquisition of scarce nutrients produced reductant is more efficient than the use (e.g. phosphate, fixed nitrogen) over the whole of this reductant to generate carbohydrate with 24-h cycle. This would be particularly advantageous subsequent regeneration of NADPH in the oxidative in migratory organisms which can spend the day PPP, as the direct use of NADPH yields, in addi- in a location optimal for light absorption and the tion, 1.33 ATP per NADPH in noncyclic photo- night in regions optimal for nutrient acquisition; this phosphorylation, while regeneration of NADPH behavior has been documented for Dinophyceae from carbohydrate via the oxidative PPP involves (Eppley et al. 1968), which have (as has been dis- the input of at least another 0.17 ATP per NADPH cussed earlier) high respiration rates (cf. Raven generated to give the minimal ATP required for CO., 1976a, b, 1980a). Similar considerations apply to fixation. Fermentation as a source of ATP is some the energy supply required to adapt to stresses which 20-fold less efficient than is the oxidative phospho- can occur in the dark as well as in the light: a good rylation in terms of running costs. Thus the direct example is changes in osmolarity, where the rate of use of photoproduced reductant for growth processes adaptation is siruilar in light and in darkness (Kirst is always favored over reductant generation by the 1975; Kirst and Keller 1976; Brown and Hellebust oxidative PPP in terrils of running costs, while the 1978). direct use of ATP generated in photophosphorylation An efficient respiratory supply of ATP for main- has no such clear advantage. tenance is important for dark survival of phytoplank- The analysis of respiratory versus direct supply ton organisms which are likely to be subjected to of ATP and NADPH for growth processes as a prolonged darkness by dropping out of the euphotic function of the capital invested requires cognizance zone (e.g. Antia and Cheng 1970; Smayda and Mit- to be taken of both the production and consumption chell-Innes 1974; Antia 1976). Survival is improved costs of the cofactors. Clearly the capital costs of by exposure to subcompensation irradiances which the direct use of cofactors are less than those involved may work via photosynthesis or sonie partial reactions in the use of photoproduced cofactors to fix CO., thereof serving to "spare" respiratory substrates. with subsequent regeneration of the cofactors in There seems to be no direct information relating respiration; two further tiers of catalytic machinery respiratory capacity and efficiency to dark survival are involved (the stromal CO., fixation system and of microalgae; it would be expected that efficient the respiratory catalysts). However, a consideration respiration working on large reserves vvould lead of the energy-consuming reactions shows that more to prolonged survival, as would a cell composition 74 which needed less maintenance energy. In the dark as a phototroph. This is not to say that the processes in the absence of oxygen, the relatively inefficient cannot be modulated or made to serve other desirable process of fermentation is presumably used as a ends; the Mehler reaction can function (via pseudo- source of maintenance energy. Moss (1977) has cyclic photophosphorylation) to generate ATP, demonstrated a positive correlation between the while both the Mehler reaction and RuBPo may be ability to survive dark anaerobiosis and likelihood involved in the dissipation of excess, and possibly of the organism encountering dark anaerobiosis in harmful, excitation energy in the photosynthetic nature for a number of (nonplanktonic) microalgae. apparatus. This "energy-dissipating" role of RuBPo Kessler (1973) has suggested that the possession of activity under low CO 2-high 0 2 conditions is perhaps hydrogenase, with a corresponding ability to produce not as significant in most microalgae as in C3 higher less toxic fermentation products, can favor dark- plants, as many microalgae have a "CO., con- anaerobic survival in microalgae (cf. the possible role centrating mechanism" which substantially sup- of hydrogenase in energy dissipation under conditions presses RuBPo activity under natural conditions. of high light input-low redox acceptor availability Either the operation or the suppression of RuBPo as discussed above). is a costly procedure, with respect to running costs and to the production of the catalytic machinery, and this must be taken into account in considering Conclusions the growth energetics and kinetics of algae which lack overt photorespiratory CO, production as well The respiratory processes of microalgae have as those which possess it. been considered here in terms of their contributions These physiological considerations can be to the efficiency of growth (from the point of view applied to ecology. The competitive dominance of of running costs and capital investment) and to the Oscillatoria agardhii in cultures or in shallow lakes safety and survival aspects of the organism's exist- at low irradiances (below 10 W • m -2) and of Scene- ence. "Dark" respiration has clear survival value desmus protuberans at higher irradiances is clearly from the viewpoint of the generation of essential C related to the lower specific maintenance coefficient skeletons, and in generating ATP and NADPH for of Oscillatoria and to the higher light-saturated spe- growth and maintenance processes in the dark. Effi- cific growth rate and resistance to photoinhibition ciency and capital cost considerations may be in- of Scenedesmus (Mur and Beijdorf 1978). However, volved in the evolutionary determination of how we do not know the mechanisms behind the low much of the conversion of immediate (mainly carbo- maintenance requirement in the cyanophytes, or th hydrate) products of photosynthesis into "cell ma- relation of the Mehler reaction and RuBPo-PCOC terial" occurs in the light when the direct use of activity to the higher irradiances required for photo- photoproduced cofactors is possible as a supplement inhibition in the chlorophyte. to the use of respiratory ATP and NADPH for these "growth" processes. The wide variation in the dark Acknowledgments respiratory rate as a fraction of growth rate between We are grateful to the Science Research Council for different algal classes suggests that the selective supporting our work (and that of Dr S. M. Glidewell) by balance between respiratory generation of cofactors Grants B/RG/1403, B/RG191966, and GR/A/69896. and the direct use of cofactors from photosynthesis We dedicate this paper to Professor Jack Myers, Zoologist is struck at different points and for unknown reasons Extraordinary. in different algae. The wide variations in dark res- piratory rate between algae, and environmental influences on these rates, are not amenable to detailed explanation at the moment; clearly a process which References consumes between 5 and 50% of gross photosynthate must be better understood before any comprehensive ANDREWS, T. J., AND G. H. LORIMER. 1978. Photo- mechanistic analysis of the energetics and kinetics respiration — still unavoidable? FEBS Lett. 90: 1-9. of phytoplankton growth is possible. ANTIA, M. J. 1976. Effects of temperature on the darkness The photorespiratory processes (Mehler reac- survival of marine microplanktonic algae Microb. tion, and RuBPo and its appendages) are best re- Ecol. 3: 41-54. garded as unavoidable consequences of the generation ANTIA, N. J., AND J. Y. CHENG. 1970. The survival of axenic of photosynthetic reductant, of the operation of cultures of marine planktonic algae from pro- longed exposure to darkness at 20°C. Phycologia 9: RuBPc-o, in "natural" CO., and 02 concentrations; 179-183. their survival value is primarily, on this view, that APPLEBY, G., J. COLBECK, E. S. HOLDSWORTH, AND the alga has to carry out the processes of which H. WADMAN. 1980. The /3-carboxylation enzymes they are necessary concomitants if it is to survive in marine phytoplankton and isolation and purification

75 of pyruvate carboxylase from Amphidinium carterae concentration. Cargenie Inst. Wash. Yearn. 75: 423- (Dinophyceae). J. Phycol. 16: 290-295. 432. AP REES, T. 1974. Pathways of carbohydrate breakdown BIDWELL, R. G. S. 1977. Photosynthesis and light and in higher plants, p. 89-128. In D. H. Northcote dark respiration in freshwater algae. Can. J. Bot. [ed.] M.T.P. International reviews of science, bio- 55: 809-818. chemistry series one. Vol. II. MTP Press. BIRMINGHAM, B. C., AND B. COLMAN. 1979. Measure- ASADA, K., AND Y. NAKANO. 1978. Affinity for oxygen ment of carbon dioxide compensation points of fresh- in photoreduction of molecular oxygen and scavenging water algae. Plant Physiol. 64: 892-895. of H2 02 in spinach chloroplasts. Photochem. Photo- BRAND, M. D. 1979. Stoichiometry of charge and proton biol. 28: 917-920. translocation in mitochondria: steady-state measure- BADGER, M. R. 1980. Kinetic properties of RuBPc-o from ment of charge/0 and P/0 ratios. Biochem. Soc. Anabaena variabilis. Arch. Biochem. Biophys. 207: Trans. 7: 874-880. 247-254. BRAND, M. D., W. G. HARPER, D. G. NICHOLLS, AND BADGER, M. R., A. KAPLAN, AND J. A. BERRY. 1977. W. J. INGELDEW. 1978. Unequal charge separation The internal CO 2 pool of Chlatnydonzonas reinhardtii: by different coupling spans of the mitochondria] response to external CO 2 . Carnegie Inst. Wash. Yearb. electron transport chain. FEBS Lett. 95: 125-129. 76: 362-366. BRANDEN, R. 1978. Ribulose bisphosphate carboxylase 1978. A mechanism for concentrating CO 2 in and oxygenase from green plants are two different Chlatnydomonas reinhardtii and Anabaena variabilis enzymes. Biochem. Biophys. Res. Commun. 81: and its role in photosynthetic CO2 fixation. Carnegie 539-546. Inst. Wash. Yearb. 77: 251-261. BROWN, A. H., AND D. WEIS. 1959. Relation between 1980. The internal inorganic carbon pool of respiration and photosynthesis in the green alga Chlatnydotnonas reinhardtii: evidence for a CO2 Ankistrodesnuts Invite. Plant Physiol. 34: 224-234. concentrating mechanism. Plant Physiol. 66: 407-413. BROWN, D. L., AND E. B. TREGUNNA. 1967. Inhibition BADOUR, S. S., AND E. R. WAYGOOD. 1971a. Glyoxylate of respiration during photosynthesis by some algae. carboligase activity in the unicellular green alga Can. J. Bot, 45: 1135-1143. Gloeotnonas sp. Biochim. Biophys. Acta 242: 493- BROWN, L. M., AND J. A. HELLEBUST. 1978. Ionic 499. dependence of deplasmolysis in the euryhaline diatom 1971b. Excretion of an acid semialdehyde by Cyclotella ctrica. Can. J. Bot. 56: 408-412. Gloeotnonas. Phytochemistry 10: 967-976. BROWN, R. H. 1978. A difference in N use efficiency 1972. Evidence for the operation of the glycerate in C3 and C I plants and its implications for adaptation pathway in Chlatnydomonas segnis. Plant Physiol. and evolution. Crop Sci. 18: 93-98. 50: 572. BUETOW, D. E. 1968. The biology of Euglena. Vol.I and BANSE, K. 1976. Rates of growth, respiration and photo- II. Academic Press, New York, NY. synthesis of unicellular algae as related to cell size - BULLOCK, K. W., T. R. DEASON, AND J. C. O'KELLEY. a review. J. Phycol. 12: 135-140. 1979. Occurrence of glycolate dehydrogenase and BEALE, S. I. 1978. Aminolaevulinic acid in plants: its glycolate oxidase in some coccoid, zoospore-producing biosynthesis, regulation and role in plastid develop- green algae. J. Phycol. 15: 142-146. ment. Annu. Rev. Plant Physiol. 29: 95-120. BUNT, J. S. 1965. Measurements of photosynthesis and BEARDALL, J., AND I. MORRIS. 1975. Effects Of environ- respiration in a marine diatom with the mass spec- mental factors on photosynthesis patterns in Phaeo- trometer and with C14. Nature (London) 207: 1373- dactylum tricornutum. II. Effect of oxygen. J. Phycol. 1375. 11: 430-434. 1971. Levels of dissolved oxygen and carbon 1976. The concept of light intensity adaptation fixation by marine microalgae. Limnol. Oceanogr. in marine phytoplankton: some experiments with 16: 564-566. Phaeodactylum tricornututn. Mar. Biol. 37: 377-388. BuNT, J. S., O. van H. OWENS, AND G. HOCFI. 1966. BEARDALL, J., D. MUKERJI, H. E. GLOVER, AND I. Exploratory measurements on the physiology of a MORRIS. 1976. The path of carbon in photosynthesis psychrophilic marine diatom. J. Phycol. 2: 96-100. by marine phytoplankton. J. Phycol. 12: 409-417. BuRRIS, J. E. 1977. Photosynthesis, photorespiration and BEARDALL, J., AND J. A. RAVEN. 1981. Transport of dark respiration in eight species of algae. Mar. Biol. inorganic carbon and the CO2 concentrating mechanism 39: 371-379. in alorelia etnersonni (Chlorophyceae). J. Phycol. CHRISTEN, P., AND A. GASSER. 1980. Production of 17: 134-141. glycolate by oxidation of the 1,2 dihydroxyethyl- BEEVERS, H. 1970. Respiration in plants and its regulation, thiamin-diphosphate intermediate of transketolase with p. 209-214. In Measurement of photosynthetic pro- hexacyanoferrate (III) or H2 02. Eur. J. Biochem. ductivity. PUDOC, Wageningen. 107: 73-77. BENED1c-r, C. R. 1978. Nature of obligate photo-auto- CODD, G. A., AND M. MERRETT. 1971a. Photosynthetic trophy. Annu. Rev. Plant Physiol. 29: 67-93. products of division synchronized cultures of Euglena. BERRY, J. A., AND G. BOWES. 1973. Oxygen uptalce Plant Physiol. 47: 635-639. in vitro by RuDP carboxylase of Chlamydomonas 197 lb. The regulation of glycolate metabolism reinhardtii. Carnegie Inst. Wash. Yearb. 72: 405-407. in division synchronized cultures of Euglena. Plant BERRY, J. A., J. BOYNTON, A. KAPLAN, AND M. Physiol. 47: 640-643. BADGER. 1976. Growth and photosynthesis of CODD, G. A., AND G. H. SCHMID. 1972. Serological Chlatnydomonas reinhanitii as a function of CO2 characterization of the glycolate-oxidizing enzymes

76 from tobacco, Euglena gracilis and a yellow mutant FALKOWSKI, P. G., AND T. G. OWENS. 1978. Effects of Chlorella vulgaris. Plant Physiol. 50: 769-773. of light intensity on photosynthesis and dark respi- CODD, G. A., AND A. K. J. SELLAL. 1978. Glycolate ration in six species of marine phytoplankton. Mar. oxidation by thylakoids of the Cyanobacteria Ana- Biol. 45: 289-295. baetza cyclindrica, Nostoc muscorum and Chlorogloea FARQUHAR, G. D., S. VON CAEMMERER, AND J. A. fritschii . Planta 139: 177-182. BERRY. 1980. A biochemical model of photosynthetic CODD, G. A., AND W. D. P. STEWART. 1973. Pathways CO, fixation in leaves of C3 plants. Planta 149: 78-90. of glycolate metabolism in the blue-alga Anabaena FERGUSSON, S. J., M. C. SORGATO, D. B. KELL, AND cyclindrica. Arch. Mikrobiol. 94: 11-28. P. JOHN. 1979. Comparative aspects of the energetics 1974. Glycolate oxidation and utilization by of oxidative phosphorylation in bacteria and mito- Anabaena cylindrica. Plant Sci. Lett. 3: 199-205. chondria. Biochem. Soc. Trans. 7: 870-873. COHEN, D., AND H. PARNAS. 1976. An optimal policy FINDENEG, G. R., AND E. FISHER. 1978. Apparent for the metabolism of storage materials in unicellular photorespiration of Scenedesmus obliquus: decrease algae. J. Theor. Biol. 56: 1-18. during adaptation to low CO, level. Z. Pflanzen- COLEMAN, J. R., AND B. COLMAN. 1980a. Effect of physiol. 89: 363-371. oxygen and temperature on photosynthetic carbon FLOYD, G. L., AND J. L. SALISBURY. 1977. Glycolate assimilation in two microscopic algae. Plant Physiol. dehydrogenase in primitive green algae. Am. J. Bot. 65: 980-983. 64: 1294-1296. 1980b. Demonstration of C3 -photosynthesis in FOCK, H. P., M. R. BADGER, D. T. CAN VIN, AND S. C. a blue-green alga, Coccochloris peniocystis. Planta WONG, 1981. Mass spectrometric determination of 149: 318-320. oxygen evolution and uptake in Chlamydomonas reinhardtii. (In preparation) COLLINS, N., AND M. J. MERRETT. 1975. The localization FOGG, G. E. 1975. Biochemical pathways in unicellular of glycolate-pathway enzymes in Euglena. Biochem. J. 148: 321-328. plants, p. 437-457. In J. P. Cooper [ed.] Photo- synthesis and production in different environments. COOK, J. R. 1961. Euglena gracilis in synchronous divi- Cambridge University Press, Cambridge. sion. II. Biosynthetic rates over the life cycle. Biol. Foy, R. H., C. E. GIBSON, AND R. V. SMITH. 1976. Bull. 121: 277-289. The influence of daylength, light intensity and tem- COUGHLAN, S. 1977. Glycolate metabolism in Thalas- perature on the growth rates of planktonic blue-green siosira pseudonana. J. Exp. Bot. 28: 78-83, algae. Br. Phycol. J. 11: 151-163. DAVIES, D. D., J. GIOVANELLI, AND T. AP REES. FOY, R. H., AND R. V. SMITH. 1980. The role of carbo- 1964. Plant biochemistry. Blackwell Scientific Publi- hydrate accumulation in the growth of planktonic cations, Oxford, Oscillatoria species. Br. Phycol. J. 15: 139-150. DIMON, B., AND R. GERSTER, 1976. Incorporation FOYER, C. H., AND D. O. HALL. 1980. Oxygen meta- d'oxygène dans le glycolate excrété à la lumière par bolism in the active chloroplast. Trends Biochem. Euglena gracilis . C. R. Acad. Sci. Paris Ser. D 283: Sci. 5: 188-191. 507-510. FREDERICK, S. E., P. GRUBER, AND N. E. TOLBERT. DOOLITTLE, W. F., AND R. A. Singer. 1974. Mutational 1973. The occurrence of glycolate dehydrogenase and analysis of dark endogenous metabolism of the blue- glycolate oxidase in green plants. Plant Physiol. 52: green bacterium Anacystis nidulans. J. Bacteriol. 318-323. 119: 677-683. FRENCH, C. S., AND D. C. FORK. 1960. Two primary DROOP, M. R. 1974. Heterotrophy of carbon, p. 530- photochemical reactions in photosynthesis driven by 559. In W. D. P. Stewart [ed.] Algal physiology different pigments. Carnegie Inst. Wash. Yearb. 60: and biochemistry. Blackwell Scientific Publications, 351-357. Oxford. GARLAND, P. B. 1977. Energy transduction and trans- DUNSTAN, W. H. 1973. A comparison of the photosyn- mission in microbial systems. Symp. Soc. Gen. thesis-light intensity relationship in phylogenetically Microbiol. 27: 1-21. different marine microalgae. J. Exp. Mar. Biol. Ecol. GERSTER, R., AND P. TOURNIER. 1977. Metabolic 13: 181-187. pathway of oxygen during photorespiration: incorpo- EGLE, K., AND W. SCHENK. 1952. Untersuchungen ration of IHO into glycolate, glycine and serine, p. über die Reassimilation der Atmungkohlensaure bei 129-130. In Abstracts of the Fourth International der Photosynthese der Pflanzen. Beitr. Biol. Pflanz. Congress on Photosynthesis, Reading, September, 29: 75-105. 1977, U.K. ISES, 21 Albemarle St., London. ELSTNER, E. F., AND J. KONZE. 1976. Wege der Sauer- GIMMLER, H. 1977. Photophosphorylation in vivo, p. stoffaktivierung in verschedien Kompartimenten von 448-472. In A. Trebst and M. Avron [ed.] Ency- Pflanzenzellen. Ber. Dtsch. Bot. Ges. 89: 335-348. clopedia of plant physiology, new series. Vol. 5. Part I: EPPLEY, R. W., O. HOLM - HANSEN, AND J. D. H. Photosynthetic electron transport and photophos- STRICKLAND. 1968. Some observations on the phorylation. Springer-Verlag, Berlin. vertical migration of dinoflagellates. J. Phycol. 4: GLIDEWELL, S. M., AND J. A. RAVEN. 1975. Measure- 333-340. ments of simultaneous oxygen uptake and evolution ERECINSKA, M., D. F. WULSON, AND K. NISHIKI. in Hydrodictyon africanwn . J. Exp. Bot. 26: 479- 1978. Homeostatic regulation of cellular energy 488. metabolism: experimental characterization in vivo 1976. Photorespiration: RuDP oxygenase or and fit to a model. Am. J. Physiol. 234: C82-C89. hydrogen peroxide? J. Exp. Bot. 27: 200-204.

77 GOLDMAN, J. C., J. J. MCCARTHY, AND D. PEAVEY. 1979. Role of carbonic anhydrase in photosyn- 1979. Growth rate influence on the chemical compo- thetic CO2 fixation in Chlorella. Plant Cell Physiol. sition of phytoplankton in oceanic waters. Nature 20: 747-756. (London) 279: 210-215. HUMPHREY, G. F. 1975. The photosynthesis: respiration GOVINDJEE, O. van H. OWENS, AND G. E. HOCH. 1963. ratio of some unicellular marine algae. J. Exp. Mar. A mass spectroscopic study of the Emerson enhance- Biol. Ecol. 18: 111-119. ment effect. Biochim. Biophys. Acta 75: 281-284. 1979. Photosynthetic characteristics of algae GRAHAM, D., AND M. L. REED. 1971. Carbonic anhy- grown under constant illumination and light-dark drase and the regulation of photosynthesis. Nature regimes. J. Exp. Mar. Biol. Ecol. 40: 63-70. (London) 231: 81-83. JACKSON, W. A., AND R. J. VOLK. 1970. Photorespira- GRODZINSKI, B. 1979. A study of formate production tion. Annu. Rev. Plant Physiol. 21: 385-432. and oxidation in leaf peroxisomes during photo- JAKOBSSON, E. 1980. Interaction of cell volume, membrane respiration. Plant Physiol. 63: 289-293. potential and membrane transport parameters. Am. GRODZINSKI, B., AND B. COLMAN. 1975. The effect of J. Physiol. 238: C196-C206. osmotic stress on the oxidation of glycolate by the KAPLAN, A., M. R. BADGER, AND J. A. BERRY. 1980a. blue-green alga Anacystis nidulans. Planta 124: 125- Photosynthesis and the intracellular inorganic carbon 134. pool in the blue-green alga Anabaena varinbilis: GRUBER, P. J., S. E. FREDERICK, AND N. E. TOLBERT. response to external CO 2 concentration. Planta 149: 1974. Enzymes related to lactate metabolism in green 219-226. algae and lower land plants. Plant Physiol. 53: 167- KAPLAN, A., U. SCHREIBER, AND M. AvRoN 1980b. 170. Salt-induced metabolic changes in Dunalielln sauna. GUILLARD, R. R. L., AND P. KILHAM. 1977. The ecology Plant Physiol. 65: 810-813. of marine planktonic diatoms, p. 372-469. In D. KESSLER, E. 1973. Effect of anaerobiosis on photosyn- Werner [ed.] The biology of diatoms. Blackwell thetic reactions and nitrogen metabolism of algae Scientific Publications, Oxford. with and without hydrogenase. Arch. Mikrobiol. GUILLARD, R. R. L., AND J. H. RYTHER. 1962. Studies 93: 91-100. on marine planktonic diatoms. I. Cyclotelln nana KEYS, A. J., I. F. BIRD, M. J. CORNELIUS, P. J. LEA, Hustedt and Detomda confervacea (Cleve) Gran. AND R. M. WALLSGROVE. 1978. Photorespiratory Can. J. Microbiol. 8: 229-239. nitrogen cycle. Nature (London) 275: 741-743. HALLIWELL, B. 1974. Superoxide dismutase, catalase KIRST, G. O. 1975. Beziehungen unterschiedlicher and glutathione peroxidase: solutions to the problem Konzentrationen von NaCI und anderen osmotisch of living with oxygen. New Phytol. 73: 1075-1086. wirksatnan substanzen auf die CO 2 -Fixierung der 1978. The chloroplast at work. A review of einzelligen alge Platymonns subcordifonnis. Oecol- modern developments in our understanding of chloro- ogia 20: 237-254. plast metabolism. Prog. Biophys. Mol. Biol. 33: KIRST, G. O., AND H. J. KELLER. 1976. Der Einfluss 1-54. unterscheidlicher NaCI-Konzentrationen auf die HARTMANN, T., AND A. EHMKE. 1980. Role of mito- Atmung der enizelligen Alge Platymonas subcordi- chondrial glutamic dehydrogenase in the reassimilation formis Hazen. Bot. Mar. 19: 241-244. of ammonia produced by the glycine-serine transfor- 1980. Phosphate transport in Platymonas subcor- mation. Planta 149: 207-208. (tennis after osmotic shock. Z. Pflanzenphysiol. HEBER, U., AND G. H. KRAUSE. 1980. What is the 97: 289-297. physiological role of photorespiration? Trends Bio- KLINGENBERG, M. 1979. The ADP, ATP shuttle of the chem. Sci. 5: 32-34. mitochondrion. Trends Biochem. Sci. 4: 249-252. HEBER, U., AND D. A. WALKER. 1979. The chloroplast KRATz, A. W., AND J. MYERS. 1955. Photosynthesis envelope - barrier or bridge? Trends Biochem. Sci. and respiration of blue-green algae. Plant Physiol. 4: 252-256. 30: 257-280. HENRY, H. F., AND E. J. NYNS. 1975. Cyanide-insensitive KREBS, H. A., AND R. L. VEECH. 1969. Equilibrium respiration. An alternative mitochondrial pathway. relations between pyridine nucleotides and adenine Sub-Cell. Biochem. 4: 1-66. nucleotides and their roles in the regulation of metabolic HOBSON, L. A., F. A. HARTLEY, AND D. E. KETCHUM. processes. Adv. Enzyme Regul. 7: 397-413. 1979. Effects of variation in daylength and temperature LAING, W. A., W. L. OGREN, AND R. H. HAGEMANN. on net rates of photosynthesis, dark respiration and 1974. Regulation of soybean net photosynthetic CO2 excretion by Isochrysis galbana Parke. Plant Physiol. fixation by the interaction of CO2, 0 2 and ribulose 63: 947-951. 1,5 diphosphate carboxylase. Plant Physiol. 54: 678- HOCH, G. E., O. van H. OWENS, AND B. KOK. 1963. 685. Photosynthesis and respiration. Arch. Biochem. LAWFORD, H. G., AND P. B. GARLAND. 1972. Proton Biophys. 101: 171-180. translocation coupled to quinone reduction by reduced HOCHACHKA, P. W., AND J. M. TEAL. 1964. Respiratory NAD in rat liver and ox heart mitochondria. Biochem. metabolism in a marine dinoflagellate. Biol. Bull. J. 130: 1029-1044. 126: 274-281. LAWS, E. A. 1975. The importance of respiration losses HOGETSU, D., AND S. MIYACHI. 1977. Some effects in controlling the size distribution of marine phyto- of CO2 concentration during growth on subsequent plankton. Ecology 56: 419-426. photosynthetic CO 2 fixation in Chlorella. Plant Cell LAWS, E. A., AND T. T. BANNISTER. 1980. Nutrient Physiol. 18: 347-352. and light-limiting growth of Thalassiosira fluviatilis

78 in continuous culture, with implications for phyto- carboxylase from Euglena gracilis. Eur. J. Biochem. plankton growth in the ocean. Limnol. Oceanogr. 54: 195-206. 25: 457-473. MARRE, E. 1961. Phosphorylation in higher plants. Annu. LAWS, E. A., AND J. CAPERON. 1976. Carbon and nitrogen Rev. Plant Physiol. 12: 195-219. metabolism by Monochlysis 'tithed: measurement of MEHLER, A. H., AND A. H. BROWN. 1952. Studies on growth-rate-dependent respiration rates. Mar. Biol. reactions of illuminated chloroplasts. II. Simultaneous 36: 85-97. photoproduction and consumption of oxygen: studies LAWS, E. A., AND D. C. L. WONG. 1978. Studies of with oxygen isotopes. Arch. Biochem. Biophys. 38: carbon and nitrogen metabolism by three marine 365-371. phytoplankton species in nitrate-limited continuous MERRETT, M. J AND J. M. LORD. 1973. Glycolate cultures. J. Phycol. 14: 406-416. formation and metabolism by algae. New Phytol. LEX, M., J. B. SILVESTER, AND W. D. P. STEWART. 72: 751-757. 1972. Photorespiration and nitrogenase activity in the MITCHELL, P. 1976. Possible molecular mechanisms of blue-green alga, Anabaena cyclindrica. Proc. R. Soc. the protonmotive function of cytochrome systems. London Ser. B 180: 87-102. J. Theor. Biol. 62: 327-367. MOLLOY, G. R., AND R. R. SCHMIDT. 1970. Studies LLOYD, D. 1974a. The mitochondria of micro-organisms. on the regulation of ribulose 1,5 diphosphate Academic Press, London. car- boxylase synthesis during the cell cycle of the eukaryote 1974b. Dark respiration, p. 505-529. In W. D. P. Chlorella. Biochem. Biophys. Res. Commun. 40: Stewart [cd.] Algal physiology and biochemistry. 1125-1133. Blackwell Scientific Publications, Oxford. MOORE, A. L., C. JACKSON, B. HALLIWELL, J. E. LLOYD, D., AND G. TuRNER. 1980. Structure, function, DENCH, AND D. O. HALL. 1977. Intramitochondrial biogenesis and genetics of mitochondria. Symp. Soc. localization of glycine decarboxylase in spinach leaves. Gen. Microbiol. 30: 143-150. Biochem. Biophys. Res. Communs. 78: 483-491. LLOYD, N. D. H., D. T. CAN VIN, AND CULVER. D. A. MOSHK1ND, L. V. 1961. Photosynthesis by dinoflagellates 1977. Photosynthesis and photorespiration in algae. of the Black Sea. Sov. Plant Physiol. 8: 129-132. Plant Physiol. 59: 936-940. Moss, B. 1977. Adaptations of epipelic and epipsammic LORD, J. M., AND R. H. BROWN. 1975. Purification and freshwater algae. Oecologia 28: 103-108. some properties of Chlorella fusca ribulose bisphos- MOYLE, J., AND P. MITCHELL. 1978a. Cytochrome c phate carboxylase. Plant Physiol. 55: 310-314. oxidase is not a proton pump. FEBS Lett. 88: 268-272. LORD, J. M., AND M. J. MERRETT. 1970. The regulation 1978b. Measurements of mitochondrial H/0 of glycolic oxidoreductase with the photosynthetic quotients: effects of phosphate and N-ethylmaleimide. capacity. Biochem. J. 119: 125-127. FEBS Lett. 90: 361-365. 1971. The intracellular localization of glycolate MUJAJI, B. W. 1980. The pentose phosphate pathway oxidoreductase in Euglena gmcilis. Biochem. J. 124: revised. Biochem. Educ. 8: 76-78. 175-181. MUR, L. R., AND R. O. BEIJDORF. 1978. A model for LORIMER, G. M., AND ANDREWS. T. J. 1973. Plant the succession from green to blue-green algae based photorespiration - an inevitable consequence of on light limitation. Verh. Int. Verein. Limnol. 20: the existence of atmospheric oxygen. Nature (London) 2314-2321. 243: 359-360. MYERS, J., AND J. R. GRAHAM. 1961. On the mass LORIMER, G. M., T. J. ANDREWS, AND N. E. TOLBERT. culture of algae. III. Light diffusors: high versus low 1973. Ribulose diphosphate oxygenase. II. Further temperature Chlorellas. Plant Physiol. 36: 342-346. proof of reaction products and mechanism of action. NEIJSSEL, O. H., AND D. W. TEMPEST. 1976. Bio- Biochemistry 12: 18-23. energetic aspects of aerobic growth of Klebsiella LoRimEa, G. M., M. R. BADGER, AND T. J. ANDREWS. aerogenes NCTC in carbon-limited and carbon- 1977. Ribulose bisphosphate carboxylase-oxygenase. sufficient chemostat cultures. Arch. Microbiol. 107: Improved methods for the activation and assay of 215-222. catalytic activities. Anal. Biochem. 78: 66-75. NEILSON, A. H., AND R. R. LEWIN. 1974. The uptake LORIMER, G. M., K. C. Woo, J. A. BERRY, AND C. B. and utilization of organic carbon by algae: an essay OSMOND. 1978a. The C2 photosynthetic carbon in comparative biochemistry. Phycologia 13: 227- oxidation cycle in leaves of higher plants: pathway 264. and consequences, p. 311-322. In J. Coombs, D. O. NELSON, E. B., AND S. J. SURZYCKI. I976a. A mutant Hall, and T. W. Goodwin [ed.] Photosynthesis 77. strain of Chlamydomonasreinharchii exhibiting altered The Biochemical Society, London. ribulose diphosphate carboxylase. Eur. J. Biochem. LORIMER, G. M., C. B. OSMOND, T. AKAZAWA, AND 61: 465-474. S. ASAMI. 1978b. On the mechanism of glycolate 1976b. Characterization of oxygenase activity in synthesis by Chromatium and Chlorella. Arch. a mutant of Chlamydotnonas reinhardtii exhibiting Biochem. Biophys. 185: 49-56. altered ribulose diphosphate carboxylase. Eur. J. MCCURRY, S. D., N. P. HALL, J. PIERcE, C. PAECH, Biochem. 61: 465-474. AND N. E. TOLBERT. 1978. Ribulose bisphosphate NELSON, E. B., AND N. E. TOLBERT. 1969. The regulation carboxylase-oxygenase from parsley. Biochem. of glycolate metabolism in Chlamydomonas reinhard- Biophys. Res. Commun. 84: 895-900. tii. Biochim. Biophys. Acta 184: 263-270. MCFADDEN, B. A., J. M. LORD, A. ROWE, AND S. NELSON, E. B., A. CENEDELLA, AND N. E. TOLBERT. DILKS. 1975. Composition, quarternary structure and 1969. Carbonic anhydrase levels in Chlatnyclomonas. catalytic properties of D-ribulose-I,5-bisphosphate Phytochemistry 8: 2305-2306.

79 NICHOLLS, D. G. 1979. Brown adipose tissue mitochon- PENNING DE VRIES, F. W. T., A. H. M. BRUNSTING, dria. Biochim. Biophys. Acta 549: 1-29. AND M. M. VAN LAAR. 1974. Products, require- ODUM, H. T., AND R. C. PINKERTON. 1955. Time's ments and efficiency of biosynthesis: a quantitative speed regulator: the optimum efficiency for maximum approach. J. Theor. Biol. 45: 339-377. power output in physical and biological systems. Am. PESCHEK, G. A. 1980. Electron transport reactions in Sci. 43: 331-343. respiratory particles of hydrogenase-induced Anacystis OGAWARA, T., T. FUJII, AND S. AIBA. 1980. Effect indu/ans. Arch. Microbiol. 125: 123-133. of oxygen on the growth (yield) of Chlorella vulgaris PESCHEK, G. A., AND E. BRODA. 1973. Utilization of Arch. Microbiol. 127: 25-31. fructose by a unicellular blue-green alga, Anacystis OKABE, K. I., G. A. CODD, AND W. D. P. STEWART. nidulans Naturwissenschaften 60: 479-480. 1979. Hydroxylamine stimulates carboxylase activity PESCHEK, G. A., AND G. SCHMETTERER. 1978. Re- and inhibits oxygenase activity of cyanobacterial versible photooxidative loss of pigments and of intra- ribulose bisphosphate carboxylase-oxygenase. Nature cellular membranes in the blue-green alga Anacystis (London) 279: 525-527. niclulans FEMS Microbiol. Lett.3: 295-297. OLSEN, L. F., A. TELFER, AND J. BARBER. 1980. A PIRT, S. J., Y. K. LEE, A. RICHMOND, AND M. W. flash spectroscopic study of the kinetics of the electro- PI RT. 1980. The photosynthetic efficiency of Chlore/la chromic shift, proton release and the redox behaviour biomass growth with reference to solar energy utili- of cytochromes f and b-563 during cyclic electron flow. zation. J. Chem. Tech. Bioeng. 30: 25-34. FEBS Lett. 118: 11-17. POWLES, S. B. 1979. The role of carbon assimilation and OWENS, O. van H., AND G. E. HOCH. 1963. Enhance- photorespiratory carbon cycling in the avoidance of ment and de-enhancement effects in Anacystis nidu- photoinhibition in intact leaves of C3 and C.1 plants. fans . Biochim. Biophys. Acta 75: 183-186. Ph.D. thesis, Australian National University, Canberra. PALMER, J. M. 1979. The 'uniqueness' of plant mito- PRÉZELIN, B. B. 1981. Light reactions in photosynthesis. chondria. Trans. Biochem. Soc. 7: 246-252. Can. Bull. Fish. Aquat. Sci. 210: 1-43. PARROTT, L. M., AND J. H. SLATER. 1980. The DNA, ALBERTE. 1978. Photo- RNA and protein composition of the cyanobacterium PRÉZELIN, B. B., AND R. S. synthetic characteristics and organization of chloro- Anacystis nidulans grown in light and carbon dioxide- phyll in marine dinoflagellates. Proc. Natl. Acad , Sci. limited chemostats. Arch. Microbiol. 127: 53-58. Wash , 75: 1801-1804. PATTERSON, C. O. P., AND J. MYERS. 1973. Photosyn- PRÉZELIN, B. B., AND B. M. SWEENEY. 1978. Photo- thetic production of H2 02 by Anacystis nidulans Plant Physiol. 51: 104-109. adaptation of photosynthesis in Gonyaulax polyedra PAUL, J. S., C. W. SULLIVAN, AND B. E. VOLCAN!. Mar. Biol. 48: 37-35. 1975. Photorespiration in diatoms. H. Mitochondrial PRITCHARD, G. G., C. P. WHITTINGHAM, AND W. J. ttcolate dehydrogenase in Cylindrotheca fusifonnis GRIFFIN. 1961. Effect of isonicotinyl hydrazide on Nature (London) and Nitzschia alba. Arch. Biochem. Biophys. 169: the path of carbon in photosynthesis 153-159. 191: 553-554. PAUL, J. S., AND B. E. VOLCANI. 1974. Photorespiration 1963. The effect of isonicotinyl hydrazide on 2 in diatoms. I. The oxidation of glycolic acid in Thalas- the photosynthetic incorporation of radioactive CO siosira pseudonana (Cyclotella nana). Arch Microbiol. into ethanol-soluble compounds in Chlorella . J. Exp. 101: 115-120. Bot. 14: 281-289. 1975. Photorespiration in diatoms. III. Glycolate: RADMER, R. J., AND B. KOK. 1976. Photoreduction of cytochrome c reductase in the diatom Cylindrotheca 02 primes and replaces CO, fixation. Plant Physiol. fusiformis . Plant Sci. Lett. 5: 281-285. 58: 336-340. 1976a. A mitochondrial glycolate: cytochrome RADMER, R. J., B. KOK, AND O. OLLINGER. 1978. c reductase in Chlamydonzonas reinhardtii . Planta Kinetics and apparent Ku, of 02 cycle under conditions 129: 59-61. of limiting carbon dioxide fixation. Plant Physiol. 61: 1976b. Photorespiration in diatoms. IV. Two 915-917. pathways of glycolate metabolism in synchronized RADMER, R. J., AND O. OLLINGER. 1980. Light-driven cultures of Cylindrotheca fusiformis . Arch. Microbiol. uptalce of oxygen, carbon dioxide and bicarbonate 110: 247-252. by the green alga Scenedesnius . Plant Physiol. 65: PELLEGRINI, M. 1980. Three-dimensional reconstruction 723-729. of organelles in Euglena gracilis Z. I. Qualitative RAVEN, J. A. 1971. Energy metabolism in green cells. and quantitative changes of chloroplasts and mito- Trans. Bot. Soc. Edinb. 41: 219-225. chondrial reticulum in synchronous photoautotrophic 1972a. Endogenous inorganic carbon sources in culture. J. Cell Sci. 43: 137-166. plant photosynthesis. I. Occurrence of dark respiratory PELROY, R. A., AND J. A. BASSHAM. 1973. Efficiency pathways in illuminated green cells. New Phytol. of energy conversion by aerobic glucose metabolism 71: 227-247. in Aphatzocapsa 6714. J. Bacteriol. 115: 937-942. 1972b. Endogenous inorganic carbon sources in PELuov, R. A., M. R. KIRK, AND J. A. BASSHAM. plant photosynthesis. II. Comparison of total CO2 1976. Photosystem two regulation of macromolecule light with measured CO2 evolution production in the synthesis in the blue-green alga Aphanocapsa 6714. in the light. New Phytol. 71: 995-1014. J. Bacteriol. 128: 623-632. 1976a. Division of labour between chloroplasts PENNING DE VRIES, F. W. T. 1975. The cost of main- and cytoplasm, p. 403-443. In J. Barber [ed.] The tenance processes in plant cells. Ann. Bot. 39: 77-92. intact chloroplast. Elsevier, Amsterdam.

80 1976b. The quantitative role of 'dark' respiratory RIED, A., C. J. SOEDER, AND I. MULLER. 1963. Uber processes in heterotrophic and photolithotrophic plant die Atmung synchron kultivierter Chlorella. I. Veran- growth. Ann. Bot. 40: 587-602. deringen des respiratorisches Gaswechsels im Laufe 1976e. Transport in algal cells, p. 129-188. des Entwicklungscyclus. Arch. Mikrobiol. 45: 343- In U. Lüttge and M. G. Pitman [ed.] Encyclopedia of 358. plant physiology. New Series. Vol. 2A. Transport in ROSA, L. 1979. The ATP/2e ratio during photosynthesis Cells and Tissues. Springer-Verlag, Berlin. in intact Spinacia chloroplasts. Biochem. Biophys. 1977. Ribulose bisphosphate carboxlyase activity Res. Commun. 88: 154-162. in terrestrial plants: significance of 02 and CO2 diffu- RYTHER, J. H., AND R. R. L. GUILLARD. 1962. Studies sion. Curr. Adv. Plant Sel. 9: 579-590. of marine planktonic diatoms. III. Some effects of 1980a. Nutrient transport in microalgae. Adv. temperature on respiration of five species. Can. J. Microb. Physiol. 21: 47-226. Microbiol. 8: 447-453. 1980b. Chloroplasts of eukaryotic micro-or- SARGENT, D. F., AND C. P. S. TAYLOR. 1972. Terminal ganisms. Symp. Soc. Gen. Microbiol. 30: 181-205. oxidases of Chlore/la pyrenoidosa. Plant Physiol. 49: 1981. Nutritional strategies of submerged benthic 775-778. plants: The acquisition of C, N and P by rhizophytes 1975. On the respiratory enhancement in Chlore!- and haptophytes. New Phytol. 86: 1-30. la pyrenoidosa by blue light. Planta 127: 171-175. RAVEN, J. A., AND J. BEARDALL. 1981. The intrinsic SAYRE, P. T., AND P. H. HOMANN. 1979. Correlation permeability of biological membranes to H': signifi- between photorespiratory activity and hydrogenase cance for the efficiency of low rates of energy trans- content in Chlorella strains. Plant Physiol. 63: 153s. formation. FEMS Microbiol. Lett. 10: 1-7. SCOTT, J. M. 1980. Effect of growth rate of the food RAVEN, J. A., AND S. M. GLIDEWELL. 1975a. Sources alga on the growth/ingestion efficiency of a marine of ATP for active phosphate transport in Hydrodictyon herbivore. J. Mar. Biol. Assoc. U.K. 60: 681-702. africanum: evidence for pseudocyclic photophos- SENGER, H., AND P. FLEISCHENHACKER. 1978. Adap- phorylation in vivo. New Phytol. 75: 197-204. tation of Scenedesmus to light intensities. Physiol. 1975b. Photosynthesis, respiration and growth Plant 43: 35-42. in the shade alga Hydrodictyon africanum. Photo- SHARPLESS, T. K., AND R. A. BUETOW. 1970. An synthetica 9: 361-371. inducible alternate oxidase in Euglena gracilis mito- 1978. C 1 characteristics of photosynthesis in the chondria. J. Biol. Chem. 245: 58-70. C3 alga Hydrodictyon africanum. Plant Cell Environ. SHELP, B. J., AND D. T. CAN VIN. 1980a. Utilization of 1: 185-197. exogenous inorganic carbon species in photosynthesis 1981. Processes limiting photosynthetic conduct- by Chlorella pyrenoidosa. Plant Physiol. 65: 774-779. ance, p. 109-136. In C. B. Johnson [ed.] Physiological 19806. Photorespiration and oxygen inhibition processes limiting crop yield. Butterworths, London. of photosynthesis in Chlore/la pyrenoidosa. Plant RAVEN, J. A., F. A. SMITH, AND S. M. GLIDEWELL. Physiol. 65: 780-784. 1979. Photosynthetic capacities and biological strate- SHIMAZAKI, K., K. TAKAMIYA, AND M. NISHIMURA. gies of giant-celled and small-celled macro-algae. 1978. Studies on electron transfer systems in the New Phytol. 83: 299-309. marine diatom Phaeodactylum tricornutum. II. Identi- REED, M. L., AND D. G. GRAHAM. 1977. Carbon dioxide fication and determination of quinones, cytochromes and the regulation of photosynthesis: activities of and flavoproteins. J. Biochem. Tokyo 83: 1639- photosynthetic enzy and carbonate dehydratase 1642. (carbonic anhydrase) Chlorella after growth or SLEIGH, M. A. [ed.] 1974. Cilia and flagella. Academic adaptation in different carbon dioxide concentrations. Press, London. Austr. J. Plant Physiol. 4: 87-98. SMAYDA, T. J., AND B. MITCHELL - INNES. 1974. Dark RICH, P. R., A. L. MOORE, W. J. INGLEDEW, AND survival of autotrophie, planktonic marine diatoms. W. D. BONNER JR. 1977. EPR studies of higher plant Mar. Biol. 25: 195-202. mitochondria. I. Ubisemiquinone and its relation to SMILLIE, R. M. 1963. Formation and function of soluble alternative respiratory oxidations. Biochim. Biophys. proteins in chloroplasts. Can. J. Bot. 41: 123-154. Acta 462: 501-514. SOEDER, C. J., I. MULLER, AND A. RIED. 1962. Uber RIED, A. 1970. Energetic aspects of the interaction between der Einfluss von Salzkonzentration und Nitratangebout photosynthesis and respiration, p. 231-246. In Meas- auf der RQ und das N-reduzierten-System von voll- urement of photosynthetic productivity, PUDOC, synchroner Chlore/la pyrenoidosa. Vortr. Gesamtgeb. Wageningen. Bot. ptsch. Ges. I: 195-200. RIED, A., I. MULLER, AND C. J. SOEDER. 1962. Wirkung SOMERVILLE, C. R., AND W. L. OGREN. 1979. A phos- von Stoffweehselinhibitoren auf den respiratorischen phoglycolate phosphatase-deficient mutant of Arabi- Gaswechsel verschiedener Entwicklungsstadien von dopsis . Nature (London) 280: 833-836. Chlorella pyrenoidosa. Vortr. Gesamtgeb. Bot. Dtsch. 1980. Photorespiration mutants of Arabidopsis Ges. 1: 187-194. thaliana deficient in serine-glyoxylate aminotrans- RIED, A., AND B. REINHARDT. 1980. Distribution of ferase activity. Proc. Natl. Acad. Sci. Wash. 77: excitation energy between photosystem two and 2684-2687. photosystem one in red algae. III. Quantum require- SPALDING, M. H., AND W. L. OGREN. 1980. Properties ments of the induction of a State 2-State 1 transition. of the " CO2 pump" in Chlatnydomonas . Plant Physiol. Biochim. Biophys. Acta 592: 76-86. 65: 69a.

81 STABENAU, H. 1974a. Vereilung von Microbody-Enzymen native oxidases of Neurospora crassa by nucleoside aus Chlatnydomonas in Dichtegradienten. Planta 118: phosphates. Biochem. J. 188: 141-144. 35-42. VAN GEMERDEN, H. 1980. Survival of Chromatium I974b. Localization of enzymes of glycolate vinosum at low light intensities Arch. Microbiol. 125: metabolism in the alga Chlorobium elongation . Plant 115-121. Physiol. 54: 921-924. VAN LIERE, L., AND L. R. MUR. 1979. Growth kinetics 1975. Zur localization von Glycolsaure-oxidase of Oscillatoria agardhii Gomont in continuous culture, in Spirogyra. Ber. Dtsch. Bot. Ges. 88: 469-472. limited in its growth by light energy supply. J. Gen. 1977. Einfluss von CO2 -Konzentration auf Microbiol. 115: 153-160. Wachstum und Stoffwechsel von Chlorogonium VAN LIERE, L., L. R. MUR, C. E. GIVSON, AND M. elongatum. Biochem. Physiol. Pflanz. 171: 449-454. HERDMAN. 1979. Growth and physiology of Oscil- 1980. Enzymes of glycolate metabolism in the alga latoria agardhii Gomont cultivated in continuous Mougeottia. Plant Physiol. 65: 15a. culture and with a light-dark cycle. Arch. Microbiol. STEWART, R., AND G. A. CODD. 1980. Glycolate and 123: 315-318. glyoxylate excretion by the green alga Sphaerocystis VENNESLAND, B., AND K. JETSCHMANN. 1976. The shroeteri. Abstracts Intemational Phycological Union effect of cyanide and sonie other carbonyl binding Symposium, Glasgow, August 1980, P. 48. reagents on glycolate excretion by Chlorella vulgaris. STEWART, K. D., AND K. R. MATTOX. 1975. Com- Planta 128: 81-84. parative cytology, evolution and classification of the WARNCKE, J., AND C. L. SLAYMAN. 1980. Metabolic green algae with some consideration of the origin of modulation of stoichiometry in a proton pump. Bio- other organisms with chlorophylls a and b. Bot. Rev. chem. Biophys. Acta 591: 224-233. 41: 104-135. WEIS, D., AND A. H. BROWN. 1959. Kinetic relation- STITT, M., AND T. AP REES. 1979. Capacities of pea ships between photosynthesis and respiration in the chloroplasts to catalyse the oxidative pentose phosphate algal flagellate Ochromonas malhamensis. Plant pathway and glycolysis. Phytochemistry 18: 1905- Physiol. 34: 235-239. 1912. WERNER, D. [cd.] 1977. The biology of diatoms. Black- 1980. Estimation of the activity of the oxidative wells Scientific Publications, Oxford. pentose phosphate pathway in pea chloroplasts. Phy- WHITTENBURY, R., AND D. P. KELLY. 1977. Auto- tochemistry 19: 1583-1585. trophy: a conceptual phoenix. Symp. Soc. Gen. SYRETT, P. J. 1955. The assimilation of ammonia and Microbiol. 27: 121-149. nitrate by nitrogen-starved cells of Chlorella vulgaris. WHITTINGHAM, C. P. 1952. Rate of photosynthesis and I. The assimilation of small quantities of nitrogen. concentration of carbon dioxide in Chlore/la. Nature Physiol. Plant 8: 924-929. (London) 170: 1017-1018. 1981. Nitrogen metabolism of microalgae. Can. WIKSTROM, M., AND K. KRAB. 1979. Mechanism and Bull. Fish. Aquat. Sel. 210: 182-210. stoichiometry of redox-linked proton translocation THOMAS, W. H. 1955. Heterotrophic nutrition and respi- in mitochondria. Biochem. Soc. Trans. 7: 880-886. ration in Gonyoulax polyedra . J. Protozool. 2: 2s-3s. WI5K1CH, J. T. 1977. Mitochondrial metabolite transport. TOLBERT, N. E. 1974. Photorespiration, p. 474-504. Annu. Rev. Plant Physiol. 28: 45-69. In W. D. P. Stewart [ed.] Algal physiology and Woo, K. C., AND C. B. OSMOND. 1976. Glycine decar- biochemistry. Blackwell Scientific Publications, boxylation in mitochondria isolated from spinach Oxford. leaves. Aust. J. Plant Physiol. 3: 771-785. TOMAS, C. R. 1980. Olisthodiscus Wens (Chrysophyceae). YADER, J. A. 1979. Effect of temperature on light-limited IV. Effects of light intensity and temperature on photo- growth and chemical composition of Skeletonema synthesis and cellular composition. J. Phycol. 16: costatum (Bacillariophyceae). J. Phycol. 15: 362- 1-15. 370. YAMANAKA, T., AND K. FUJII. 1980. Cytochrome a-type TROXLER, R. F., AND C. D. OFFNER. 1979. 8-Amino- laevulinic acid synthesis in a Cyanidium caldarium terminal oxidase derived from Thiobacillus novellus. mutant unable to make chlorophyll a and phycobili- Molecular and enzymatic properties. Biochem. proteins. Arch. Biochem. Biophys. 195: 53-65. Biophys. Acta 591: 53-62. ZENVIRT, D., A. KAPLAN, AND L. REINHOLD. 1980. VANDERLEYDEN, J., E. VAN DEN FYNDE, AND H. Accumulation of inorganic C in Dunaliella sauna. VERACHTENT. 1980a. Nature of the effect of adeno- Plant Physiol. 65: 70a. sine-5'-monophosphate on the cyanide-insensitive ZIEM - HANCK, U., AND U. HEBER. 1980. Oxygen re- respiration in mitochondria of Moniliella tomentosa. 2 assimilation. Bio- Biochem. J. 186: 309-316. quirement of photosynthetic CO chim. Biophys. Acta 591: 266-274. VANDERLEYDEN, J., C. PEETERS, H. VERACHTENT, AND H. BERTRAND. 1980b. Stimulation of the alter-

82 Photosynthetic Products, Physiological State, and Phytoplankton Growth

IAN MORRIS1 Marine Program, University of New Hampshire, Durham, NH 03824, USA

Introduction 3) Little consideration is given to the assimilation The ultimate end-product of photosynthesis by of elements other than carbon. A proper under- phytoplankton is the synthesis of new cell material standing of the factors that control the chemical that permits the component cells to increase in mass composition of phytoplankton and the ultimate end- and leads, eventually, to an increase in the number products of synthesis depends on understanding the of cells in the population. The nature of the newly assimilation of elements such as N, P, Si, S, etc., synthesized cell material will depend on the balance as well as C. Because this paper is part of a session between the various metabolic reactions occurring on "photosynthesis," and consideration of other ele- in the cell. The nature of such a balance might be ments is the concern of other papers, emphasis is expected to be regulated by the environmental con- placed on C assimilation. ditions prevailing at any given time and the physio- 4) Finally, little mention is made of the pigment logical state of the algae that make up the phyto- composition of phytoplankton. Clearly, this aspect plankton population. is of profound importance in distinguishing the major In this broad view of photosynthesis, it is groups of algae and in any understanding of the difficult to make any clear distinctions between response of cells to varying illumination. However, those reactions specific for the photosynthetic process few data permit one to consider the pigments as and those of other aspects of cellular metabolism. end-products of phytosynthetic C assimilation and Yet this broad view of photosynthesis is the approach their consideration would be inappropriate here. of this paper. Thus, the discussion will be concerned As a result of such constraints, emphasis can with the overall end-products of photosynthesis — be defined more precisely. It attempts to examine the major storage products and those essential poly- the relative synthesis of lipid, polysaccharide, and mers required for cell growth. Emphasis will be protein during the photosynthetic assimilation of placed on the way in which the distribution of fixed inorganic carbon by marine phytoplankton. The con- carbon between the various types of products and cept that such an examination might be of relevance the resulting chemical composition of the algae might to the physiological ecology of phytoplankton de- be related to the environmental factors that control pends on the assumption that the synthesis of essential the growth and distribution of phytoplankton in the growth materials such as proteins, in comparison oceans. with that of "storage" products such as lipid and Some Boundaries polysaccharide, might be an important variable. This assumption is not based entirely on a priori reasoning, It is clearly neither possible, nor desirable, but depends, also, on empirical data obtained from to consider the entire field of algal metabolism in studies with algal cultures. the way suggested by the opening remarks. It is Before continuing, it is worth emphasizing necessary to draw some boundaries (albeit artificial) that by limiting the discussion to the overall products around the subject. Therefore, the following con- of C assimilation, several approaches to the physio- straints have been put on the subject matter to be logical state of phytoplankton that have attracted discussed: considerable attention in recent years will be omitted. 1) No consideration is given to the production of For example, a hyperbolic relationship between the small-molecular-weight metabolites such as the cellular content of a limiting nutrient and the growth individual amino acids, fatty acids, sugars, nucleo- rate has been observed for a number of different tides, etc. systems (Fuhs 1969; Caperon and Meyer 1972; 2) Similarly, no consideration is given to the detailed Paasche 1973; Droop 1974). Thus, the reasoning chemical composition of the various end-products suggests, measurement of the cellular content of a — the amino acid composition of proteins, the fatty- particular nutrient might indicate its possible role as acid components of lipids, the nature of the poly- a limiting saccharide material, etc. factor in the environment. Similarly, the kinetics of uptake of such nutrients might be expected ' Present address: Center for Environmental & to be a useful indicator of the factor that controls Estuarine Studies, University of Maryland, P.O. Box 775, phytoplankton growth. Other examples of attempts Cambridge, MD. to measure the physiological state of phytoplankton 83 include measurements of C: chlorophyll ratios with the "laboratory weed" Ch/ore/la. Differences (Steele and Baird 1961), carotenoid: chlorophyll in algal types, densities of suspensions, concen- ratios (Yentsch and Vaccaro 1958), adenylate energy trations of essential nutrients, etc., make such doubts charge (Falkowski 1977), and cellular content of such understandable. However, failure to recognize the parameters as chlorophyll a, ATP, N, C, and P (e.g. earlier work with this important alga can lead to Salcshaug and Holm-Hansen 1977; Sakshaug 1977). an embarrassing announcement of a discovery made Thus, an approach that emphasizes the relative three decades earlier. The papers of Spoehr and importance of selected end-products of C assimilation Milner (1949), Myers (1949), Myers and Johnston is only one of several possible methods of relating (1949), and Myers and Cramer (1948) are examples characteristics of cell chemistry and metabolism of work with Chlore/la that formed the starting to the controlling environmental factors. Lack of point for later studies. consideration of the other approaches is not to deny their importance. It merely recognizes the context CHEMICAL COMPOSITION into which this particular paper fits, as well as the restriction of space. The work of Spoehr and Milner (1949) estab- lished the essential characteristics of metabolism in The Approach Ch/ore/la, i.e. that the alga is, in essence, a protein- synthesizing organism and that its metabolism can vary considerably with changing environmental con- The paper is divided into four basic parts: ditions. A typical elementary composition of organic 1) A consideration of early work with Chlorella H, 28.5% 0, and 2) Some discussion of data from cultures of other matter in Chlore/la is 53% C, 7.5% points out, 8-10% N algae, notably marine diatoms. 10.8% N. As Myers (1962) 3) A summary of the major factors that influence implies a protein content of 50-60%. This predo- photosynthetic products in algal cultures. minantly protein-synthesizing type of metabolism can 4) An analysis of available data from natural popu- be forced into a form involving synthesis of non-N- lations of marine phytoplankton. containing compounds such as carbohydrate or lipid. In each part, two aspects of the data are con- This can be achieved by prolonged N depletion, or by sidered: (a) the chemical composition of the cells transferring cells that grow at low light intensities to and (b) the biochemical characteristics of the photo- higher intensities. synthetic process, as revealed by the products of Myers (1962) pointed to an anomaly in the "an essentially protein- C assimilation. The cellular composition of the history of biology, i.e. that Ch/ore/la, was selected as a cells represents the "true" end-product of all syn- synthesizing organism, organism for study of photosynthesis, a pro- thetic reactions (including those of photosynthesis) standard of carbo- in which the cell integrates the various processes cess thought to be an exclusive synthesis over time periods equivalent to generation times.. hydrate." Studies of the biochemical characteristics of photo- synthetic C assimilation can be viewed as attempts PHOTOSYNTHETIC CHARACTERISTICS on the part of the scientist to understand some reac- The predominant synthesis of protein in healthy tions (generally measured over relatively short times) growing cells of Chlore//a , and the variability of such that contribute to the integrated picture presented a phenomenon with changing environmental condi- by the cellular composition. Perhaps surprisingly, tions, are confirmed by studies of certain charac- there are few studies in which both the chemical teristics of the photosynthetic process. The work of composition and the patterns of C assimilation have Myers (1949) illustrates the amount of information been included in the same investigation. Also, there that can be inferred from a simple measurement of the are many more data on the chemical composition assimilatory quotient (A.Q. = CO., absorbed/O., of algae, particularly with natural populations, than evolved). Table 1 summarizes such data and illustrates on the products of C assimilation. Thus, precise how the previous history of the cell and environmental comparisons are difficult. However, it will be seen conditions côntrol the relative synthesis of carbo- that the juxtaposition of chemical composition and hydrate and protein. biochemical aspects of C assimilation raises impor- One other point emerges from the early work of tant questions and such questions will become a Myers (1946a, b, 1949). Growth of Chlore/la sat- central theme for the paper. urates at a lower light intensity than does photo- synthesis. This observation puzzled Myers and it Early Work with Chlorella continues to be of relevance to phytoplankton growth and photosynthesis in the sea. It is popular for present-day phytoplanktologists Thus, this early work with Chlore/la established to doubt the validity of considering results obtained the essential features of algal photosynthesis. It also

84 TABLE 1. Some characteristics of photosynthesis in Chlorella. (From Myers 1949.)

Balanced equations for growth with:

Nitrate 1.0 NO i :+ 5.7 CO2 + 5.4 H20 --> C 2119,02.2N + 8.502 + 1.0 OH-

A.Q. = -0.69

and Ammonium 1.0 NH; + 5.7 CO2 + 3.4 H20 -> C 71-192022N + 6.25 02 + 1.0 H+

A.Q. = -0.91

Previous history A.Q. (CO2 /02 ) Cho/N (inferred)

low light high light Grown low light -0.68 -0.88 normal 4 h high light -0.4 - > normal 3 d dark -0.91 - 0.96 < normal Normal -0.86 N-deficient + NO!/ - 0.74 2 N-deficient - NO - 0.99

initiated an awareness of the close links between observations also apply to other types of algae, C and N assimilation, an awareness that continues notably those that might be more important con- to be prominent in present-day studies of the physio- stituents of phytoplankton populations. logical ecology of marine phytoplankton. CHEMICAL COMPOSITION Other Species Tables 2 and 3 reproduce data that support the idea that the prominent synthesis of protein observed After the initial work with Chlorella , it became with Chlorella can also be measured in other micro- important to establish the extent to which the major algae from several different groups. Data such as these

TABLE 2. Chemical composition of various species of marine phytoplankton. The algae were harvested in exponential phase. ("Continuous" light, 18°C, harvested during exponential growth, g's between 8 and 36 h.) (After Parsons et al. 1961.)

% dry wt. Protein carbohydrate Fat

Chlorophyceae Tetraselmis maculata 52 15.0 2.9 Dunaliella sauna 57 31.6 6.4 Chrysophyceae Monochrysis lutheri 49 31.4 11.6 Syracosphaera catterae 56 17.8 4.6 Bacillariophyceae Chaetoceros sp. 35 6.6 6.9 Skeletonema costattun 37 20.8 4.7 Coscinodiscus sp. 17 4.1 1.8 Phaeodactylum tricornutum 33 24.0 6.6 Dinophyceae Amphidiniwn car/eni 28 30.5 18.0 Exuviella sp. 31 37.0 15.0 Myxophyceae Agmenellum quadriplicatum 36 31.5 12.8

85 TABLE 3. Lipid, carbohydrate, and protein content of diatoms (expressed as a percentage of the ash-free "dry" weight). (After Lewin and Guillard 1963.)

Organism Fat Carbohydrate Protein

Rhabdonenta adriaticum (from plankton) 44.16 13.6 13.25 Chaetoceros decipiens (from plankton) 27.94 13.2 21.50 Chaetoceros sp. (unialgal culture, growing exponentially) 9.5 9.2 48.6 Skeletonema costatunt (unialgal culture, growing exponentially) 7.7 34.1 60.6 Coscindiscus sp. (unialgal culture, growing exponentially) 4.2 9.5 39.5 Phaeodactylunz tricormitum (unialgal culture, growing exponentially) 7.1 25.9 35.7 Phaeodactylum tricolmitunt (fusiform cells from 16-d pure culture) 38.6 2.2 46.5 Phaeodactylum tricornuttan (oval cells from 16-d pure culture) 26.6 21.1 37.7 Cerataulina bergonii (unialgal culture, grown 2-4 wk) 14.76 26.72" 58.52 Chaetoceros lauderi (unialgal culture, grown 2-4 wk) 16.46 21.10" 62.44 Skeletonema costatum (unialgal culture, grown 2-4 wk) 21.93 34.55" 43.52 Leptocylindrus dan icus (unialgal culture, grown 2-4 wk) 20.73 33.02" 46.25

"By difference.

form the basis for the idea that species differences are TABLE 4. The protein:carbohydrate ratios of seven spe- not significant in determining the gross chemical cies of marine diatoms grown for various periods in batch composition of the cells. Possible uncertainty sur- culture. (After Myklestad 1974.) rounds the question of lipid synthesis in diatoms. It has sometimes been suggested that diatoms synthesize fats Days of growth rather than carbohydrates as storage products (re- 0 2 6 12 ferences cited in Lewin and Guillard 1963). In young exponentially growing cultures of diatoms, there is Chaetoceros debilis 2.40 1.10 0.63 0.53 little evidence to support this generalization. How- Chaetoceros ever, it appears valid for certain species of diatoms curvisetus 1.36 1.24 0.88 0.58 (Lewin and Guillard). Chaetoceros affinis var willei 2.16 1.51 0.31 0.23 The same variability of metabolism described Chaetoceros socialis 1.61 1.26 0.42 0.40 earlier for ChloreIla has been observed in cultures Thalassiosira gravida 2.31 1.59 0.83 0.55 of other types of algae. Thus, prolonged growth in Skeletonema costatum 1.86 1.21 0.34 0.18 batch cultures under conditions that create severe Thalassiosira N deficiency cause an increased synthesis of non-N- fluviatilis 0.60 0.39 0.16 0.11 containing compounds (lipids or carbohydrates) at the expense of protein. Collyer and Fogg (1955) described the reciprocal relationship between protein Healey (1975) provides an extensive review and fat content during batch growth of six species of literature that deals with the physiological indi- belonging to the Chlorophyceae, Euglenophyceae, cators of nutrient deficiency in algae. Part of that Xanthophyceae, and Bacillariophyceae. Table 4 review considered the cellular contents of protein describes the observations of Myklestad (1974) on and carbohydrate. Healey recognized an overlap the changing protein:carbohydrate ratio during between nutrient-sufficient and nutrient-deficient batch growth of seven species of marine diatoms algae in their protein or carbohydrate contents. (a result also observed by Myklestad and Haug However, use of the protein: carbohydrate ratio 1972). Other data on the changing nutritional patterns reduced this overlap considerably. Healey also of algae are reviewed by Fogg (1959, 1965). recognized that use of the protein: carbohydrate

86 ratio as an indicator of nutrient deficiency can be tion. The pronounced synthesis of carbohydrate and unreliable when considering algae that store lipid lipid (at the expense of protein) under conditions (rather than carbohydrate) under nutrient deficient of N deficiency depends on relatively prolonged N conditions, so that a ratio of protein:carbohydrate starvation. The effects of more moderate, short- and lipid reduced the overlap still further. term N depletion are less clear. For example, several observations report little change in protein content PHOTOSYNTHETIC CHARACTERISTICS during batch growth of a number of different algae Few studies have considered the question of (e.g. Skeletonema obliquus, Thomas and Krauss whether the changes in chemical composition de- 1955; Nitzschia, Badour and Gergis 1965; and Ske- scribed above are paralleled by comparable changes in letonetna costatum, Handa 1969). Also, in studies of photosynthetic characteristic measured over relatively nitrate reductase activities in Chlorella, I. Morris short time periods. Fogg (1956) observed prominent (unpubl ished data) observed 1 ittle change in the protein of 'IC (supplied as P 4C]bicarbonate) incorporation content of cell-free extracts during the first 24 h after into protein by actively growing cells of Navicula transfer to N-free medium. It appears that changes in pelliculosa, and that this was replaced by incor- soluble, nonprotein N might be expected to reflect poration into lipid in N-deficient algae. Few other initial or mild N deficiency (compare data with Syrett studies have considered such aspects of photosynthesis (1953) on the patterns of metabolism during the re- and these are considered later, in a more detailed covery from N deficiency. analysis of the effects of environmental factors on the products of photosynthesis. Glover (1974) measured the photosynthetic assimilation of [HC]bicarbonate into the major end- products during N starvation of Phaeodactylum Summary of Major Effects tricontitum. The proportion of 14C incorporated into of Environmental Factors on protein increased in the 2 d immediately following Photosynthetic Products transfer to N-free medium (Table 5). This increase was most marked when a N source was provided Information in the previous two sections pro- during the assimilation of HC but also occurred vides a basic description of the central features of to some extent in the absence of such a source. Thus, the photosynthetic products in algae. In essence, during the initial stages of N deficiency, there appears microscopic algae are protein-synthesizing organisms to be an atempt to conserve the synthesis of protein with a capacity for metabolic diversity that allows at the expense of storage (soluble?) N, so the dramatic extensive variability to be superimposed on this changes observed with prolonged N starvation cannot predominant synthesis of protein. The nature of this be extrapolated immediately to less severe conditions. variability is regulated by the environmental con- The original work on the effects of N depletion ditions and the physiological state of the cell. In generally involved starvation of batch cultures. In particular, high light intensities and N depletion essence, this represents an unbalanced situation. This appear to impose an enhanced synthesis of non-N- might be of relevance to cerain conditions in natural containing storage products such as carbohydrates or phytoplankton populations, for example, the transient lipids. This effect of environmental factors appears conditions that occur immediately after a spring to be more important than any effects of species bloom and the accompanying depletion of N from differences. the water. Of relevance to other situations, however, This section considers more precise effects of might be the balanced growth that occurs in N-limited specific environmental factors, in order to establish chemostat continuous cultures. Studies of the effects some basic framework of understanding from cultures of increasing N limitation on the chemical compo- that can be used to analyze data with natural popula- sition of cells growing in chemostats have tended tions. The variables affecting the nature of the photo- to emphasize the C:N ratio. Increases in this ratio synthetic products to be considered are the following: with increasing N limitation have been reported nutrient status, light, and temperature. widely (see references, Goldman et al. 1979). The extent to which changes in the N content can be NUTRIENT STATUS equated with changing protein contents is uncertain. The only nutrient to be considered here is ni- The possible importance of stored soluble N com- trogen (N). This restriction is not intended to deny pounds under conditions of N sufficiency and any the importance of studies of other nutrients, notably tendency to conserce protein synthesis during mod- P and Si. It simply reflects the fact that studies of erate N limitation makes such a practice. Despite the effects of nutritional status on photosynthetic this danger, it seems reasonable to assume that part products have tended to emphasize N. of the changing N content with different degrees of The effects of N depletion on the products of limitation is related to changing protein levels. For photosynthesis presented earlier is an over simplifica- example, Glover (1974) observed a reduction of

87

TABLE 5. Pattern of photosynthesis following transfer of Phaeodactylum tricornutuni to nitrogen-free medium (d 3). Cells were resuspended in fresh, nitrogen-free, and old media before the addition of radioactive bicarbonate. Results are expressed as a percentage of the total carbon incorporated. (From Glover 1974.)

Nitrogen-free medium

Days 3 5 7 11 15

Ethanol soluble 54.6 46.9 52.9 48.2 64.4 Polysaccharide 30.3 34.4 28.2 45.2 25.9 Protein 15.1 18.7 18.9 6.6 9.7

Old medium

Days 3 5 7 11 15

Ethanol soluble 38.9 35.0 53.2 42.4 55.5 Polysaccharide 44.6 44.1 21.2 42.7 32.4 Protein 16.5 20.9 25.6 14.9 12.1

Fresh medium

Days 3 5 7 11 15

Ethanol soluble 49.9 52.6 40.4 56.2 61.0 Polysaccharide 40.0 21.1 31.8 20.3 30.5 Protein 10.1 26.3 27.8 23.5 8.5

75% in protein content (per cell) with a 10-fold algae: alternating light/dark periods, effects of light decrease in growth rate of Phaeodactylum tricormt- intensity, and influence of light quality. tum growing in NOi-limited chemostats. Morris et al. (1974) examined changes in the Alternating light 'dark periods - Changing products of ["C]bicarbonate assimilation with in- chemical composition and products of photosynthesis creasing degrees of N limitation of P. tricornutum during the life cycle of Chlore/la ellipsoidea syn- growing in a NO:i-limited chemostat. These workers chronized in light/dark cycles have been studied emphasized the way in which the proportion of extensively by Tamiya's group (e.g. review of Tamiya "C incorporated into protein increased with increas- 1957). Increasing ratios of carbohydrates to proteins ing N limitation. However, a N source was present during the light period appear to be general. The during ''C assimilation and thus reflects the pattern of dark period is a time when carbohydrates are trans- photosynthesis accompanying addition of a N source formed into a number of cell materials, including to N-limited algae. Konopka and Schnur (1980b) re- proteins. The changes in carbohydrate:protein ratio ported that N-deficient chemostats of a cyanobac- do not appear to result from any significant changes terium Merismopedia tenuissima showed a reduced in the protein content (per cell), but result from in- proportion of "C, incorporated into protein when creasing carbohydrate content when the light is super- compared with cultures grown at maximum growth imposed on a more-or-less constant protein level rate in nutrient-sufficient conditions. Unlike the (Kanazawa 1964). Comparable data have been re- experiments of Morris et al. (1974), the measurements ported by Darley et al. (1976) for synchronized of Konopka and Schnur (1980b) were made in the cultures of Navicula pelliculosa. absence of added N. Interestingly, increasing degrees A simplistic assumption states that this synthesis of P, S, or C deficiency had little or no effect on the of storage products during the day occurs because proportion of "C incorporated into protein. Also, of an excess of energy supply over the immediate Glover (1977) observed increased proportions of "C requirements of the cells for growth. Cohen and incorporated into protein with increasing Fe limitation Parnas (1976) doubted this assumption. Rather, they in Phaeodacty/um tricormmon and Isochrysis gal- suggested that the production of storage material is bana a tightly regulated phenomenon, linked directly to future requirements and designed to produce an opti- LIGHT mal policy for the metabolism of storage materials. Three aspects of light availability influence the For the most part, their arguments are theoretical nature of the photosynthetic products in cultures of and lead to the expression of a model (related to

88

the more general model for reserve materials in all TABLE 6. Effect of light intensity on the cellular compo- microorganisms (Palmas and Cohen 1976). Although sition of Eglena cells harvested in the logarithmic phase Cohen and Parnas provide supporting evidence from of growth. (After Cook 1963.) experiments with cultures of Chlatnydomonas rein- hardii, it is uncertain how general are their con- pg/cell clusions. In particular, their model predicts (and their experimental data confirm) (1) that the synthesis Protein/ of the storage product (starch) occurs towards the Light intensity Pro- Para- para- (foot candles) tein mylum Lipids mylum end of the light period and (2) that the duration of starch synthesis decreases with increasing light 65 178 38 188 4.67 intensity. The authors recognize the fact that such 120 234 27 176 8.65 observations conflict with others (e.g. Cook 1966). 190 242 64 158 3.78 400 246 194 159 1.27 Light intensity — Most works on the effect of 1200 235 378 158 0.62 377 112 0.62 light intensity on photosynthesis have emphasized 3000 236 changes in pigment content and physiological char- acteristics of the photosynthetic process. Less infor- mation is available on how light intensity modifies protein fraction increased at reduced light intensities. the overall synthesis of the end-products of photo- An alternative way to state the same fact is to say that synthesis. Myers (1946a, b, 1949) reports the way in protein synthesis saturates at lower light intensities which higher light intensities cause enhanced syn- than does total photosynthesis. Comparable data thesis of carbohydrate to be superimposed on the showing lower irradiances required to saturate incor- predominantly protein-synthesizing metabolism of poration of '''CO, into protein than that into a Chlorella. Cook (1963) observed the same phenom- "polysaccharide and nucleic acid" fraction have been enon in Euglena (Table 6, Fig. 1). This effect of high reported for a freshwater cyanobacterium Meris- light intensity can be observed by measuring both the mopedia tenuissima (Konopka and Schnur 1980a). chemical composition of Cells grown at different light Interestingly, this increased proportion of "C incor- intensities and from measurements of the biochemical porated into protein at reduced irradiances could not be characteristics of photosynthesis (e.g. assimilation observed with nutrient-deficient cultures (cf. later quotients in Myers 1949, and products of "C discussion with natural populations). Myers (1949) assimilation in Cook 1963). Comparable results were pointed out the interesting puzzle that growth of reported by Morris et al. (1974) for Phaeodactylum Chlorella saturated at a lower light intensity than did tricornutum. This last named work emphasized the fact photosynthesis. Beardall and Morris (1976) reported that the proportion of fixed "C incorporated into the the same observation for Phaeoclactylum tricornutum

40

32

24

e 16

0 200 400 600 800 1000 1200 2800 3000 Irradiance ( foot candles) FIG. 1. Effect of light intensity on the paramylum and protein contents of Euglena gracilis. (After Cook 1963.)

89 120 X-- X 0.06

0 Cells grown ot 0.7 K I x -0 0 0 100 0.05 -• Cells grown at 12 KI x 70< 80 0.04 ezt 0 .e ; iü 4-e 60 0.03

• 0 2 2 c.; 40 0.02

20 0.01

0 2 4 6 8 10 12 14 " 24 Irradiance ( K1 x )

FIG. 2. Effects of light intensity on photosynthesis (•, 0) and growth (x) of Phaeodactyhun tricornutum. (After Beardall and Morris 1976.)

(Fig. 2). To some extent, this difference between the number of different algae (Hauschild et al. 1962a, light saturation curve for photosynthesis and that for b; Wallen and Geen 1971a). Some of the data from growth depends on the light history of the cells. The Wallen and Geen's work are presented in Table 7. difference is greatest when photosynthesis is measured Experiments with different light qualities are not as with cells previously grown at high light intensities straightforward as might be supposed. Ensuring equal (Fig. 2). Previous growth at low light intensities energy availability and absorption is not sufficiently reduces the saturating intensity for photosynthesis and rigorous to rule out the possibility that apparent effects the curve for carbon assimilation resembles that for of the spectral quality of light are related more to the growth (Fig. 2). amounts of light energy absorbed. It is unlikely that such doubts modify interpretation of results such as

Light quality - The stimulatory effect of blue those of Table 7, but they do make it difficult to light on protein synthesis has been observed for a extrapolate from such data to natural populations.

TABLE 7. Effect of light on the distribution of NC in Dunaliella tertiolecta and Cyclotella nana. (After Wallen and Geen 1971a.)

%total "C fixed % ' 1C in ethanol-insoluble Spectral quality of light Et0H-sol Et0H-insol Protein Carbohydrate

D. tertiolecta white 89.9 10.1 92.0 6.8 blue 28.8 71.2 96.1 1.3 green 33.2 66.8 95.9 1.8

C. nana white 67.3 32.7 92.6 5.4 blue 35.9 64.1 95.3 1.3 green 39.8 60.2 98.8 1.2

90 TEMPERATURE saccharide synthesis or in the incorporation of radio- activity into an ethanol-soluble fraction (containing Most studies of the effects of temperature on the lipids) (Table 8). chemical composition of algae and on the products of photosynthesis originate with the papers of JOrgensen and Steemann Nielsen (1965), Steemann Nielsen and Natural Phytoplankton Populations JOrgensen (1968), and JOrgensen (1968). These workers proposed that algae adapt to suboptimal SUMMARY OF CONCLUSIONS temperatures by synthesizing more of the enzymes FROM CULTURES required for photosynthesis. The evidence was indirect The work with laboratory cultures described and was apparently supported by the fact that cells of earlier forms the basis for measurements of photo- Skeletonema costatum grown at 7°C contained more synthetic products in natural populations of marine protein than did those grown at 20°C (Jegensen phytoplankton. The essential characteristics may be 1968). This observation was apparently confirmed summarized thus: by Morris and Farrel (1971) with Phaeodactylum 1) Algae such as those that constitute phytoplankton tricornututn and Dunaliella tertiolecta. Later, how- populations are protein-synthesizing microorganisms . ever, the observations of Morris and Glover (1974) 2) There is a potential for considerable metabolic suggested that the reported differences between algae diversity and the nature of this is determined by grown at suboptimal temperatures and those from environmental factors and the physiological state of higher temperatures resulted from the changes that the cells. occur during batch growth. Parameters such as photo- 3) High light intensities and prolonged N depletion synthetic ability and enzyme activities were maximal can impose excessive synthesis of storage materials early in batch growth at higher temperatures. Apparent (lipids or polysaccharides) in addition to the essential enhanced activities in cells grown at lower tem- synthesis of protein. peratures resulted from the comparison made late in 4) The effect of short-term or moderate N depletion exponential growth. When earlier peaks at higher is less clear. There is some evidence that the synthesis temperatures were taken into account, the apparent of essential proteins is conserved so the proportion effects of reduced temperatures were not observed. of fixed C incorporated into protein increases during Morris et al. (1974) reported how temperature moderate N deficiency. affected the assimilation of [''CI bicarbonate into the 5) Species differences appear to be less important major end-products of photosynthesis (Table 8). The than environmental factors and the physiological proportion of HC incorporated into protein was state of the cells. greater at the higher temperatures used in the experi- Most of the relevant work with natural popu- ment. However, cells previously grown at the lower lation has been concerned with the chemical compo- temperatures incorporated a higher proportion of sition of particulate matter in the oceans. Much less '4 C into protein than did those algae previously emphasis has been placed on following the fixation grown at higher temperatures. These changes in the of [ HC]bicarbonate into the major end-products of proportion of "C incorporated into protein were photos ynthesis . accompanied by reciprocal changes in either poly- CHEMICAL COMPOSITION TABLE 8. Effects of temperature on the products of photo- Most work on the chemical composition of synthetic ["C]bicarbonate assimilation by Phaeorlactylunz suspended matter in the oceans has been concerned tricotnuturn. The results are expressed as % of total "C with the major elements , in particular, the C:N:P ratio fixed. (After Morris et al. 1974.) (the so-called Redfield ratio). The recent paper of Goldman et al. (1979) documents the literature on this Experi- subject. These authors make the interesting point that Growth mental % "C fixed over wide areas of the oceans the ratio approximates to tempe- tempe- rature, rature, Ethanol- Hot-TCA- 100:16:1 (C:N:P) and that this ratio is characteristic °C °C soluble soluble Protein of algae growing at or near their maximum growth rate. Such observations raise the interesting possibility 18 18 45.2 33.5 21.3 that phytoplankton growing in the nutrient-poor sur- 12 42.8 41.8 16.1 face layers of many parts of the open ocean may be 7 34.9 55.8 9.3 growing at, or near, their maximum growth rate. Other observations have also failed to find physiological 7 18 44.2 22.7 33.1 indications of nutrient (particularly N) deficiency 12 46.8 26.1 27.1 7 58.1 20.9 21.0 among phytoplankton of subtropical oceanic regions (e.g. Morris et al. 1971).

91 Goldman et al. (1979) recognized that such turnover times for both total C and protein C. The physiological indices as the Redfield ratio can only latter values were generally less than the former indicate the relative growth rates. Organisms with but the values were comparable. Later in this review, widely different absolute values of p. max can show there will be further discussion of protein turnover similar C:N:P ratios when growing near those maxi- times. For the time being, it is worth noting that mum growth rates. This is a specific example of Maita and Yanada did not measure incorporation of how an emphasis on the physiological state of phyto- "C into protein (contrast later). Rather, they meas- plankton fails to address the problem of the absolute ured total C assimilation and, from the measured rate at which the population is growing. protein C: total C ratio calculated the rates of protein There have been fewer studies of the chemical synthesis; therefore, agreement between the two composition of phytoplankton populations at more types of turnover values might be expected. detailed levels than the elementary composition. Handa et al. (1972) reported the way in which the PATTERNS OF C ASSIMILATION IN protein and carbohydrate contents of particulate NATURAL POPULATIONS OF carbon varied along a transect that extended from MARINE PHYTOPLANKTON 48°N to 68°S in the Pacific Ocean. The proportion C and N assimilation ratios — Goldman et al. of particulate carbon found in carbohydrate did not (1979) emphasized the way in which the C:N compo- vary significantly. However, the proportion that sition ratios of phytoplankton populations varied over occurred as protein was greatest at the higher latitudes a small range of values (generally between 5:1 and (lower temperatures). Packard and Dortch (1975) 9-10:1). The variability in the measured C:N assimi- reported the way particulate protein N varied between lation ratios is much greater. Table 9 summarizes oceanic and upwelling regions in the North Atlantic. some reported values for C: N assimilation ratios (cited The ratio of protein N: chlorophyll was higher at the by Slawyk et al. 1978). It is clear that the range of oceanic stations (2.83) than at the upwelling stations values is much greater than that observed for the C:N (0.54). These authors calculated that only 20% of composition ratios. For example, in the work of the sestonic protein N at the oceanic stations was Slawyk et al., C:N assimilation ratios from 4-45:1 associated with phytoplankton. The comparable occurred in a region where the composition ratio only figure at the upwelling stations was 65%. Hitchcock varied from 5.4-9.1:1 (integrated over the euphotic (1977) described the changes in particulate carbo- zone). Also, observed changes in the assimilation ra- hydrate in an upwelling zone off West Africa. The tios did not parallel those in the composition ratios. For concentration of particulate carbohydrate in the example, Eppley and Renger (1974) noted that the surface waters was directly proportional to the con- C:N composition ratios of natural phytoplankton centration of chlorophyll and the carbohydrate: populations were lowest at the highest growth rates. chlorophyll ratio decreased with depth in the euphotic The reverse trend was noted for the C:N assimilation zone. Handa (1975) also observed marked diurnal ratio. Eppley and Renger pointed out that part of the changes in the carbohydrate content of particulate reason for such a discrepancy lies in the methodology. matter that appeared in the day and disappeared at night in Mikawa Bay. This author did not measure protein. TABLE 9. Some examples of C:N assimilation ratios from Few of these studies have tried to link such natural populations of marine phytoplankton. (Cited by distributions of particular constituents to the physio- Slawyk et al. 1978.) logical state of the phytoplankton. Haug et al. (1973) and Sakshaug and Myklestad (1973) observed a de- C:N assi- crease in the protein:carbohydrate ratio when natural milation ratio Location Reference phytoplankton populations in Trondheimsfjord be- came N or P deficient. In an upwelling area off 0-12:13-2 Discontinuity layer Goering et al. (1970) South Africa, Barlow (1980) also observed a de- in Pacific Ocean clining ratio of protein:carbohydrate when a diatom McCarthy (1972) bloom collapsed after nutrient depletion. Also, Maita 12:40 Off S California and Yanada (1978) reported changing ratios of par- 12:76 Various eutrophic MacIsaac and Dugdale ticulate protein C to total particulate carbon during and oligotrophic (1972) seasonal changes in waters off Hokkaido. The ratio waters vas highest during early spring, late fall, and winter, 0.2:18 Central Gyre, Eppley et al. (1973) when the standing stock (measured either as chloro- N Pacific phyll or as particulate C) was least. In other words, 4:45 NW African Slawyk et al. (1978) the concentration of particulate protein changed less upwelling region than did that of total C. These authors also calculated

92 The 11 C technique for measuring C assimilation is a relationship between rates of C assimilation and the genuine tracer technique. Use of ' 5N, however, can growth rates of the phytoplankton populations. At perturb the system in the sense that carrier N is added the basis of such questions, is a comparison between with the isotope. The extent of this perturbation is observations with natural populations and those made greatest in those waters where the ambient concentra- with laboratory cultures, where the effects of par- tions of the appropriate N compounds are lowest. ticular environmental conditions can be identified Under such conditions, addition of carrier N can result under controlled conditions. It is convenient to con- in observed rates of ' 5N assimilation that are greater sider the way in which the products of photosynthetic than those at ambient N concentrations. C assimilation are influenced by the three major Assimilation of C into major end-products of environmental factors: light, temperature, and nu- photosynthesis — Little work is available on the trient availability. patterns of ' 1 C assimilation by natural populations of phytoplankton. The papers of Olive and Morrison Light — Three aspects of the role of light can be (1967), Olive et al. (1969), Wallen and Geen (1971b), considered: (a) diurnal changes, (b) light intensity, and Morris et al. (1974), Morris and Skea (1978), Hitch- (c) light quality. cock (1978), Smith and Morris (1980a, b), and Morris et al. (1981) present most of the available data. Also, a) Diurnal patterns of l'C assimilation Li et al. (1980) made such measurements with Oscil- Morris and Skea (1978) described the time latoria (Trichodesmium) thiebautii in the Caribbean courses for [ HC]bicarbonate incorporation into na- Sea. This section summarizes the main findings from tural phytoplankton populations from the temperate such observations, as well as some unpublished obser- waters of the Gulf of Maine. Incorporation into protein vations from our laboratory. continued during both the light and dark periods; the The broad aim of such work on photosynthetic assimilation into protein in the dark was paralleled by products of natural phytoplankton populations is loss from the polysaccharide fraction. Thus, the pro- to question whether such measurements can (1) de- portion of NC incorporated depends on the length of scribe the physiological state of the populations, (2) incubation and increases after the period in darkness. identify those environmental factors that control Comparable results were obtained in the oligotrophic phytoplankton growth, and (3) comment on the waters of the Sargasso Sea (Fig. 3), although the

60

50

to b 40

30

o 20

O)

10

0 0 6 12 18 24 0 6 12 18 24 TIME IN HOURS FIG. 3. Incorporation of ["C]bicarbonate into the ethanol-soluble (0), the hot-TCA-soluble (A), and the protein (0) fractions when surface phytoplankton populations from the southem Sargasso Sea (27 °00'N, 64°00'W; July 27-30, 1976) are incubated under natural light for 24h. (A) Radioactivity in each fraction and the total incorporation (x); (B) data expressed as percentages of the total "C assimilated.

93 into polysaccharide is greatest at the optimal irra- diance levels and is reduced at both lower and higher 50 levels. The proportion incorporated into protein, low- molecular-weight metabolites and (to a lesser extent) lipid showed the reverse trend, being lowest at the 40 optimal intensities and increasing at both lower and (r) higher levels. Morris and Skea (1978) emphasized the way 0 30 cn in which such an effect of light intensity depends ce on the nutrient status of the cells. That is, enhance- 1— ment of the relative incorporation of ' 'C into protein z tu 20 o by reduced light intensities was less marked with summer populations from the Gulf of Maine (sampled from surface layers after the establishment of strati- 10 fication and depletion of nutrients from the surface 0600 1000 HOUR S waters) than with populations sampled before the spring bloom; Fig. 5 and 6 present a similar observa- FIG. 4. Proportion of "C assimilated to ethanol-soluble tion for coastal and oceanic regions of tropical waters. (D), hot-TCA-soluble (€3), and protein (MI) fractions when These figures describe data from two different cruises were surface phytoplankton (the same station as in Fig. 3) and compare stations off the mouth of the Orinoco incubated for 4 h at different times of day. River with oceanic stations from the Caribbean Sea (Fig. 5) or the Western Atlantic Ocean (Fig. 6). Stimu- disappearance of radioactivity from the polysaccha- lation of the proportion of 11 C incorporated into ride fraction during the dark was less marked than that protein by reduced light intensities was observed at the observed by Morris and Skea (1978). coastal stations but was not significant at the oceanic In another type of experiment, surface popu- stations. Possibly, , the physiological state of phyto- lations from the Sargasso Sea were harvested at plankton populations from oceanic regions might be different times of the day and incubated for a constant revealed by emphasizing this effect of light intensity time (4 h). The proportion of ' 'C incorporated into rather than the direct measurements of the patterns of protein was greatest early in the morning and de- '1 C assimilation (see below). creased during the afternoon (Fig. 4). This decrease In waters that show the stimulating effect of was accompanied largely by increasing proportion of reduced irradiances on the proportion of ''C incor- '1C incorporated into the ethanol-soluble fraction porated into protein, the observation can be made (also containing lipid material). In such experiments, by incubating a single sample at a range of light changes due to diurnal fluctuations in the metabolism intensities or in simulated in situ experiments, in of phytoplankton are superimposed on effects of light which samples from different depths are incubated intensity. It is unlikely that changes in irradiances at light intensities corresponding to the depth from during the day can explain wholly the results in which they were taken. Fig. 4. If the reduced proportion of ' IC incorporated It is interesting to question whether light inten- into protein was caused solely by elevated light sities influence the time course of ' 'C incorporation intensities, one might have expected to observe a into different products during a 24-h light/dark cycle. similar difference between midday and the end of Earlier, I commented on the hypothesis of Cohen the day. and Parnas (1976) in proposing that, in unicellular b) Effects of light intensity algae that grow in light/dark cycles, synthesis of Work with both cultures and natural populations storage products such as polysaccharide during a light (Morris et al. 1974; Morris and Skea 1978) has period was a regulated process. These authors pro- confirmed the generalization that incorporation of posed that synthesis of such storage products would [ HC] bicarbonate into protein saturates at a lower light occur late in the light period and that higher light intensity than does incorporation into storage products intensities would reduce the period over which syn- such as polysaccharides. Thus, in such experiments thesis of storage products would occur. We have the proportion of H C incorporated into protein in- observed no such phenomenon in natural phyto- creases at reduced light intensities . plankton populations. For example, Fig. 7 describes Li et al. (1980) observed similar effects of sub- the way in which the light intensity of incubation optimal and supraoptimal light intensities on the influences the proportion of "C incorporated into patterns of ''C assimilation by the marine cyanobac- the various products but does not appear to alter terium Oscillatoria thiebautii (Trichodesinium). In the "shape of the curve" for incorporation during this organism, the proportion of ''C incorporated the light period.

94 STATION No.1 0.02)4 chl a • L -1 STATION No.9 I.79pg chl a• 70 • 60

z 50

IiJ 1— • cr o-o a 40 • •I cr 30 — \c, •

20

l0 24 48 72 96 120 0 24 48 72 96 120 E

FIG. 5. Effect of irradiance on the proportion of [ 14 Olbicarbonate incorporated into protein (•), polysaccharide (•), and ethanol-soluble material (0) when surface populations from two stations were incubated for 4 h in an incubator illuminated with warm white fluorescent tubes. Data were collected on a cruise of theRV Eastward between Feb. 14 and 19, 1974. Station 1 was south of Puerto Rico (17°18'N, 65°00'W) and station 9 at the mouth of the Orinoco River (09°33'N, 60°31'W).

50

40 d te ra

o 30

Incorp 20 C 14 10 %

0 Low m.w. Lipid Polysocc. Protein Low m.w. Lipid Polysocc. Protein

FIG. 6. Effect of irradiance on the proportion of [Hgbicarbonate incorporated into the major end-products of photo- synthesis at two stations during cruise EN-034 of the RV Endeavor. Station 17 is at the mouth of the Orinoco River (5.69 t.t.g chl a • L- ') and station 24 (0.07 ,r..tg chl a •L - I) — 200 km NNE of station 17. Samples collected from a depth corresponding to 30% of surface irradiance were incubated at the range of intensities.

95

i 70, 2000 p I I C I E '•'\ I IMO d., \ 1 • I 80 0 N 1 • I A .______.y• 10 - I • 70 - - I .11Nii / /111----...__--• 60 _ I 8 - 1 1 - 7 / 6- I 40 - _ i . • 1 • I • I -_. / I / 4 - /I / //•itô''...... -- ---111 .-----À 11 • / I a 1 •->111■., I -•...... ---1 11,-,-....z.À_____.4 ,t:.:>,,,,...... „.N.,,,, j____• 2 10 °----1°J-----. •,0 1 o °1-----1 --o 1 1 1 I 1 V ; i o , . ,, ,0 , , 0 4 8 12 16 20 240 4 8 12 16 20 24 0 4 8 12 16 20 24 Hours Hours Hours FIG. 7. Time courses of incorporation of [nClbicarbonate into polysaccharide (broken line), protein (A), lipid (x), and low-molecular-weight metabolites (0); (•) shows total incorporation. Water samples were taken from a depth corresponding to 30% of surface irradiance and incubated at (A) 100%, (B) 30%, and (C) 1% of this surface value. The top part of Fig. 8 shows the changing irradiances during the period of incubation. Experiments were performed at station 17 (9°33.3'N, 60°42.7'W) of EN-034 cruise of the RV Endeavor, Apr. 5, 1979.

c) Light quality Temperature —It is difficult to identify precise and The pattern of [ HC]bicarbonate assimilation also direct effects of temperature on the patterns of photo- changes with depth in the water column. Figure 8 synthesis. Temperature changes with geography, or illustrates the increasing proportion of "C incor- with time in temperature waters, are generally ac- porated into protein with increasing depth. This is companied by changes in the degree of stratification generally accompanied by a decreasing proportion and, thus, in nutrient availability. One of the most incorporated into polysaccharide. The experiments interesting effects of temperature has been reported by described in Fig. 8 involved incubating surface water Smith and Morris (1980a, b) for phytoplankton popu- at the various depths. Comparable data are obtained lations from the Southern Atlantic Ocean. At locations with in situ incubations when water sampled from with extremely low ambient temperatures ( < -1.0°C) a particular depth is incubated at that depth (Fig. 9). and low incident irradiance levels (maximum, 0.2 The effects of incubating samples at increasing ly • min- ) as much as 80-90% of the [''C] bicar- depths described in Fig. 8 and 9 resemble those bonate assimilated during an 8-h day was incorporated obtained by alterations in light intensities achieved into lipid. At such stations, there was insignificant with neutral density filters. Work in our laboratory has incorporation into protein. At other locations with not detected any significant difference between the " higher" temperatures (- 0°C) the pattern of photo- two approaches. From such work, therefore, it might synthesis resembled that observed in other regions be suggested that changing patterns of photosynthesis of the oceans, with prominent incorporation into with depth are related solely to changes in light inten- polysaccharide and significant synthesis of protein. sity and not to light quality. The opposite conclusion The significance of such a striking synthesis was reached by Wallen and Geen (1971b). These of lipid material at the extremely low temperatures authors observed increasing proportions of [NC] of polar regions is unknown. Populations that showed ethanol-insoluble-bicarbon ate being incorporated into such a phenomenon did not show high lipid content compounds (mainly protein) with increasing depth in in the particulate matter. Thus, incorporation of Saanich Inlet and Indian Arm, B .C. Comparisons with "C into lipid material might reflect synthesis of their earlier work with laboratory cultures (Wallen and specialized lipids (possibly associated with mem- Geen 1971a) led them to emphasize light quality and branes) or the synthesis of lipid which is degraded not light intensity. at night.

96

% CARBON ASSIMILATED forward as might be supposed. Earlier, it was em- 10 20 30 40 50 60 70 phasized that a shift from protein synthesis to exces- . o • sive synthesis of non-N-containing compounds such as APRIL 2,1975 lipid and carbohydrate accompanied N deficiency. \ However, the result depends on extreme N starvation. 10 i • During more moderate N deficiency, it appeared that the cellular Content (and rates of synthesis) of non- protein N might be more affected than that of protein. 20 — \ \ / • Also, enhanced synthesis of materials such as lipids and polysaccharides at the expense of protein appears to be specific for N deficiency. Deficiency of other 30 _ \ 7 • nutrients (e.g. Fe, P, etc.) appears to be accompanied by a conservation of protein synthesis so that, under (,) 40 \ / such conditions, the proportion of C incorporated into — • protein can be enhanced. cc 0 • Attempts to relate patterns of C assimilation to r- the nutritional status of natural phytoplankton popu- w \ / M A Y 29,1975 lations have generally been of two kinds: (a) studies of 10 — 1 • à y seasonal changes in temperate waters and (b) geogra- phical comparisons of coastal and oceanic regions. Both types of studies are complicated by changes in the 20 _ \ •//' . species composition of the phytoplankton populations. That is , observed changes may be related to changes in \ / \ the dominant algae of the population and not to a 30 - • changing environemental variable such as nutrient availability. However, the earlier discussion of data / ( \ from laboratory cultures emphasized the way in which 40 - • differences between species appeared to be less im- • portant than effects of physiological state. It therefore seems reasonable to relate measured changes in pat- \ \ AUGUST 11,1975 i \ \ terns of C assimilation to the physiological state of the 10 - • phytoplankton .

20 a) Seasonal studies _ / \ 7 The papers of Hitchcock (1978) and Morris and Skea (1978) present the few available data on seasonal 30 _ changes in the products of ric] bicarbonate assimila- tion. Hitchcock observed increasing proportions of HC being incorporated into ethanol-soluble material oà. 40 i during the development of a spring diatom bloom in lower NatTagansett Bay. This increase was ac- FIG. 8. Proportion of P1 C]bicarbonate incorporated into companied by a decrease in the proportion incorpo- the ethanol-soluble fraction (0), the polysaccharide fraction rated into both the polysaccharide and protein frac- (A), and protein (•) when surface water from a station tions. These changes could be correlated with a in the Gulf of Maine (43°42'N, 69°39'W) was incubated decline of nutrients (notably N) and a for 4 h at various depths. reduction in the protein:carbohydrate ratio in the particulate matter. The data of Morris and Skea (1978) are less Nutrient status — It is attractive to consider the possi- clear. In their studies at a coastal station in the Gulf bility that measurements of the flow of C into the major of Maine, Morris and Skea could not detect any end-products of photosynthesis might be a means of increased incorporation of "C into ethanol-soluble detecting nutrient deficiency among phytoplankton or polysaccharide material with the decline of nu- populations. It is clear from the earlier discussion that trients from the surface waters. Indeed, these authors the nutritional status of algae in laboratory culture has emphasized the way in which the proportion of ' 'C a profound effect on the products of photosynthesis. incorporated into protein was higher in the summer However, immediate extrapolation from such work populations. They also detected a transient and with cultures to natural populations is not as straight- striking increase in the proportion incorporated into 97 CARBON ASSIMILATED INTO PROTEIN 20 30 30 40 50 60 20 30 40 50 ... i i I • 1 I i

I\ APRIL 2,1975 \ MAY 29,1975 • 0 AUG 11,1975

\\IN. L.1 10 • 0 •,,...0 ...... 0 • cc 1- w 2 20 0 • a_ lii o \ i 30 0

\ \ 40 0 •

FIG. 9. The proportion of [1 'C]bicarbonate incorporated into protein when surface water ( • ) and water from the various depths (0) were incubated for 4 h throughout the euphotic zone; same station location as in Fig. 8. protein after the collapse of the first (of two) spring Moving offshore was accompanied by a reduction blooms. The authors questioned whether this in- in the proportion incorporated into lipid and (to a creased relative incorporation into protein might lesser extent) low-molecular-weight metabolites. indicate nutrient limitation. Accompanying this change was an increase in the proportion of "C appearing in protein and poly- b) Geographical comparisons of coastal saccharide (64-72%). The significance of these and oceanic regions changes is unknown and cannot be related too imme- To date, the only measurements of this kind diately to observations from laboratory cultures. are those pursued in the author's laboratory. The summarizes the data from the paper present section PRODUCTS OF PHOTOSYNTHESIS of Morris et al. (1981) and emphasizes the coastal AND PHYTOPLANKTON GROWTH and oceanic regions of the Caribbean Sea as well as the adjacent regions of the Western Atlantic. Table Elsewhere in this Bulletin, Eppley points out the 10 summarizes the main observations on the pro- impossibility of extrapolating from measured rates of portion of [' IC]bicarbonate incorporated into the main assimilation of an element (e.g. carbon) to growth end-products of photosynthesis. At the inshore sta- rates (doubling times) of the algae constituting the tions (showing higher chlorophyll concentrations and phytoplankton population. The reason most corn- measurable levels of inorganic nutrients in the surface monly recognized for this difficulty is the difficulty of layers) there was approximately equal incorporation measuring phytoplankton biomass, separate from over an 8-h light period into all four fractions studied. other particulate matter (including detritus). However,

TABLE 10. Rates of photosynthesis and major end-products of [wC]bicarbonate assimilation at coastal (17 and 15) and oceanic regions of the Caribbean Sea (28) and the Western Atlantic Ocean (24). All measurements were made on samples taken from depth corresponding to 30% of surface light and incubated at that intensity for 8 h. Data were collected on cruise EN-034 of RV Endeavor (Mar. 22-Apr. 14, 1979). (After Morris et al. 1981.)

Photosynthesis % total "C assimilated

Station Chi a jig C • L- ' •11-1 g C • i.tg Chl - ' •11-1 Low m.w. Lipid Polysacc. Protein

17 5.69 10.0 1.76 26.4 27.1 26.5 20.0 15 2.05 2.70 1.32 27.7 24.1 24.9 23.3 24 0.07 0.67 10.4 21.1 17.9 41.0 23.0 28 0.09 0.36 4.0 19.2 8.9 41.1 30.8

98

Eppley describes an alternative (and fundamental) dif- biomass as chlorophyll and the rates of C assimilation ficulty. Eppley recognizes "unbalanced growth" when might suggest that the growth of phytoplankton was at any one time, the relative rates of assimilation of comparable at both types of stations. Yet the "skew- elements or of the synthesis of various cellular com- ed" pattern of C assimilation with little detectable ponents can differ from their relative proportions in the protein synthesis and dominant incorporation into cell. The particular problem centers around the fact lipid at certain stations makes such an extrapolation that algae of the phytoplankton population grow and from C fixation to growth untenable. Indeed, the divide in an environment of alternating light/dark stations at which significant lipid synthesis occurred periods. Under such conditions, rates of processes showed lower C-specific rates of C assimilation than measured over relatively short periods in the light did the others (Table 11). The PC: chl ratios were might not reflect net rates occurring over a 24-h (or abnormally high at such stations and would normally longer) light/dark cycle. Eppley also identified other be interpreted as indicating the influence of non- conditions under which phytoplankton populations phytoplankton carbon in the measurement of PC. may exhibit unbalanced growth. However, an abnormal C metabolism might also lead The rates of synthesis of the different end- to such unusual PC: chl ratios. products of photosynthesis might be expected to The difficulty in extrapolating from rates of C indicate the conditions under which measurements assimilation to growth rates of the phytoplankton of C assimilation during the light period might more population introduces the question of whether rates or less closely reflect growth rates of the algal popu- of C incorporation into selected polymers might be lation. Thus, conditions promoting enhanced syn- a closer indication of growth rate. That is, if one of thesis of storage products might yield rates of C the reasons for a discrepancy between total rates of assimilation that greatly overestimate the growth C assimilation and doubling times is in the variable rates. Under conditions of reduced synthesis of stor- synthesis of storage products during the light period, age products, the rate of C assimilation might reflect measurement of C incorporated into a compound more closely the synthesis of those essential polymers more closely linked to growth and division might required for growth and division. be expected to be a closer measure of growth rates. The data from studies of Smith and Morris From our studies in coastal and oceanic regions of (1980b) in the waters around Antarctica illustrate the Caribbean Sea and Western Atlantic Ocean we such a problem (Table 11). At stations 13 and 18 have attempted calculations of turnover times based significant incorporation of ''C into lipid was ob- on C incorporation into the various products (Table served. Such stations show higher assimilation ratios 12). Such calculations are difficult and the results (C fixed per unit chlorophyll) than do stations 23 must be interpreted as preliminary and with caution. and 29 where there is no such marked incorporation Assumptions about the C-content of the lipid, poly- into lipid. Thus, measurements of phytoplankton saccharide, and protein of the particulate matter

TABLE 11. Characteristics of photosynthesis at four stations from the Southern Atlantic Ocean. (After Smith and Morris 1980b.)

Biomass Rates of photosynthesis (8 h) % of total wC fixed (8 h)

Assimi- Station Chl a PC Per volume lation No. C specific "Polysacc/ (p,g• PC:chl a (g C- ' • h- ') (g C•g chl- ' • h- ') (h- `) Lipid Protein metabolite"

13 0.04 167.8 4195 0.25 5.70 0.002 70.3 3.7 26.0 18 0.21 328.0 1433 0.92 4.44 0.003 48.6 9.7 41.7 23 0.85 196.2 231 1.80 2.11 0.009 4.2 11.9 83.9 29 0.62 144.3 233 1.03 1.67 0.007 25.1 17.9 57.0

Rates of photosynthesis (24 h) % of total ' 4C fixed (24 h)

Assimi- "Polysacc/ Per volume lation No. C specific Lipid Protein metabolite" 13 0.11 2.53 0.001 88.9 2.8 8.3 18 0.37 1.77 0.001 79.2 5.8 15.0 23 0.75 0.87 0.004 13.6 21.2 65.2 29 0.67 1.09 0.005 6.3 26.9 66.8

99 TABLE 12. Carbon-specific rates of rqbicarbonate assimilation by phytoplankton populations at two coastal stations (15 and 17) and two oceanic stations in the Caribbean Sea (28) and the Western Atlantic (24). Rates of incorporation into particulate matter ("total") and into the various end-products were normalized to particulate C or to the C content of the various fractions (particulate protein, polysaccharide, and lipid were measured colorimetrically and the C contents estimated from published chemical compositions of laboratory cultures of marine algae). Further details are in legend to Table 10. (After Morris et al. 1981.)

Assimilation rates (11-1 )

Station Time of incubation, Total Lipid Polysaccharide Protein

15 8 0.021 0.003 0.009 0.001 17 8 0.051 0.028 0.012 0.002 24 8 0.015 0.009 0.042 0.007 28 8 0.006 0.003 0.029 0.004

15 20 0.014 0.014 0.032 0.005 17 20 0.017 0.066 0.030 0.013 24 20 0.006 0.003 0.014 0.006 28 20 0.002 0.001 0.008 0.001

were based on values from the literature for various have failed to approach the problems of growth and cultures of marine algae (see Morris et al. 1981 for division of a population of microbes growing in an details). Also, distinguishing phytoplankton-protein, environment of alternating light/dark periods. phytoplankton-lipid, etc., from nonphytoplankton However, from Table 12, it is apparent that the material in the particulate matter is not possible. turnover times calculated from incorporation into Although, therefore, the precise values presented in protein are more comparable to those calculated Table 12 are open to considerable question, the dis- from total 'C incorporation when an incubation crepancy between rates calculated from incorporation time of 20 h (includes the dark period) is considered. into protein and those from total C assimilation is interesting. Not only are the calculated turnover times Acknowledgments for particulate protein longer than those for par- ticulate C, the direction of change in comparing The work associated with the author's laboratrny coastal and oceanic locations is opposite. Calculations was undertaken while he was at the Bigelow Labo- based on total C assimilation show expected higher ratory for Ocean Sciences, West Boothbay Harbor, turnover times at the inshore stations. The reverse Maine, USA 04575. Colleagues who participated is true when rates of C incorporation into protein are in the work and to whom I am grateful are H. E. normalized to particulate protein C. This similar Glover, A. E. Smith, W. Skea, J. C. Laird, and W. reverse trend is also apparent (but with much shorter Li. The work was supported by NSF grants OCE doubling times) when inccirporation into polysac- 75-15104, OCE 75-21128, OCE 77-18722, and charide is used as a basis for the calculations. DPP 78-23833. It might be argued that measurements of C as- similation over 24 h might reduce some of the dis- References crepancy between such assimilation and growth rates. "unbalanced" syn- If the discrepancy results from the BADOUR, S. S., AND M. S. GERGIS. 1965. Cell division thesis of storage products in the light period, inclusion and fat accumulation in continuously illuminated mass of the dark period in the time over which assimilation cultures. Arch. Microbiol. 51: 94-102. is measured might be expected to "cancel out" this BARLOW, R. G. 1980. The biochemical composition of effect. Data fiom 20 or 24 h are included in Tables 11 phytoplankton in an upwelling region off South Africa. and 12, and some basic questions raised above remain. J. Exp. Mar. Biol. Ecol. 45: 83-93. During the last 20-30 yr, there has been a growing BEARDALL, J., AND I. MORRIS. 1976. The concept of awareness of the need to measure "instantaneous" light intensity adaptation in mariihe phytoplankton: some experiments with Phaeodactyhun tricormitum. rates and to avoid enclosing water samples for mea- Mar. Biol. 37: 377-387. surements over periods as long as 24 h. Being forced to CAPERON, J., AND J. MEYER. 1972. Nitrogen limited consider 24-h incubations, therefore, is a dramatic growth of marine phytoplankton. I. Changes in popu- illustration of the way in which measurements of pro- lation characteristics with steady-state growth rate. ductivity in terms of ["C] bicarbonate assimilation Deep-Sea Res. 19: 601-618.

100 COHEN, D., AND M. PARNAS. 1976. An optimal policy HAUG, A., S. MYKLESTAD, AND E. SAKSHAUG. 1973. for the metabolism of storage products in unicellular Studies on the phytoplankton ecology of the Trond- algae. J. Theor. Biol. 56: 1-18. heimsfjord. I. The chemical composition of phyto- COLLYER, D. M., AND G. E. FOGG. 1955. Studies on fat plankton populations. J. Exp. Mar. Biol. Ecol. 1 1: accumulation by algae, I. Exp. Bot. 6: 256-275. 15-26. COOK, J. R. 1963. Adaptations in growth and division in HAUSCH1LD, A. H. W., C. D. NELSON, AND G. KROT- Euglena effected by energy supply. J. Protozool. 10: KOV. 1962a. The effect of light quality on the products 436-444. of photosynthesis in Chlore/la vulgaris. Can. J. Bot. 1966. Photosynthesis activity during the division 40: 179-189. cycle in synchronized Euglena gracilis. Plant Physiol. 1962b. The effect of light quality on the products 41: 821-825. of photosynthesis in green and blue-green algae and DARLEY, W. M., C. W. SULLIVAN, AND B. E. VOLCANI. in photosynthetic bacteria. Can. J. Bot. 40: 1619- 1976. Studies on the biochemistry and fine structure 1630. of silica shell formation in diatoms. Planta 130: 159- HEALEY, F. P. 1975. Physiological indicators of nutrient 167. deficiency in algae. Res. Dey. Dir. Freshwater Inst. DROOP, M. R. 1974. The nutrient status of algal cell in Tech. Rep. 585: 00-00. continuous culture. J. Mar. Biol. Assoc. U.K. 54: 825- HITCHCOCK , G. 1977. The concentration of particulate 855. carbohydrate in a region of the West Africa upwelling EPPLEY, R. W., AND E. M. RENGER. 1974. Nitrogen zone during March 1974. Deep-Sea Res. 24: 83-93. assimilation of an oceanic diatom in nitrogen limited HITCHCOCK, G. L. 1978. Labelling patterns of carbon-14 continuous culture. J. Phycol. 10: 15-23. in net plankton during a winter-spring bloom. J. Exp. EPPLEY, R. W., J. M. SHARP, E. H. RENGER, M. J. Mar. Biol. Eco!. 31: 141-153. PERRY, AND W. G. MORRISON. 1977. Nitrogen JeGENSEN, E. G. 1968. The adaptation of plankton algae assimilation by phytoplankton and other microorgan- II. Aspects of temperature adaptation of Skeletonetna isms in the surface waters of the central North Pacific costatum. Physiol. Plant. 21: 423-427. Ocean. Mar. Biol. 39: 111-120. JORGENSEN, E. G., AND E. STEEMANN NIELSEN. 1965. FALKOWSKI, P. G. 1977. The adenylate energy charge in Adaptation in plankton algae. Mem. Ist. Ital. Idrobiol. marine phytoplankton: the effect of temperature on 18: 37-46. the physiological state ofSkeletonema costatum (Grey.) KANAZAWA, T. 1964. Changes of amino acid composition Cleve. J. Exp. Mar. Biol. Ecol. 27: 37-45. of Chlore/la cells during their life cycle. Plant Cell FOGG, G. E. 1956. Photosynthesis and formation of fats Physiol. 5: 333-354. in a diatom. Ann. Biol. London N. S. 20: 265-285. KONOPKA, A., AND M. SCHNUR. 1980a. Effect of light 1959. Nitrogen nutrition and metabolic patterns intensity on macromolecular synthesis in cyanobac- in algae. Symp. Soc. Exp. Biol. 13: 106-125. teria. Microbiol. Ecol. 6: 291-301. 1965. Algal cultures and phytoplankton ecology. 1980b. Biochemical composition and photo- Univ. Wisconsin Press, Madison, Milwaukee, and synthetic carbon metabolism of nutrient-limited cul- London. p. 11-126. tures of Merismopedia tenuissima (Cyanophyceae). FUHS, G. W. 1969. Phosphorus content and rate of growth J. Phycol. (In press) in the diatoms Cyclotella nana and Thalassiosira LEWIN, J. C., AND R. R. L. GUILLARD. 1963. Diatoms. fluviatilis . J. Phycol. 5: 312-321. Annu. Rev. Microbiol. 17: 373-414. GtovEn, H. E. 1974. Studies on the biochemistry and Li, W. K. W., M. E. GLOVER, AND I. MORRIS. 1980. physiology of Phaeodactylum tricornututn, Ph. D. Physiology of carbon photoassimilation by Oscil- thesis, London. 195 p. Univ. London, latoria thiebautii in the Caribbean Sea. Limnol. 1977. Effects of iron deficiency on Csochrysis Oceanogr. 25: 447-456. galbana (Chrysophyceae) and Phaeodactylunt tri cor- MACISAAC, J. J., AND R. C. DUGDALE. 1972. Interactions nutum (Bacillariophyceae). J. Phycol. 13: 208-212. of light and inorganic nitrogen in controlling nitrogen GOERING, J. J., D. D. WALLEN, AND R. M. NANNAN. uptake in the sea. Deep-Sea Res. 19: 209-232. 1970. Nitrogen uptake by phytoplankton in the dis- MALTA, Y., AND M. YANADA. 1978. Particulate protein continuity layer of the eastern subtropical Pacific in coastal waters, with special reference to seasonal Ocean. Limnol. Oceanogr. 15: 789-796. variation. Mar. Biol. 44: 329-336. GOLDMAN, J. C., J. J. MCCARTHY, AND D. G. PEAVEY. MCCARTHY, J. J. 1972. The uptake of urea by natural 1979. Growth rate influence on the chemical compo- populations of marine phytoplankton. Limnol. Ocea- sition of phytoplankton in oceanic waters. Nature nogr. 17: 738-748. (London) 279: 210-215. MORRIS, I., AND K. FARRELL. 1971. Photosynthetic HANDA, N. 1969. Carbohydrate metabolism in the marine rates, gross patterns of carbon dioxide assimilation diatom Skeletonema costatum. Mar. Biol. 4: 208-214. and activities of ribulose diphosphate carboxylase in 1975. The diurnal variation of organic constituents marine algae grown at different temperatures. Physiol. of particulate matter in coastal water, p. 125-132. Plant 25: 372-377. In S. Mon i and G. Yamamoto [ed.] ITBP synthesis. MORRIS, I., AND H. E. GLOVER. 1974. Questions on Univ. Tokyo Press. the mechanisms of temperature adaptation in marine HANDA, N.,K. YANAGI, AND K. MATSUNAGA. 1972. phytoplankton. Mar. Biol. 24. 147-154. Distribution of detrital materials in the Western MORRIS, I., AND W. SKEA. 1978. Products of photosyn- Pacific Ocean and their biological nature. Mem. Ist. thesis in natural populations of marine phytoplankton Ital. Idrobial. 29 Suppl. 53-71. from the Gulf of Maine. Mar. Biol. 47: 303-312. 101 MORRIS, I., A. E. SMITH, AND H. E. GLOVER. 1981. PARSONS, T. R., K. STEPHENS, AND J. D. M. STRICK- Products of photosynthesis in phytoplankton off the LAND. 1961. On the chemical composition of eleven Orinoco River and in the Caribbean Sea. Limnol. species of marine phytoplankton. J. Fish. Res. Board Oceanogr. (In press) Can. 18: 1001-1016. MORRIS, I., C. M. YENTSCH, AND C. S. YENTSCH. SAKSHAUG, E. 1977. Limiting nutrients and maximum 1971. The physiological state of phytoplankton from growth rates for diatoms in Narragansett Bay. J. Exp. low-nutrient sub-tropical water with respect to nitrogen Mar. Biol. Ecol. 28: 109-123.

as measured by the effect of ammonium on dark carbon SAKSHAUG, E., AND O. HOLM - HANSEN. 1977. Chemical dioxide fixation. Limnol. Oceanogr. 16: 859-868. composition of Skeletonema nostrum?, and Pavlova MORRIS, I., H. E. GLOVER, AND C. S. YENTSCH. 1974. (Monchusis) Imbed as a function of nitrate- phos- Products of photosynthesis by marine phytoplankton: phate- and iron-limited growth. J. Exp. Mar. Biol. the effect of environmental factors on the relative rates Ecol. 29: 1-34. of protein synthesis. Mar. Biol. 27: 1-9. SAKSHAUG, E., AND S. MYKLESTAD. 1973. Studies on MYERS, J. 1946a. Culture conditions and the development the phytoplankton ecology of the Trondheimsfjord of the photosynthetic mechanism III. Influence of light III. Dynamics of phytoplankton blooms in relation intensity on cellular characteristics of Chlore/la. J. to environmental factors, bioassay experiments and Gen. Physiol. 29: 419-427. parameters for the physiological state of populations. 1946b. Culture conditions and the development J. Exp. Mar. Biol. Ecol. 11: 157-188. of the photosynthetic mechanism IV. Influence of light SLAWYK, G., Y. COLLOS, M. MINAS, AND J-R. GRALL. intensity on photosynthetic characteristics of Chlorella. relationship between carbon to nitrogen J. Gen. Physiol. 29: 429-440. 1978. On the particulate matter and growth 1949. The pattern of photosynthesis in Chlore/la, composition ratios of the from the northwest p. 349-364. In J. Franck and W. E. Loomis [ed.] rate of marine phytoplankton Mar. Biol. Ecol. 33: Photosynthesis in plants. Am. Soc. Plant Physiol. Iowa African upwelling area. J. Exp. State College Press. 119-131. 1962. Variability of metabolism in algae. Dtsch. SMITH, A. E., AND I. MORRIS. 1980a. Synthesis of lipid Bot. Ges. Nene. Folge Nr. 1: 13-19. during photosynthesis of phytoplankton of the Southern MYERS, J., AND M. CRAMER. 1948. Nitrate reduction Ocean. Science (Washington, D.C.) 207: 197-199. and assimilation in Chlore/la. J. Gen. Physiol. 32: 1980b. Pathways of carbon assimilation in phyto- 93-102. plankton from the Antarctic Ocean. Limnol. Oceanogr. MYERS, J., AND J. A. JOHNSTON. 1949. Carbon and 25: 865-872. nitrogen balance of Ch/ore//o during growth. Plant STEELE, J. H., AND I. E. BAIRD. 1961. Relations between Physiol. 24: 111-119. primary production, chlorophyll and particulate MYKLESTAD, S. 1974. Production of carbohydrates by carbon. Limnol. Oceanogr. 6: 68-78. marine planktonic diatoms, I. Comparison of nine SPOEHR, H. A., AND H. W. Mt LNER. 1949. The chemical different species in culture. J. Exp. Mar. Biol. Ecol. composition of Chlore/la; effect of environmental 15: 261-274. conditions. Plant Physiol. 24: 120-149. MYKLESTAD, S., AND A. HAUG. 1972. Production of STEEMANN NIELSEN, E., AND E. G. JORGENSEN. 1968. carbohydrates by the marine diatom Chaetoceros The adaptation of plankton algae. I. General Part. affinis var willei (gran) Mustedt. I. Effect of the con- Physiol. Plant 21:401-403. centration of nutrients in the culture medium. J. Exp. SYRETT, P. J. 1953. The assimilation of ammonia by Mar. Biol. Ecol. 9: 137-144. nitrogen-starved cells of Chlore/la unlgaris. Ann. OLIVE, J. D. M., D. M. BENTON, AND J. KISHLER. Bot. London N.S. 17: 1-18. 1969. Distribution of C-14 products of photosynthesis TAMIYA, H. 1957. Synchronous cultures of algae. Annu. and its relationship to phytoplankton composition and Rev. Plant Physiol. 8: 309-334. rate of photosynthesis. Ecology 50: 380-386. THOMAS, W. AND KRAUSS. 1955. Nitrogen OLIVE, J. D. M., AND J. M. MORRISON. 1967. Variations H., R. W. in distribution of I IC in cell extracts of phytoplankton metabolism in Scenedesmus as affected by environmen- living under natural conditions. Limnol. Oceanogr. 13: tal changes. Plant Physiol. 30: 113-122. 383-391. WALLEN, D. G., AND G. H. GEEN. 1971a. Light quality PAASCHE, E. 1973. Silicon and the ecology of marine in relation to growth, photosynthetic rates and carbon plankton diatoms. II. Silicate uptake kinetics in five metabolism in two species of marine plankton algae. diatom species. Mar. Biol. 19: 262-269. Mar. Biol. 10: 34-43. PACKARD, T. T., AND Q. DORTCH. 1975. Particulate 197 lb. The nature of the photosynthate in natural protein-nitrogen in North Atlantic surface waters. phytoplankton populations in relation to light quality. Mar. Biol. 33: 347-354. Mar. Biol. 10: 157-168. PARNAS, H., AND D. COHEN. 1976. The optimal strategy YENTSCH, C. S., AND R. F. VACCARO. 1958. Phyto- for the metabolism of reserve materials in microorgan- plankton nitrogen in the oceans. Limnol. Oceanogr. isms, J. Theor. Biol. 56: 19-55. 3: 443-448.

102 Photosynthesis Measurements on Natural Populations of Phytoplankton: Numerical Analysis

CHARLES L. GALLEGOS AND TREVOR PLATT Department of Fisheries and Oceans, Marine Ecology Laboratory, Bedford Institute of Oceanography, Dartmouth, N.S. B2Y 4A2

Introduction is linear. The curve exhibits its maximum slope here (the initial slope). At some higher value of In this paper we shall consider the numerical light intensity, the curve usually attains an upper analysis of the data from that kind of experiment bound (light saturated), where the slope has fallen in which a natural phytoplankton assemblage from to zero. At still higher light intensities, the rate of a particular depth is incubated at various light inten- change of photosynthesis with light may become sities in the presence of radioactive bicarbonate such negative (the photoinhibited range). When photo- that a curve can be constructed describing the inhibition is observed, the peak in the curve may relationship between photosynthetic rate and available be sharply defined, Fig. lb, or relatively broad, light for that particular sample. The principles in- Fig. le. Sometimes, no photoinhibition is observed volved will also be applicable to photosynthesis- over this range of light intensities, Fig. Id. light measurements made on cultures by means of Selection of a suitable model will depend on the oxygen electrode. the light range over which the measurements are Experiments of the type described can be used made. Many workers will be interested in working to map geographic distributions of assimilation in a restricted range of light intensities where number, or in models for computing production photoinhibition is unlikely ever to occur. In this under a square metre of ocean surface. However, case the models can be relatively simple compared our emphasis is on the physiological content of the with those necessary to represent photoinhibition. models. In that respect, we gain the maximum Mathematical representation of the photo- information when the light-saturation experiments synthesis—light curve below the onset of photo- are used to test hypotheses, for example, about the inhibition is dealt with in Jassby and Platt (1976), relative values of physiological parameters of popu- Platt and Jassby (1976), and Platt et al. (1977). lations sampled from different physical environments Models expanded to admit the possibility of photo- (e.g. within versus below surface mixed layers) or inhibition are treated in Platt et al. (1980) and Platt about the effects on parameter values of some imposed and Gallegos (1980). It is important to note that treatment (e.g. temperature variations). Numerical these models are of the empirical kind rather than analysis therefore involves choosing a mathematical the mechanistic kind. That is, they each describe a model of the photosynthesis—light relationship; family of curves according to a certain equation, choosing a parameterization of the model that is with particular members of the family being specified consistent with the way the measurements are made; by particular sets of the parameter values. The basic objectively estimating the parameters of the model equation describing each family is constructed, not and their confidence regions from experimental data; from physiological arguments, but from application and interpreting the parameter values in terms of of mathematical intuition to find a curve that matches what is known about their physiological basis. typical data. This being the case, the criterion of merit has Choosing the Model to be the relative success with which a particular family of curves can describe experimental data. We have to select a mathematical description Stated another way, that equation representing a for data of the kind shown in Fig. 1 (from Platt family of curves that describe the and Gallegos 1980), where the range of light inten- results of an experiment with the least residual variance when sities is from zero to the maximum that can occur averaged over a large number of experimental trials under natural circumstances. Figure la is a composite will be the equation of choice. In this way, Jassby schematic showing the various possibilities that can and Platt (1976) found that the hyperbolic tangent occur. Depending on how the measurements are equation was a consistently successful description made, there may be a small, negative assimilation of their experiments: at zero light intensity, usually interpreted as the magnitude of the dark respiration. At low light intensities, the photosynthesis—light relationship ( 1 ) P(I) = P„, tanh (od/Pm)

103 where P is the "C assimilation at light level /, P„, by Platt et al. (1977) and by Platt and Gallegos is the assimilation at saturating light, and a is the (1980). The number of parameters that are required initial slope. will depend on the degree of complexity that is sought Following such a routine for the selection of in the description. For example, referring again to an equation to describe a particular set of experiments Fig. 1, the dependance at low light can be described presupposes an objective method for fitting the by a single parameter, the initial slope a: equations to the data. This will be dealt with in aPD I detail below. (2) a — Again, we emphasize that this hyperbolic tangent al I- 0 equation was used as no more than an empirical description, and that its initial trial (Jassby and Here the superscript B implies that the value of Platt 1976) was based on no more sophisticated ' C assimilation (P) has been normalized to chlo- arguments than that it was known to describe a rophyll a. The adequacy of this one-parameter saturation-type equation. However, recent work description breaks down if P 74 0 at / = 0. In by Chalker (1980) has shown that, if the photo- other words, if it is required to allow for the dark synthesis-light curve is represented as a generalized respiration, an additional parameter R', equal to expansion in P, then the only possible solution for the magnitude of the intercept, will have to be the quadratic approximation is the hyperbolic tangent carried as an additive term. equation. This will be taken up in detail below. A linear description is justified only at low light levels. Otherwise, the curvature must be allowed Choosing the Parameters for through another parameter. A good choice is the height of the plateau Pe, . This is not the only Selection of the appropriate parameters for a choice. One could equally use the light level at photosynthesis-light model has been dealt with which the straight line with slope a cuts the plateau

0. b. 0.8 Peak and/or plateau 0

TION ci 0.6 ,manolonically ? decreasing region "pholoinhibition range" o 0.4 PRODUC re m 0.2 0_ IMARY

AVAILABLE LIGHT 140 280 420 560 700 intercept

) ( Wm-2

2.4 C. d. 5.0- 2.0 o 1.6 4.0 '- o 1.2 o 3.0

08 C.) 2.0 ›- co Cfl û- 0.4 — 1.0-i co Q_ I I • I 200 400 600 200 400 600

(W M-2 ) I (Wm" 2 ) FIG. 1. (a) A generalized photosynthesis-light curve; (b-d) Curves from experiments on natural assemblages of marine phytoplankton.

104 (/k ) or the light level at which P = 112 Pm (half- where Pg and a are the best estimates of their "true" saturation constant). Once the parameters are fitted, values. If the errors are normally distributed but not these alternatives may be found, one from the other, of constant variance, then the weighted sum of through simple identities. squared residuals To describe experimental data on photoinhi- bition, at least one additional parameter is required. 1 This can be f3, •a parameter independant of a, (5) J =-, — f(I J ; ,â, .M2 describing the negative slope at high irradiances. Those curves that show an extended plateau before where ol is the variance associated with the meas- the onset of photoinhibition (Fig. le) require a ured P'] , is the appropriate criterion to minimize final parameter IT to describe the irradiance at which (Bevington 1969). When the PI are means of rep- the slope of the curve becomes negative. licate samples, the computed variances may be used These summarize the minimum parameter sets for b-Y. However, uniform light fields are difficult necessary at each level of description. In every to produce, and minor differences in light intensity case, more parameters could be added and a better can produce large variations in P'3, particularly at fit obtained. But the additional or superfluous param- low light, such that true replicates may be difficult eters would have no biological meaning and no to obtain. It is possible that this problem may be further information could be gained. Also, the curves circumvented by modulating the position of rep- should be fitted by an objective method to a data licate bottles relative to the light field, thereby set containing sufficient points that reasonably precise exposing all bottles to the same average light during estimates can be made for each of the parameters. the course of an incubation. When replicates are not This is the subject of the next section. available, some assumption about the variances of the ei must be made. Usually it is assumed either that they are constant, or that they are proportional Estimating the Parameters to the observed value of P']. In either case the ap- and Confidence Regions propriateness of such an assumption should be examined at the conclusion of the fitting exercise, The fitting problem is to find the set of param- for example, by forming a plot of the residuals as eters that yield the "best" (in some sense) description a function of predicted value (Draper and Smith of the data in terms of the chosen model formulation 1966). and parameterization. Measurements of production In an earlier paper (Platt et al. 1980) we gave always contain errors, so that we may represent some attention to whether or not a weighted sum of the observed data squared residuals should be minimized when fitting parameters of light-saturation models. We displayed (3) P = f(lt; Pg, a, • • • ) + a figure showing the standard deviation of estimated production (from in situ estimates) plotted against where PI is the measured production, f(1; ; Pe the mean. The relationship was roughly linear, but a, . . .) is the functional form of the light-saturation when a linear weighting function was used to fit curve written in terms of irradiance, I, and pe, , the parameters, the fitting routine failed to converge a, and one or more photoinhibition parameters, on stable parameter estimates. The reason seems and/or respiration; ei is an error term that accounts to be that, across a large number of experiments, for any effects not incorporated in the model, as the variance in estimated production is roughly well as measurement error. The choice of the func- correlated with , but that within any given experi- tional form, f(li ; Pei , a, . . . ) has been discussed ment, the variance is independent of P for all except previously, but we note that the model formulation the very lowest values. Figure 2 illustrates this point. and parameter estimation problems are not entirely In most instances an unweighted sum of squared independent (see below). It is the statistical distri- residuals can probably be used provided the model bution of the error term, ei , that determines the is a gbod descriptor at the low light level. details of the fitting procedure. Man'y algorithms have been developed for In the simplest case in which the errors are performing the sum of squares minimization in non- independent and normally distributed with a constant linear parameter estimation problems. We shall not variance, the best parameters are those that minimize enter a detailed discussion of the mathematics of the sum of squared residuals (Bard 1974; Smith these techniques, but instead refer the reader to Smith 1979) (1979) and Bard (1974) for a comprehensive survey of the field. Basically all of the techniques are itera- tive, requiring initial or trial estimates of the param- (4) J = — f(1 .) ; , â , . )1 2 eters from which a search is conducted until the 105 0.75 ascertain that the fitted parameters indeed yield a the minimum 0.50 minimum of the error criterion, and that is unique within the region of reasonable parameter 0.25 estimates.

00 • If a relatively small number of experiments is 7 to be fitted, it may be simplest to write a short -0.25 routine that computes the sum of squared residuals -0.50 while stepping through a grid of trial estimates, and search for the minimum interactively (Silvert F 00 0.10 0.20 0.30 040 0.50 1979). This is the crudest of the pattern search tech- niques but has the advantage that it is simple to write, can be implemented on relatively unsophis- O• 0.75 ticated computers or programable calculators, and •«2. 0.50 also provides information that is useful for deter- -_1 0.25 mining confidence regions about the parameter estimates (see below). C 0,0 .. • ' Silvert (1979) presented two methods for de- cc -0.25 termining confidence regions for parameters in general, nonlinear, biological models. We now -0.50 discuss the application of these methods to a model of the light-saturation curve of photosynthesis. 00 0.5 1.0 1.5 2.0 2.5 30 3.5 ESTIMATED P B (mgC(mg Chle h- I The first method is a Monte Carlo approach, which is applicable when a packaged routine is used FIG. 2. Residuals about two fitted curves as a function of to obtain the parameter estimates. The procedure predicted PR. Variance is larger for curve with larger has been described by Bard (1974) and Smith (1979) Pe,, but for both curves, the variance is roughly independent and is implemented as follows. The best parameter of PB over most of the domain. estimates are obtained from a set of measurements of photosynthesis and irradiance. Those parameter parameter set yielding a minimum of the objective estimates and measured irradiance values are used function is located. The major difference among in the chosen model of the light-saturation curve to the algorithms is in the method of choosing the generate a set of simulated P I' values. The simulated updated parameter set at each iteration. The simplest ' values are "noise-corrupted" by adding a random algorithms, both conceptually and computationally, error terni having the distribution (known or assumed) are the pattern search techniques which employ only in the original analysis. If the errors are of constant information about present and past values of the variance, then an estimate of the variance, s 2 , is objective function and parameter estimates. These given by (Silvert 1979). are simple to program, and several examples of (6) s 2 = J/(11 — in) FORTRAN subroutines are given by Bevington (1969). More sophisticated are the so-called descent where J is given by Eq. (4), m is the number of methods, which employ the first (and sometimes parameters, and n is the number of data points. The higher) derivatives of the objective function with fitting routine is then used to estimate the parameters respect to the parameters in updating the parameter of the light-saturation model using the simulated, estimates. For electronic computers, various pack- noise-corrupted data. This procedure is repeated aged routines are available to minimize the sum of enough times (-100) to obtain a distribution of squared residuals by these methods, and a list of errors in the parameter estimates, from which con- programs available has been compiled by Bard fidence regions can be determined. A drawback of (1974). We have found that the modified Gauss — this procedure in the context of light-saturation mod- Newton method performs satisfactorily, fitting up to els is that the procedure implicitly assumes that the at least four parameters simultaneously (see e.g. model formulation is a perfect description of the Fig. lç). process. In models that are based on mechanistic Packaged routines have the advantage that assumptions (e.g. physical models in which the minimal programming effort is required of the user. equations are derived from Newton's laws) this is Also, it is reasonably certain that a minimum sum not a serious drawback. However, the models of of squared residuals will be found in an efficient the light-saturation curve are mainly geometric in (in the sense of minimal computation time) manner. their formulation. The Monte Carlo approach does However, a certain amount of checking, at least in not account for any systematic tendencies for the a pilot analysis, should be done. It is advisable to equations to over- or under-estimate production

106 the J squared residuals above, thissearchprocedurecanbeused confidence less than trial estimates interactive way, of sional spacewasdefined,centered a rangeofvaluesand combination and collected from values of below theirestimated computed byfirst cluding of productionand with where binations thatyield estimates istodetermine confidence by than somecritical for to account in (7) o is certain parameters yieldingsums (Smith 1979) same computing computing parameters „ Contours ofthe We illustrate the minimum, The second m - ,„ and an J - time. regions R is of the the is regions intercept B for limits tobespuriouslynarrow. (n the , of the -in) critical co E P FIG. below parameters the LabradorSea. such the confidence ; as principalaxes. parameters. upper parameters, with method discussedbySilvert(1979) of the (mg C(mg Chla)-1 h- 1) critical sum to locate is roughly±5%,10% and [ +( irradiance estimates. of the holding aand value. The sum the degrees 3. Two-dimensional m 15.25 1500 a term, 15.75 trends and sum and sum value are values, and define confidenceregionsat curve. Failure -20 y pointoftheF-distribution procedure belowwith of best estimates. contours of P, above itsestimated of R n n R A squared residualsateach Pe,, of m regions can E B are as of the minimum This squared residualswere a, and , in squared residualsless critical of (mgC was fittedto43 of freedom. ) until cause the FI71 -1.25 i i 1 t within 117. 1 squared residuals, squared residuals Pe, the rangesof All combinations was repeated Equation (1),in- ■ 'stepping through given above, \ about - \ (mg Chla) ... parameter com- RB. of about the e a \ value \ constant , sections this approach three-dimen- _ — \ As \ \ 111 the (1-y) Errors ■ for a,and±50%R -0.5 ■ parameter •.:.: computed ■ we noted

sum is given

\ :/

I -1 I i value, I in an h through pairs in well best data and -1 for of

Pe, 0.25 and the 95%jointconfidence different opticaldepths lations plankton way Scotia. data improve ourunderstanding gain regarding variation inthe tionship thathasbeen tical tation canbe is of objective nevertheless we phytoplankton populations. physiological mechanismsregulatinggrowth it canbeestimated. improvements were we mightanticipatethattherewouldbelittleto among samples, estimates allow precision eters. understanding stage of ration experimentsisto for for a.Theconfidence side confidence because errors slices through a R which furtherstudyiswarranted. 0.18 B of of prerequisite that The Figure 3shows One aspectofthe are collected application of from In Fig.3athe its estimated are examining this n purpose (mgC(mgChla) populations numerical analysisthatweadvanceour highly correlated.Precision . on the separated by I regions an ...... procedures ateverystep 0.22 of • Pe, • significance of the the confidence experimental programdesignedto off thecontinental in parameters have but much • us of A Case II, is roughly±5%, value, made inthe about also toidentifythoseparameters numerical analysis these proceduresbelow,with not onlytomakestatements tried toemphasizethat of region is physiology must 0.26 is bysamplingfromwidely value. two-dimensional variation and regions regions a much interestis 1 in photosynthesis-light rela- density whereas le h - Study of of the gain insightaboutthe It isat R be I w areas where of theterm, We illustrate B extend 50% are region differences observed for the met precision withwhich m about the oblique 0.30 with naturalphyto- -2 gradient. of the P the the of before interpre- correlated. -I shelf of the corresponding of the For interpretation of and ±10% to sections or curve. the vertical organisms: the light satu- parameter R example, on We wish of Nova the the axes B analysis param- the use , popu- either prac- until One The 107 of to interpret the results in terms of the effect of light For the near-surface sample, an equation just history on photosynthetic parameters; this requires for light saturation seems sufficient, so we choose that we first identify those parameters that show first the formulation and parameterization of Eq. (I). significant variations with depth. Objective estimation of the parameters Pe, a, and Deciding on a model and parameters, and esti- RB requires initial estimates. In our experience, it mating the parameters is an iterative process in which has been possible to obtain initial estimates of suf- we first try one fonnulation, evaluate the results ficient accuracy by visual inspection of the curves. after fitting the parameters, and increase the com- Based on Fig. 4a we used 2.0 mg C. (mg Chl a)' • 11- ' plexity of the model formulation if necessary. The as an initial estimate of Pe, . Alpha is more difficult first step is to examine P - 1 plots of the data (Fig. to estimate by inspection, but by extrapolating the 4). The points in Fig. 4a were taken from 5 m, initial slope, it appears to intersect Pe, at /k 40 the depth of penetration of 50% of surface irradiance, W • in-2 , which by the identity /k = P„,la gives and the sample in Fig. 4b was taken from 35 m, 0.05 mg C . (mg Chl a) • 11- ' • W- ' • m' as an initial or the I% light-depth. It is immediately apparent estimate for a . We use 0.0 as an initial estimate that different levels of complexity will be required for le . These initial estimates are then entered into to fit the two experiments because the deep sample whatever fitting procedure is being used. With the exhibits obvious photoinhibition at high irradiance, modified Gauss-Newton procedure we obtained whereas photoinhibition is lacking in the near-surface final estimates pg = 2.22, and a = 0.045, and sample. Although photoinhibition is outside the scope R B = —0,19. of our main discussion, we include it here because It is apparent that some formulation including it illustrates a situation in which the choice of model photoinhibition is necessary to describe the data forniulation is not trivial. of Fig. 4b. In the interest of parsimony we first a. SCOTIAN SHELF Sm b SCOTIAN SHELF Sm

2.5 2.5

-1 . • 2.0 -c 2.0 1- h -1

a) f I 1.5 4 1.5 I- Ch E ; 1.0 8 1.0 I- C(mg

mg

0.5 1- PB

0.0 0.0 100 200 300 400 500 600 700 100 200 300 400 500 600 700 - 0. 201i F I (Wm -2) 1 (WM -2)

C.SCOT IAN SHELF 35m d.SCOTIAN SHELF 35m

2.5 r 2.5 r

1 -1 - 2.0 2.0 h h -1 )' a) cl • I I 1.51- • • 1.5 Ch Ch /r 1. • • 1.0 C(mg C(mg

mg

mg • 0.5 0.5 PB PB 6

h J 0.0 0 MO 20 0 300 400 500 600 700 0 100 200 300 400 500 600 700 I (Wm -2) 1 (Wm -2 )

FIG. 4. (a-b) Photosynthesis-light experiment for station off the continental shelf of Nova Scotia from (a) 5 m and (b) 35 m; (c-d) empirical light-saturation models fit to data in (a) and (b).

108 tried the three parameter photoinhibition equation We applied this test to the residuals in Fig. 5. given in Platt et al. (1980). Trial estimates of the The computed F for the near-surface sample was initial slope and maximum were chosen as before; 1.73, well below the critical value of 7.2 needed we have found that it is sufficient to use a very small to reject the model formulation. For the deep sample, number (about 10 -4) as an initial estimate of p, the the computed F was 8.09, indicating that increasing parameter controlling the strength of photoinhibition. the level of complexity in the model formulation However, if a grid search technique is being used is worthwhile to reduce the systematic error. to estimate the parameters it may be worthwhile to This means that we must return to the model for- obtain a closer initial estimate of p. This can be mulation stage of the numerical analysis procedure. done by noting that the intensity, //, , at which the We note that the systematic trend in the residuals for initial slope of the photoinhibited portion of the the deep sample occurs at high irradiances, indicating curve extrapolates to 0 is given by the identity that an equation permitting an extended plateau /b = P3 113 (see Platt et al. 1980). From visual before the onset of photoinhibition is required to inspection of the curve we might estimate 1, 700 eliminate the trend; Platt and Gallegos (1980) pre- W • m-2 = 1.5 / /3, so that /3 = 0.0021 mg C • (mg sented an equation with that property. The model Chl a)- ' •11- ' • W- ' • m2 would be a reasonable initial is parameterized in terms of the intensity lb' at which estimate. Objective fitting of the parameters using the photoinhibition reduces the value to half the maxi- modified Gauss-Newton routine gave Ps = 2.50, mum rate, and the intensity /I, such that 1 — = a = 0.042, and /3 = 0.0052 as final estimates. IT , the intensity at which photoinhibition begins. It is necessary to evaluate the fit of the models Since the revised model formulation contains two by examining the residuals, i.e. the difference bet- parameters that are characteristic intensities, there ween predicted and observed values. In doing so may be some value to substituting Ij for a, using we are examining both the validity of our assumption the identity given above; the range of intensities about the constancy of the variance, and the adequacy from 'k tO 17 can be considered an optimal "window" of the model formulation. We discussed the test of for photosynthesis (Platt and Gallegos 1980), which the error variance previously (Fig. 2). Testing may provide a useful way to compare the two samples. the adequacy of the model formulation involves plotting the residuals as a function of irradiance 075 a (Fig. 5). In Fig. 5a we see that positive and neg- 050 ative residuals occur at all regions of the curve, whereas in Fig. 5b there is a noticeable curvature 025 to the residuals, i.e. there is a tendency for residuals of a certain algebraic sign to be concentrated in 00 certain regions of the 1 axis. We determine if the • • (-) curvature is sufficient to warrant increasing the level cf, -Q25 of complexity in the formulation by testing if there is significant parabolic dependence in the residuals (-) as a function of irradiance. An alternative procedure 075 for analysis of Fig. 5 is the runs test (Platt and 0 Gallegos 1980), but we believe the parabolic cur- 075 vature analysis given here is a more powerful test. -0 The procedure is as follows. Fit a second-order 0.50 polynomial to the n residuals, with irradiance as D 025 the independent variable; form J2, the sum of squared :. • • residuals about the polynomial regression, and Jo , LL1 00 the original sum of squared residuals (Eq. 4). Further .• complexity is warranted if the statistic -0.25

JO J2 -050 F — ■ J2 /(n — 2) -075 , . ■ . ■ 100 200 300 400 500 600 700 is greater than the critical value of the F-distribution I RRADIANC E (W m-2 ) with 1 and n - 2 degrees of freedom (Bevington FIG. 5. Residuals about the fitted curves as a function of 1969). We recommend using the 99% significance irradiance for samples off the continental shelf of Nova level in making this decision, because estimating Scotia, (a) from 5 m (hyperbolic tangent equation) and an additional parameter should not be undertalcen (b) from 35 m (three-parameter photoinhibition equation without very strong indications. of Platt et al. 1980). 109 Also, the characteristic intensities are invariant the t value is (n - in), not the number of cases of under a change in vertical scaling, which is advan- pseudo-data generated in the Monte Carlo procedure tageous should it be decided at a later stage of the (Smith 1979). analysis to normalize photosynthesis to cell number, In comparing the two samples , we are comparing cell volume, etc. We therefore made this change models with different numbers of parameters. Both in parameterization for both experiments. model formulas use the hyperbolic tangent function Following the reformulation we again estimate to describe the curve up to the threshold of photo- the parameters and evaluate the fit as above. At inhibition, so that IC and lk are directly comparable. this iteration we found that the model formulations Observed differences between those parameters were were sufficient and thus we may accept the parameter statistically significant. The equation used for photo- estimates as final. The fitting routine gave P = inhibition includes a derived parameter for the 2.22, I,,. = 49.9, and R 5 = -0.19 for the near- threshold of photoinhibition which, for the deep surface sample; for the deep sample we obtained sample, fell within the 95% confidence interval P1-,1 = 1.59, 'k = 41.4, I = 522, and br = 312. 305 l. 320 W • in-2 . For the shallow sample The final fitted curves are shown next to the observed we can say only that the threshold is greater than data point in Fig. 4c-d. the highest light intensity used, 550 W • m -2 ; in a It is now appropriate to determine confidence one-tailed test of significance, the parameter IT for intervals for the parameters. We illustrated above the deep sample is significantly below that value. the procedure based on determining contours of Similar remarks hold for the intensity at which critical sums of squared residuals. This procedure P5 = 0.5 Pei . could be applied equally well to the present example. Interpretation of the results is now in order. However, the deep sample required four parameters We can see that isolation below the density gradient to obtain an adequate fit; if we were to compute in temperate waters results in a lower rate of light- confidence intervals by stepping through a grid of saturated photosYnthesis and lower irradiances re- 10 values for each parameter, we would have 104 quired both to saturate and inhibit photosynthesis. parameter combinations, and six 2-dimensional Complete interpretation in terms of adaptation to "slices" to examine. In a case like this it is much low light or response to physiological stress would more convenient to use the Monte Carlo approach. involve examination of species composition of the Examination of the residuals revealed no systematic samples, as well as a suite of other environmental mors in the model formulations, so that the Monte and biological covariates. Other questions of interest Carlo approach should not give seriously biased include the time required for the differences to de- confidence intervals. velop, but these sorts of queries are beyond the scope The Monte Carlo procedure involves simulating of our paper. data sets with residual errors distributed like those in the experimental data. The sum of squared residuals Physiological Content of the Models was 0.613 for the shallow sample, and 0.501 for Although the modeling of photosynthesis- the deep sample. There were 52 data points for each light curves is still stuck in the empirical stage, the curve, so that the standard deviations were 0.111 ultimate aim of the work is a mechanistic description. and 0.102, respectively. We used the estimated For the range of light intensities below the onset of parameter sets and measured irradiance values for photoinhibition (which will satisfy the interests of each curve to generate 100 sets of pseudo-data. Each most workers), Chalker (1980) provided a bridge data point was noise-corrupted by adding a normal between the empirical and the informed represen- random variable with zero mean and the estimated tation. Suppose that the rate of change of P and I standard deviation. Parameters for each of the 100 is a function of P,dP1d1 = f(P) and expand f(P) sets of pseudo-data were then estimated using the as a power series fitting routine. Confidence limits are computed from the mean dP and standard deviation obtained by averaging over (8) — = a 1 + a 1 P + a2P2 + a:1P 3 + . . . . dl the replicate estimates from the simulated data. The confidence limits for a given parameter, say 11„ Apply what is known about the photosynthesis- are given by light curve to determine the coefficients in this ex- pansion. For example in the linear approximation nE t y /2 Sp . — ni dP — = a + a 1 P (9) dl where S, the standard deviation of the parameter and 412 is from a table of Student's t-distribution. we know that at P = 0, dP Idl = a, the initial The number of degrees of freedom associated with slope, which can also be written as a = P„, Ilk.

110 (18) (10) ao — P,n hlk P = —a,12a2

We also know that as P —> Pm , dP1d1 = 0. In Eq. In Eq. (15) we find (9) we-then find a;P,„ dP 0 = — + a,P„, (19) Pm (—dl ) max = k '4(a, +

1 Now we know that the maximum slope of the photo- (11) . • a1 = — 1k synthesis—light curve is a = P,,, /'k This implies that a l = 0. Also we know that the maximum Then quantum yield, and therefore the maximum slope, occur at P = 0, which, from (18) implies that dP a — 0,120 2 = Oora, = 0. = dl Pm — P Thus, from (14), we can write

the condition 0 at I = 0 1 which leads, under P = ' (20) 0 2 = to the solution P7;i1k

Substituting this into Eq. (16) gives us the final result (12) P = P„, (1 — e 21 This is an equation previously published by Webb e Tk- — 1 (21) P = Pm et al. (1974). 21 The quadratic approximation of Eq. (8) is e-17 + 1 dP (13) = a 0 + a,P + a2P 2 which is identical to dl al before, we find the constant term to be of mag- (22) P = tanh ) As r m nitude a = Pm 11k . The condition that dP1d1 = 0 at P = Pm yields Thus the hyperbolic tangent equation is the only possible outcome from this analysis. Further, one 1 could say that the result a, = 0 is a natural con- (14) a , = — (- a2Pm) ik sequence of the fact that only two parameters, in this case a o and 0 2 , are required to fix a curve below In Eq. (13), this gives the onset of photoinhibition. The leading parameters a and Pm have the dP 1 following interpretation. The initial slope a is related (15) — = — (Pm — P) — a2P(Pm — P) dl lk directly to the maximum quantum yield (Rabinowitch and Govindjee 1969; Platt and Jassby 1976) and Integration of (15) yields is scaled by the magnitude of the quantum require- ment. The plateau height P„, (called the assimilation (r a2Pm)1 number Pg, when normalized to chlorophyll a) is e 1 — (16) P = P„, related to the processing time of the dark reactions - (Rabinowitch and Govindjee 1969; Falkowski 1980). Pm! ka2 e The processing time is scaled to the size of the photo- synthetic unit to yield the observed value of Pg: Now let y = dP1d1 and calculate 1 Chlo (23) P = — dy 2500 (17) .-= — — a + 2a2P dP lk where te is the handling time required for one sub- For the maximum slope dP1d1, we must have strate molecule (-0.O2 s), Chl o is the concentration dyldP = 0. Substituting (14) into (17), we see of photosynthetic pigment, and 2500 is a number that this occurs at derived from flashing light experiments, which is 111 also scaled to the quantum requirement (Rabino- mechanistic models will contain terms for which witch and Govindjee 1969). there is no operational technique for measurement Thus, when observing changes in Pm in the field, in the field. ,we may be seeing, for example, changes in the photosynthetic unit size. It is even possible that References through a direct normalization to Chia, such im- BARD, Y. 1974. Nonlinear parameter estimation. Aca- portant changes in size of the photosynthetic unit demic Press, New York, NY. 332 p. may be hidden, without parallel measurements on BEVINGTON, P. R. 1969. Data reduction and error analysis the reaction center chlorophyll P700. For a mecha- for the physical sciences. McGraw Hill, New York, nistic model, parameterization in terms of photo- NY. 336 p. synthetic unit size could perhaps be more inform- CHALKER, B. E. 1980. Modelling light saturation curves for photosynthesis: An exponential function. J. Theor. ative than chlorophyll a. However, it is important Biol. 84: 205-215. to be realistic about the practical, operational ob- DRAPER, N. R., AND H. SMITH. 1966: Applied regression servables and to recognize that P700 is not accessible analysis. John Wiley & Sons, Inc., New York, NY. to routine measurement in the field with the present 407 p. state-of-the-art. FALKOWSKI, P. 1980. Light-shade adaptation in marine phytoplankton, p. 99-119. In P. FaLkowski [ed.] Similarly, changes in Pm could be mediated Prirnary productivity of the sea. Plenum Press, New by changes in 'e' accessible perhaps through activity York, NY. assays on the photosynthetic enzymes, when a JASSBY, A. D., AND T. PLATT. 1976. Mathematical quantitative interpretation of such results is formu- formulation of the relationship between photosynthesis lated. and light for phytoplankton. Limnol. Oceanogr. 21: 540-547. In the scheme that we have suggested for the KOK, B., B. FORBUSH, AND M. McGLoiN. 1970. Co- numerical analysis of photosynthesis-light data, operation of charges in photosynthetic 02 evolution 1. the effect of other environmental variables is best A linear 4-step mechanism. Photochem. Photobiol. handled through their effect on the photosynthetic 11:457-475. parameters themselves (Platt et al. 1977). Thus, for PLATT, T., AND A. D. JASSBY. 1976. The relationship example, adaptation to a new light regime with an between photosynthesis and light for natural assem- adjustment of photosynthetic unit size would be blages of coastal marine phytoplankton. J. Phycol. 12: 421-430. treated as effect of light history on P. A similar PLATT, T., AND C. L. GALLEGOS. 1981. Modelling approach is used to handle the effect of temperature. . primary production, p. 339-362. In P. Falkowski [ed.] It is possible, however, that some of the subtlety Primary productivity in the sea. Plenum Press, New of these adjustments will not be made clear until a York, NY. mechanistic treatment is used which distinguishes PLATT, T., C. L. GALLEGOS, AND W. G. HARRISON. 1980. Photoinhibition of photosynthesis in natural , between the two photosystems. Working with fresh- assemblages of marine phytoplankton. J. Mar. Res. water field communities of phytoplankton, Vincent 38: 687-701. (1979) has observed directly the redistribution of PLATT, T., K. L. DENMAN, AND A. D. JASSBY. 1977. excitation energy (spill over) with adaptation to Modelling the productivity of phytoplankton, p. 807- various light levels within the water column. The 856. In E. D. Goldberg. [ed.] The sea: ideas and time-scale for this redistribution is of order minutes. observations on progress in the study of the seas. Kok et al. (1970) has achieved new insights into the vol. VI. John Wiley, New York, NY. dynamics of photosynthesis through a model in which RABINOWITCH, E., AND GOYINDJEE. 1969. Photosyn- thesis. Wiley, New York, NY. 273 p. photosystem II is represented as a linear series of reactions involving Sit- VERT, W. 1979. Practical curve fitting. Limnol. Ocea- four distinct time constants. nogr. 24: 767-773. To conclude, we might summarize by saying SMITH, W. R. 1979. Parameter estimation in nonlinear that a numerical analysis based on semiempirical models of biological systems. Fish. Mar. Tech. Rep. models can still be a useful vehicle for organizing 889. 90 p. data on photosynthesis-light relationships. It is a WEBB, W. L., M. NEWTON, AND D. STARR. 1974. Carbon dioxide exchange of Alnus rubra: a mathe- fully objective procedure for reducing many obser- matical model. Oecologia (Berl.) 17: 281-291. vations to a few parameters capable of physiological VINCENT, W. G. 1979. Mechanisms of rapid photosyn- interpretation. For a more complete understanding, thetic adaptation in natural phytoplankton commu- we need to keep moving towards a mechanistic nities. I. Redistribution of excitation energy between representation. It may turn out, however, that viable photosystems I and IT. J. Phycol. 15: 429-434.

112 Tracer Kinetic Analysis Applied to Problems in Marine Biology

DAVID F. SMITH AND S. M. J. HORNER CSIRO, Division of Fisheries and Oceanography, P.O. Box 20, Marmion, W.A. 6020 Australia

Introduction marine scientists from the theoretical basis of tracer kinetic analysis. Moreover, many of us would have To be conscious that you are ignorant of the facts been intimidated by the rigorous mathematical is a great step to knowledge." treatment demanded by the topic. Consider even an BENJAMIN DISRAELI explanatory footnote found in such a paper (Hearon 1804-81 1969). "Choose A 1 to be irreducible. Then by lemma 2, real root a 1 + b The proposition that marine bioscientists are there is a b such that A i + bl has a where a, is the maximum real root of A l . Further largely ignorant of the body of theory required to a positive eigenvector successfully employ radioisotopes is a proposition (Gantmacher 1959b), there is Z such that (A l + bl)Z = (a, + b)Z and clearly that marine bioscientists might, justifiably, greet a Z. Thus we choose a = and v = Z. with some scepticism. However, we suggest that A I Z = a1 such that the real this proposition is true and present a brief historical C can be chosen nonsingular and Then the resume of tracer kinetic analysis to support this parts of the roots of C are less than a. negative and contention and demonstrate that the present situation real parts of the roots of C - al are was almost inevitable. (Hearon 1963, 1968) the matrix - (C - al)-1 is Technological advances of the 1940s resulted nonnegative." in large quantities of certain isotopes being made The applications of tracer kinetic analysis were, generally available for research. In particular, the like the theoretical development, published in journals availability of "C, with its very long half-life, opened that would rarely be read by the community of marine up areas of research which were previously inacces- bioscientists. The most frequently encountered sible. Earlier studies employing "C as tracer had journals containing the results of applying tracer to be conducted at the site of production of the radio- kinetic analysis were J. Clin. Invest., J. Clin. Endo- isotope because of its very short (20.3 min) half- crinol. Metab., Circ. Res., Physiol. Rev., and Radiat. life. Thus, prior to the availability of "C, isotopic Res. carbon studies were as limited as those which today This historical association between tracer employ '"N. The availability of large quantities of kinetics and medical studies was to continue almost many new isotope species resulted in a rapid expan- to the exclusion of other research areas for several sion in employment of isotopes, which in turn stimu- years, broadening out to include pharmacokinetics lated interest in the theory of tracer kinetic analysis. and, finally, basic biological research. The burst of activity in development of a general One of the very earliest explicit references to analytic treatment of data arising from tracer exper- the works of the theoretical tracer kineticists to iments probably can be said to have begun with the appear in the literature of marine science was by publication by Sheppard and Householder (1951) Conover and Francis (1973). Thus, 17-20 yr after and continued with the simultaneous publications first appearing in press, the publications of some of Berman and Schoenfeld (1956), Berman et al. early theoreticians began to be cited in the literature (1962a, b), Hart (1955, 1957, 1958, 1960, 1965a, b, of marine science. 1966, 1967), Hart and Sondheimer (1970), Hart We have not mentioned the work of Steemann and Spencer (1976), Rescigno and Segre (1964), Nielsen (1952) as contributing to the general frame- Mann and Gurpide (1969a, b), and culminating in work of tracer kinetic theory; it was never intended the middle 1960s with the publications of Bergner to contribute to that body of knowledge. It should (1964, 1965). be clearly pointed out, however, that Steemann These publications, which collectively cover Nielsen's work is, in its entirety, consistent with almost in entirety the present body of theory relating tracer kinetic theory. It was necessary for Steemann to interpretation of tracer kinetic data, appeared in Nielsen to place certain restrictions on the experi- the following journals: Bull. Math. Biophys., J. mental systems employed to obtain the unambiguous Appl. Phys. , Biophys. J., and J. Theor. Biol. Thus, interpretations of data that he demonstrated were even the choice of journals tended to isolate most possible. Had subsequent workers employed his

113 of the Phytoplankton Dissohred inorganic Microheterotrophsmicrohetero microheterotrophs. Finally, consideration carbon DIC specific activity permits calculation of the moles A21 01 X41 02 .4- 04 carbon present in each compartment appearing in the X2 Xi Xa diagram. Al2 )114

Multicompartmental Analysis: >1/4,31 03 X43 Theoretical Basis X3 The analyses currently employed to treat data arising from tracer kinetic experiments fall into Dissolved organic carbon three broad categories. These differ in complexity and, in an inverse fashion, in the number of restric- tions that must be placed on the experimental system. The first of these is exemplified by the analysis model of "C fluxes in a water sample. If, in FIG. 1. A Schoenfeld (1956). This is two experiments, radioisotope is added first as DI' 'C to a described by Berman and the seawater sample being incubated in the light and the time- a differential equation approach and requires that varying compartmental radioactivities are measured, it is system be in steady state over the course of the generally possible to estimate all the intercompartmental experiment and that the tracer and traced substance transport rates. The incorporation of radioactivity into DOC be homogenously mixed. Each and every inter- from labeled DIC is light dependent and mediated by the compartmental transfer can be elucidated and the phytoplankton compartment. The DOC radioactivity, , associated rates measured solely by tracer experi- however, does not necessitate labeling phytoplankton ments only if each and every compartment is acces- carbon to obtain a radioactive precursor. sible to tracer addition and to sampling. This latter restriction assumes that any n -compartment system original protocol there would have been far fewer has the maximum number, n(n — 1), interconnec- erroneous statements and confusion about what had tions and that all transfer rates between compartments been measured in labeling experiments. are nonzero (Fig. 2). In practice, this restriction, Tracer kinetic analysis circumnavigates the i.e. accessibility of all compartments, is not nearly frequent confusion about what is being measured as severe as it might first seem. The number of in a radioisotope experiment. This is achieved, not accessible compartments required can be decreased by imposing experimental constraints, but rather by by incorporating independently gained information extending the analysis. about the system interconnections when analyzing Figure 1 diagrammatically presents the infor- the tracer kinetic data. This analysis is by far the most mation that can be obtained by applying a tracer frequently employed and the easiest to employ due kinetic analysis to the data arising from radioisotope to later efforts of Berman et al. (1962a, b), who incorporation experiments. produced the SAAM series of programs. These In this example a known quantity (in tracer programs are extremely efficient in parametric amounts) of radioactive inorganic carbon is intro- duced to a water sample which is then incubated under constant environmental conditions. Portions of the sample are removed with time and freed of \ 4 particulate matter by filtration. A portion of each 2 filtrate is assayed for radioactivity and concentration of total dissolved inorganic carbon (DIC). Another portion of each filtrate is freed of DIC before being assayed for radioisotope content. Finally, the radio- activity of each filter is measured. a. b. The tracer kinetic analysis of such experimental data permit one to estimate the rates of production FIG. 2. Generally connected n -compartment system. (a) of particulate organic carbon (FOC) without resorting The maximum number of interconnections between a system 1). (b) Multicompartmental to any physical separation of phytoplankton from of n compartments is WI: — analysis permits one to proceed from the generally connected microheterotrophs. The analysis yields a measure of model to a specific one such as the closed catenary system. the rate of DIC production through respiration during If every compartment is accessible to labeling and sampling, the course of the experiment, as well as the rates of no infomiation other than that obtained from tracer kinetic production of dissolved organic carbon (DOC) and experiments is needed to define and measure all the transfer rates of incorporation and respiration of DOC by rates.

114 estimation employing multiple nonlinear regressions, Furthermore, the SAAM 27 program now fitting experimental data directly to a compartmental includes the option of incorporating time-varying model or the defining differential equations, and can coefficients in the system differential equations bè obtained gratis as they were produced under the allowing one to solve for rate estimates in systems auspices of NIH. (Mathematical Research Branch, patently not in steady state during the experiment. National Institute of Arthritis and Metabolic Diseases, l'he second frequently encountered problem National Institutes of Health, Bethesda, MD 20014, of intimate mixing of tracer and "tracee" can be USA. Requests for SAAM 27 should be accompanied entirely circumnavigated by choice of experimental by a 7 track magnetic tape.) technique. For example, to obtain a set of time- The second method of tracer kinetic analysis varying samples, do not add tracer to a large seawater is essentially that proposed by Hart (1955) and re- sample and then remove aliquots with time. Instead, quires only that the experimental system be in steatly dispense a set of replicate samples, add to each the state and that tracer and traced substance are in same quantity of tracer then, noting the elapsed well-mixed compartments. The results obtained by time, filter the entire amount of the subsample. As applying this technique are always consistent with shown in Fig. 3 this technique obviates the need those obtained by the differential equation approach for homogeneity in regard to added tracer or even described above. In addition, a complete solution the need to ensure that the "replicates" are equal yielding all the transfer rates can be obtained if all in volume if the original water sample was homo- compartments are accessible to sampling and a genous at the time subsamples were removed. suitable number available for introduction of tracer. In the remainder of this section a summary of The increased generality is paid for by an increase the three general classes of tracer kinetic analysis in complexity of the procedure to obtain the solutions provides insight into their differences. A complete to the system equations that are integrodifferential in form. a. The third broad category of data analysis is that proposed by Bergner (1964). This .• • • approach, which ' • • • • is obtained from maximum likelihood theory, , requires a steady-state system but has relaxed the requirement r. of intimate mixing of tracer and "tracee." The latter advantage gives one a powerful tool .• .• with which to conduct a special class of tracer kinetic experiments, i.e. "tag-recapture" experiments. Tag- vi V2 V3 V4 recapture experiments, as most frequently conducted today, are merely tracer kinetic experiments as they b. were conducted before these three approaches to multicompartmental analysis were published. Before presenting the theoretical basis of each approach in detail it might be worthwhile to comment • on two of the shared constraints, i.e. the requirement of both a "steady state" and intimate mixing of tracer a and traced substance. In spite of the oft-heard statement, "Real world vi V2 V3 = V4 systems are probably never in a steady tate!" this constraint is probably only rarely violated in a primary FIG. 3. Invariance in radioisotope incorporation. A set of production estimate. The formal requirement, that the replicate subsamples is dispensed from a single homogenous system be in steady state during the tracer experiment, seawater sample and the same quantity of tracer Na1-1 14CO3 can be met by making the duration of incubation is added to each replicate. If the replicates are incubated only a few percent of the turnover time of the fastest under identical conditions for the same period of time, compartment that is not in steady state. Furthermore, then filtered, each filter will have precisely the same the compartmental analysis either furnishes or predi- quantity of radioactivity. Neither the volumes dispensed, cates a model which defines a set of equations. The nor the degree of mixing the tracer with the sample, affects equations will explicitly relate compartment sizes and the total radioisotope incorporation. Any deviation in sample rates; volume must result in equal but inverse deviations in both thus any deviation from steady state large the total NaHCO3 content and the total number of phyto- enough to compromise the tracer kinetic analysis plankton. The total radioisotope incorporation into POC can also be used to calculate the maximum and is inversely related to NaHCO 3 content, through specific minimum value of the rates measured during a period activity, and directly proportional to the number of phyto- of growth or depletion of a compartment. plankton. The two effects must always precisely cancel.

115 presentation of the theory is provided in the Refer- Thus, in a tracer kinetic experiment data points ences. are obtained that are measures of the different fdt); each data set is then fitted to a sum of exponential TRACER KINETIC ANALYSIS ternis. If n terms are required to fit the data we I: DIFFERENTIAL EQUATIONS APPROACH assume initially the system is composed of at least The following presentation is essentially that n compartments. If we were to lack all other infor- of Berman and Schoenfeld (1956) and Berman et al. mation about the system but could take samples (1962a). from all compartments during a tracer experiment There is an experimental system that can be we should still be able to obtain the n, A u's and described by a set of n communicating compart- the n, ai's and, therefore, estimate each and every ments which are accessible to sampling and to addi- Ku by the relations' tion of tracer. (Compartments may be chemical (CO., H2 CO3 HCO COEft- ) or physi- (4) X = A aA - ' cal as in the case of the Donnan equilibrium.) If the system is in steady state and tracer is injected into Typically we cannot access all compartments; one compartment, the time-varying distribution of however we usually can bring independent knowl- label throughout the system is given by the set of edge about the system to assist in the analysis. differential equations TRACER KINETIC ANALYSIS II: INTEGRODIFFERENTIAL EQUATION APPROACH ri:1011 [— XII + X12 + . • ± X11 [fat)] f2(t) X21 - X22 + • • • + X2H f2(t) The following is a description of the approach (1) which was investigated intensively by Hart (1955, 1957, 1958, 1960, 1965a, b, 1966, and 1967). The advantage of this technique lies in the fact that it [irm i „, .. • - supplies a formal way to manipulate tracer data to test assumptions of compartmental contiguity. As an example, in numerous published experiments the separation of phytoplankton and microheterotrophs where d[fe] lee) dt has been attempted with membrane filters and a judicious selection of pore size. If one does not fi (t) = time-varying radioactivity, or specific activi- employ a tracer kinetic analysis, this separation is ty, of i th compartment (dpm or dpm • mol') required if only phytoplanIcton production of particu- = fractional turnover rates from the jth to the ith late carbon is to be measured in the presence of compartment (min- ') primary DOC procIpction which, in turn, is incor- porated by microheterotrophs. 11 The integrodifferential equation analysis can Xii = — Xki i k manner described by i=o be applied to the data in the Hart as an alternative to physical separation of the components. and X o; = transport from the ith compartment to autotrophic and heterotrophic outside the system. By way of illustration consider the three-com- diagrammati- For constant X,j, the solutions of equations (1) partment system in Fig. 4. This figure are always sums of exponentials. cally represents the distribution of tracer with time in compartment 1 (DOC), compartment 2 (DIC), 11 and compartment 3 (POC). The POC compartment (2) Mt) = Aue - ait (i = 1, 2, . . . n) cannot conveniently be used to introduce label, but i= 1 its radioactivity can be measured after introducing tracer as DIC or DOC (Wiebe and Smith 1977a). The — aj are eigenvalues of the matrix [X u ], the A u are elements of the eigenvectors. In matrix no- tation . . 1 Frequently there are only rt(ii —1) independent A d's al 0 0 -- because, in an experiment, tracer can usually be introduced 0 a, . 0 into a single compartment; as the initial conditions regarding r radioisotope are known clearly only (n —1) independent (3) X =A A -1 A exist because

fi (0) = A U L0 • • • anJ i=i

116 These equations can be generalized to treat non contiguous compartmental systems by replacing each a,,f,(t) by f a (t — T)f,(7) ch- to give . . .

E f (t —7)f i (r) dr = 0 (6) ./.;(t) — j for i = 1 . . . n

By applying the Laplace transform , the integrodifferential equations (6) can be simplified to give a set of algebraic equations:

Letting (Mt)] ',(s) (r11 - (r1 +, r2 + r3 ) ) r11 Xi(s) and Ads) FIG. 4. Integrodifferential equation analysis of the single- compartment re-entry function. This analytical method is then extremely powerful when applied to systems wherein only some compartments are accessible to tracer addition and to —A + A 1 „(i X ,(s) sampling. In the experimental system illustrated, at t = 0, Yi(s) A 12(s) + • . + Y2 (S) 4 2 ,(s) — A 22 (s) + . . . + A 2„(s) X 2 (s) tracer NaH "CO3 is added to a water sample containing r Nal-I"CO:,. X, mol of Radioisotope is lost by incorporation (7) into nonrespired cell material and DOC. Radioisotope is returned to the DIC pool via respiration of DOC and of • • labeled phytoplankton and microheterotroph cell material. A ni (s) + A 2(s) + . . . — X „(s) Even though these sources have differing specific activities and time lags, their cumulative re-entry (by respiration) If we carry out a tracer experiment by adding and the size of the fixed NaHr2C0:1 pools can be obtained label to compartment 2 (DIC) we can solve for each by only monitoring the radioactivity of the DIC pool. X, quantity of DIC in the sample (mol); r,„ fractional loss X,(t) (or express them as X of each measured rate of DIC (min-1 ); r„ r, r11 , fractional respiration compartment and write the numerical form of equa- rates of carbon fixed as phytoplankton, microheterotrophs, tion (7) as and DOC (min - ').

(8) / X 2(s) A 2 (s) + [sX1.2(s) — XJ 2 (0)1 = 13 i = I We obtaM the rate of incorporation of radio- isotope due to all the POC trapped on a membrane If the Wronskians (see Appendix I) of equation filter. We then wish to know if the POC compartment (8) do not all vanish (Le. in the case of the POC is labeled solely through autotrophie DIC incorpo- compartment) there is no completely contiguous ration or also through rnicroheterotrophic POC, representation of the system; hence , in the POC which is not contiguous with DIC but receives its radioactivity we will find a contribution that did not label via the DOC compartment. The rationale of come directly from DIC. this analysis is as follows: If the equation for DOC radioactivity is exam- The POC and DOC compamnents receive label ined and the Wronskian {X 2 (s) , [sX 1 .,(s) — from phytoplankton at rates Oh(t) and Q,(t) which X 12 (0)]} = 0 the DIC and DOC compartments depend on the radioactivity of the DIC, Q,(t). The are contiguous and constant A u 's will be found. POC compartment also receives label in the form of microheterotrophs labeled Via the DOC compart- ment; therefore, the rate of labeling rnicrohetero- TRACER KINETIC ANALYSIS trophs, 0,(t), depends not on the DIC at time t, III: INDICATOR EQUIVALENCE THEOREM but on an earlier time (t —i-). The following is a description of the basis for If equation (1) of the last section is rewritten the tracer kinetic analysis described by Bergner to correspond to the nomenclature used by Hart we (1964). have Given a steady-state compartment containing b" moles 'of an element then the total input rate for that element, is related to the mean transit time (5) fi (t) — E a Ji (t) = 0 for i = 1 to n r, =1 for the element, 0° , by . . . 117 (9) b" = r 0" presents more than the most superficial treatmerit of theoretical tracer kinetics. Any individual interested If tracer is introduced to the system at t = 0 and in applying multicompartmental analysis should b(t) is the amount of tracer in the compartment at refer to the References as well as current literature. time 1, then Superficial though it might be, any discussion of a theoretical problem in physics necessitates the b(r) d employment of mathematical notation so compact (10) 0° = f 1),(0) t as to become cryptic to one unfamiliar with its routine use. where b, (0) is the total amount of tracer supplied So as not to entomb the utility of tracer kinetics, to the whole system at t = 0. alongside the experimentalist, inside its theoretical Letting the amount of tracer present in the whole framework, two examples of actual experiments, system be given by b(1), and the specific activity, data, and analysis will be presented. Much of the a(t), given by the relation . . . same material is covered as in the preceding section but numerical values in the appendices replace most (11) a(1) b(t)11," of the functional notation. In addition to quieting the fear of unfamiliar mathematical expressions then the quantity of the elernent in the whole system shared by many of us, this section explicitly states is some problems and pitfalls not easily derived from solely theoretical considerations. fo b„(t)dt Example 1 — Labeling compounds with pre- (12) b"s - cursor-successor relationships during a constant f a(t)dt infusion experiment — The curves of time-varying a radioactivities (Fig. 5) were obtained by introducing tracer as an iodide-131 solution to a suspension of MULT1COMPARTMENTAL ANALYSIS: human thyroid cells. At varying times, portions of PRACTICAL APPLICATIONS the cell suspension were removed and the compounds It would be foolhardy to believe that the pre- of interest isolated (for experimental details see ceding section is more than the barest outline, or Smith and Holmes 1970). This particular experi- 150,-

(t) 02 (t)

• 03(t)

100 a. • 00) o • 05(t) • 06(t)

to" 50 a 8

r- 0 100

Time. (min) • Fio. 5. Radioactivities of thyroxine pathway members following introduction of 131 1 1- to a thyroid cell suspension. After addition of radioiodide, portions of the cell suspension were removed at noted times and frozen in liquid nitrogen. - The organic compounds of interest were extracted from each .sample, then isolated from one another by chromatography. The radioactivity of each spot was measured and plotted versus satnpling tune. Analysis of these curves permits one to

determine pool sizes, forward and reverse reaction rates, and precursor - product relationships. 118 mental system is one of special interest as it is an exact analogue of the one encountered in primary productivity experiments. One compartment, iodide or bicarbonate, is present in an enormous quantity relative to any other compartment or turnover rate of the system. Furthermore, it is into this compartment that label is introduced to the experimental system. 0.1 In such a system the labeled compartment approxi- mates a constant quantity of tracer present at constant specific activity during the course of the experiment. Compounds being formed from such a reactant will increase in total radioactivity until their specific activities equal that of the source of tracer. There- 0.01 after, if the system of compartments remains in steady state in regard to the moles of compounds present, the quantity of radioactivity in each com- partment also remains constant. Curves of time-varyhig radioactivities arising 0.001 from such an experimental system are, as noted 0 BO 160 240 320 earlier, graphs of sums of exponential terms. These Time (min) curves, however, contain a term having a unique Fic. 6. Radioactivity curves obtained by transforming the property. The exponential coefficient of this term experimental data. The data from experiment 1 were trans- is zero; hence the term is a constant. The origin of formed by the relation yi = 1 — Q,(t)IQ,(r—>oc). This this constant term lies in the introduction of tracer transformation plus data point error *preclude the arbitrary via the extraordinarily large compartment of the selection of the smallest values to obtain the initial X, system. estimates. The region of each curve which best approximates a straight line segment is used to obtain an initial estimate The classic approach to the analysis of data of the smallest X in the sum of exponentials. as in Fig. 5 begins with finding the number of exponential terms present in each curve and obtaining the exponent of the term that dies away most slowly initial estimates of the exponents. To do this we in time. first estimate the radioactivity maximum of each After obtaining an initial estimate of the first curve (i.e. the plateau each curve approaches as a exponent, X, , the value of the coefficient A, is lirniting value). obtained by substitution in the relation: Next, divide each data point of a curve by the radioactivity maximum. Alter dividing, change the (14) A, = R(01 e-A, `k sign of the quotient and add 1. This new quantity R(t) is then plotted against time on semilogarithmic The values of R(t) and t employed are those graph paper. used in obtaining X , and an average value of A l In theory we have eliminated the constant; the is determined. plot of R(t) may be curvilinear near the origin, but Once A, and X , are obtained, /1 . 1 e -À.` can be at larger values of time it becomes straight as it calculated for each sample time and the difference contains contributions from only the term with the taken between A, and R(t) to give R'(t). The smallest exponent. In theory, we use the R(t) values function R '(t) should lack A and its graph taken at the largest values of time to find the constant can be employed to find A 2 e -)` 21 as the new terrn slope of the R(t) plot by the relation having the smallest exponent. This process is repeated until each exponential term present in the (13) X = In [R(t2)/R(t1)]/(t 1 —t 2 ) original sum has been estimated. The decision as to whether or not to stop trying to extract additional Figure 6 shows why this is rarely a good practice. exponential terms can be based on statistical con- The values at the largest , values of time are the straints imposed by the data, on a priori knowledge smallest quotients after division by the maximum about the system, and most importantly, on the radioactivity. Any error in.either the estimation of experience of the researçher. numerator or denominator accentuates the scatter The detailed steps of this portion of the analysis in the region of the linear portion of the graph. are given in Appendix II and should be consulted The experimentalist has to judge the region of before proceeding further. the graph that will yield the best approximation for Having obtained initial estimates of the expo- the smallest exponent, i.e. the first to be found, nents of each data set, the data sets are independently

119

fitted by a program employing multiple nonlinear and : . . regression and the values of the exponents obtained by an iterative process. [3.0454 0 Exponents similar to each other in numerical (22) ). =-- 0 .0523 0.0552 0 value are tested to determine if they represent the 0.0308 0 0.0352 same exponent appearing in different curves. This is accomplished by fitting all the data sets simul- Thus, compartrnent E receives no tracer from B taneously and causing a X 1 of one tracer curve to or F, compartment F receives tracer only from E, be equal to the X, of another tracer curve. The final and compartment B receives tracer only from E. sums of squares for each data set should approach The fractional rates of transfer of material are a minimum if the X, is common. given by each of the elements of X and the absolute Using the results obtained from the fitting of values can be found from the total radioisotope compartments E, F, and B in Appendix II, we can content of each compartment divided by the specific continue the analysis from the stage at which the activity of the iodide supplied as tracer. tracer curves are expressed as sums of exponentials. Example 2 - Estimating rates of DOC pro- fitting provided equations . . . The global the duction by phytoplankton - Attempts to employ DI ' 4 C incorporation to measure the rates of DOC (15) QE(t) = 134.51 (1 - 1.003e -°451' ) production by phytoplankton can only succeed if tracer kinetic experiments are used. Data from two ( 1 ) Qr(t) = 129.52 (1-5.803e -0- 04541 such experiments (Table 1) plotted against time give + 4.801 e-"."55"r ) curves similar to that of Q(t) seen in the previous example. (1 7) Q,(1) = 113.30 (1-1.524e -4"e-1' Curves of this shape invariably result when + 0.524e -".0352' ) -...J.eawater samples are incubated in the light and tracer is introduced as DI"C (Mague et al. 1980; The matrix A, of equation (4) is then . . Wiebe and Smith 1977a) or the complementary curves are seen during incubations in which tracer .0 0 0 is introduced as labeled DOC (Wiebe and Smith (18) A = 1 5.803 -4.801 1977a, b). The shape of the DOC radioactivity 1.524 0 -0.524 curves as well as rate of production are apparently invariant, even under experimental conditions pro- The inverse of A which is A -1 is calculated to be . . . ducing dramatic changes in the POC production rate (Smith and Wiebe 1976). 1 0 0

(19) A -1 = [1.2 08 -0.2083 0 2.907 0 - 1.909 TABLE 1. Time-varying radioactivities obtained during light incubatibns of seawater samples with D1 14C. The matrix, a:, is taken from equations (15)-(17). . .

).0454 0 Experiment 1 Experiment 2 (20) « 0 .0.0553 0 0 0.0352 [ Time DO HC Tirne DO"C The intercompartmental transfer rates will be content content found by the relation given in equation (4). (min) (Bq) (min) (Bq)

(21) X = A cc A -1 = 0 0 0 1 0 0 25 154.36 1.28 1.190 -257.38 [5.803 50 2.78 2.144 -4.801 0 100 367.70 5.83 4.286 1.524 0 , -0.524 150 419.25 6.39 6.430 [1.0454 0 200 443.33 13.33 9.050 0 0.0553 0 ] 250 463.26 19.72 10.954 - 23.33 _12.859 0 0 0.0352 26.39 13.098 I 0 0 ] 29.06 13.336 [1. 208 - 0.2083 0 35.28 13.003 2.907 0 1.909 38.89 14.288

120 Isotope incorporation experiments that employ imposed on this responsibility is avoiding the intro- a control sample and a single experimental sample duction of systematic errors by dispensing subsamples processed after some arbitrary incubation period from a nonhomogenous sample, or by altering the cannot estimate rates of DOC production. sample through mechanical stirring. Subsequently, Such an approach assumes that a constant rate all that is required is that each and every replicate of DOC production will be reflected by a constant has exactly the same history throughout the experi- rate of radioisotope accumulation in the DOC com- ment. The degree of culpability regarding the first partment. This is demonstrably not true and is an of these three pitfalls may never be resolved. The error that originates from failure to distinguish latter two can be almost entirely eliminated by between radioisotope flux and carbon flux. methodically ensuring that replicates receive identical If the DOC compartment is in steady state with treatment. respect to carbon, then the DOC compartment is of Some idea of efficiency of this approach can be constant size relative to carbon, i.e. the rate of DOC gleaned from the performance of a group of cell phys- production is exactly equal to its rate of loss. iologists working under Professor F. C. Neidhardt If a constant supply of tracer is introduced to at the University of Michigan, USA. A majority the input of such a compartment, then the compart- of the problems in which they were interested required mental radioactivity will increase at a continuously precise estimates of the intrinsic growth rates of decreasing rate, until the specific activity of the bacterial cultures. Their procedure was to remove compartment is equal to that of its inputs. Thereafter, an aliquot of a bacterial suspension, growing at the compartment will be at steady state with respect approximately 37°C, and transfer the live cell sus- to both isotope and carbon content. pension to a spectrophotometer cuvette. One then Thus, isotope incubations employing only a walked a distance of perhaps 10 m to the spectro- single measurement in time must give smaller and photometer, room temperature------20°C, and read smaller production rate estimates with longer and the absorbancy of the cell suspension. The growth longer incubation periods. This is an artifact arising rate of the culture was estimated from the slope of from division of a constant quantity of radioisotope, the semilogarithmic plot of absorbancy versus time. which is equal to the DOC compartment size divided In the course of a year, several thousands of such by the DIC radioactivity, by an incubation period growth rate estimates were obtained, yet the entire of arbitrarily long duration. range of variation of growth rate estimates made The analysis of the data of Table 1 shows DOC under the same culture conditions was less than isotope incorporation curves to be described by the 0.8%. Similar results can be obtained with less care relation . • . in tracer kinetic experiments if a rigid experimental protocol has been devised. Q(t) = Some specific suggestions and techniques to assist in tracer kinetic experiments will be given, The presence of only a single exponential can only but we will attempt to avoid giving any arbitrary mean that the DIC and DOC compartments are values for universal employment. That is, we know contiguous. This does not imply that DI' 4C does of few experimental techniques that can be transferred not have to enter the cell before being converted to from one experimental system to another without DO ' 4C, but it does demonstrate that label in DO' IC careful modification. Therefore, we would be loathe does not arise from labeled POC material. to recommend any specific values as "safe" pressure differentials and sample volumes which one might "Nobody's perfect! Even I once made a mistake! universally employ without testing. Instead, we . . I thought I'd made an error, and I hadn't." would strongly recommend that each filtration ANON. manifold be equipped with an automatic pressure relief valve. The valve is adjusted to give the maxi- mum pressure differential with all manifold ports MULTICOMPARTMENTAL ANALYSIS: closed. During all subsequent sample filtrations, EXPERIMENTAL TECHNIQUE although the number of ports open to vacuum flasks will vary from 10 to 1, the pressure differential Tracer kinetic experiments and the subsequent will never rise above the maximum for which the data analysis allow only one independent variable, valve has been set. time. Therein lies both the analytical power of the technique and the demand for rigor in the experimen- Purification of DI" C — Commercially avail- tal portion of the work. able preparations of BaNCO 3 can have as much as The investigator is responsible for ensuring that 11% of the total radioactivity present in a form that measurements on a given water sample reflect prop- is nonvolatile even from highly acid, boiling solu- erties of that sample and are not artifacts. Super- tions. Moreover, the energy spectrum obtained from

121 such a solution is complex and certainly is not the interpreted as uptake by phytoplankton as opposed spectrum of NC. to microheterotrophs. Any tracer experiment that has not employed The truth is such plots occur if the radiochemical a DIHC preparation previously isolated from the has not been purified just prior to use, if a back- commercially supplied material must be considered ground obtained by accumulating less counts than suspect. Any measure of DOC production rate must the samples is subtracted, or if the samples have not be made with radioisotope that was converted to a been counted to the same number of counts. gas by addition of a nonvolatile acid to an aqueous It cannot be too strongly emphasized that if the solution. Only in this way can one hope to quantita- experimental data, y, has an associated uncertainty, tively remove DI' 'C at the end of an incubation. a-, and one fits the transformed data f(y) the un- certainties must be likewise compensated to give Employment of carrier - Quantitative removal Œf , by the relation, of a radioisotopically labeled element from an ex- perimental system generally demands scrubbing the di()') 0-f = 0- system with the unlabeled element in carrier form. dy To be a carrier for a tracer only the isotopes can be different. In a radiotracer experiment the only carrier There is a desperate need for reviewers and editors for H"C0:1 - is H' 3C01 - or H 12 . There is no to demand that manuscripts describing results that a priori means of guaranteeing that adding a non- depend on curve fitting and parameter estimations volatile acid to a water sample, then sparging with include in the methods reported, the techniques and Ar, or air will quantitatively dilute out the convergence criteria of the numerical analyses. DI'C present to background levels.

Estimation of sample radioactivity — It has already been mentioned that the only permissible variation between the samples originating from a References tracer kinetic experiment is the length of the incu- bation period. Not all references have been cited in the text. It follows then, that the precision associated with each radioactivity measurement should be the BERGNER, P-E. E. 1964. Tracer dynamics and the deter- same. To ensure this, each sample must be counted mination of pool-sizes and turnover factors in metabolic until the saine number of counts has been accumu- systems. J. Theor. Biol. 6: 137-158. 1965. Exchangeable mass: determination without if one counts lated. Systematic error is guaranteed assumption of isotopic equilibrium. Science (Wash- samples for the same length of time instead of to ington, D.C.) 150: 1048-1050. the same number of counts. BERMAN, M., AND R. SCHOENFELD. 1956. Invariants in experimental data on linear kinetics and the formu- Transformation of data — Having paid strict lation of models. Appl. Phys. 27: 1361-1370. attention to each portion of the experimental proce- BERMAN, M., E. SHAHN, AND M. F. WEISS. 1962a. dure and taken care to estimate properly sample The routine fitting of kinetic data to models: a mathe- radioactivity, it is still possible to snatch defeat matical formalism for digital computers. Biophys. from the jaws of victory. J. 2: 275-287. As innocuous a procedure as a linear regression BERMAN, M., M. F. WEISS, AND E. SHAHN. 1962b. can sow the seeds of despair if transforined data Some formal approaches to the analysis of kinetic terms of linear compartmental systems. Bio- points are employed and the ramifications of the data in phys. J. 2: 289-316. transformation are unappreciated. CONOVER, R. J., AND V. FRANCIS. 1973. The use of Heterotrophic potential experiments afford us radioactive isotopes to measure the transfer of materials with an excellent example. In these experiments, in aquatic food chains. Mar. Biol. 18: 272-283. water samples are incubated with a labeled organic ESTREICHER, J., C. REVILLARD, AND J-R. SCHERRER. compound and the rate of incorporation estimated 1978. Compartmental analysis-I: Linde, a program at varying concentrations of that compound. The using an analytical method of integration with con- incorporation rate is assumed to follow Michaelis- stituent matrices. Comput. Biol. Med. 9: 49-65. Menten kinetics and frequently the data are trans- FLECK, G. M. 1972. On the generality of first-order rates isotopic tracer kinetics. J. Theor. Biol. 34: 509- formed to obtain a straight line for purposes of in 514 . parameter estimation. Numerous examples of such GURPIDE, E., J. MANN, AND S. LIEBERMAN. 1965. transformed data that consist of not one but two Estimation of secretary rates of hormones from the straight line segments have been reported, the seg- specific activities of metabolites which have multiple ment at higher substrate concentrations having the secreted precursors. Bull. Math. Biophys. 27: 389- lesser slope. Unfortunately, this has been widely 406.

122 HALL, S. E. H., R. GOEBEL, I. BARNES, G. HETENYI JR., PERL, W. 1960. A method for curve-fitting by exponential AND M. BERMAN. 1977. The turnover and conversion functions. Int. Appl. Radiat. 8: 212-222. to glucose of alanine in newborn and grown dogs. PERL, W., R. M. EFFROS, AND F. P. CHINARD. 1969. Fed. Proc. Fed. Am. Soc. Exp. Biol. 36(2): 239-244. Indicator equivalence theorem for input rates and HART, H. E. 1955. Analysis of tracer experiments in regional masses in multi-inlet steady-state systems non-conservative steady-state systems. Bull. Math. with partially labeled input. J. Theor. Biol. 25: 297- Biophys. 17: 87-94. 316 . 1957. Analysis of tracer experiments: II. Non- PROVENCHER, S. W. 1976a. A Fourier method for the conservative non-steady-state systems. Bull. Biophys. analysis of exponential decay curves. Biophys. J. 16: 19: 61-72. 27-41. 1958. Analysis of tracer experiments: III. 1976b. An eigenfunction expansion method for Homeostatic mechanisms of fluid flow systems. Bull. the analysis of exponential decay curves. Chem. Phys. Math. Biophys. 20: 281-287. 64: 2772-2777. 1960. Analysis of tracer experiments: IV. The RESCIGNO, A., AND G. SEGRE. 1964. On some topological kinetics of general N compartment systems. Bull. properties of the systems of compartments. Bull. Math. Biophys. 22: 41-52. Math. Biophys. 26: 31-38. 1965a. Analysis of tracer experiments: V. Integral SHEPPARD, C. W., AND A. S. HOUSEHOLDER. 1951. equations of perturbation-tracer analysis. Bull. Math. The mathematical basis of the interpretation of tracer Biophys. 27: 417-429. experiments in closed steady-state systems. J. Appl. 1965b. Analysis of tracer experiments: VI. Phys. 22: 510-520. Determination of partitioned initial entry functions. SMITH, D. F. 1974a. Quantitative analysis of the func- Bull. Math. Biophys. 27: 329-332. tional relationships existing between ecosystem 1966. Analysis of tracer experiments: VII. General components. I. Analysis of the linear intercomponent multicompartment systems imbedded in non-homo- mass transfers. Oecologia 16: 97-106. geneous inaccessible media. Bull. Math. Biophys. 1974b. Quantitative analysis of the functional 28: 261-282. relationships existing between ecosystem components. 1967. Analysis of tracer experiments: VIII. II. Analysis of the nonlinear relationships. Oecologia Integrodifferential equation treatment of partly acces- 16: 107-117. sible, partly injectable multicompartment systems. 1975. Quantitative analysis of the functional Bull. Math. Biophys. 29: 319-333. relationships existing between ecosystem components. HART, H. E., AND J. H. SONDHEIMER. 1970. Discrete III. Analysis of ecosystem stability. Oecologia 21: formulation and error minimization in applying the 17-29. integro-differential equation approach to mono- 1977. Primary productivities of two foraminifera: compartment data. Comput. Biol. Med. 1: 59-74. zooxanthellae symbionts. Proc. Third Int. Symp. HART, H. E., AND H. SPENCER. 1976. Vascular and Coral Reefs. p. 593-597. extravascular calcium interchange in man determined Smn-H, D. F., AND R. A. HOLMES. 1970. Kinetics of with radioactive calcium. Radiat. Res. 67: 149-161. allosteric inhibition in vivo: a quantitative analysis HEARON, J. Z. 1969. Interpretation of tracer data. Biophys. with synchronous cultures of B lastocladiella emersonii J. 9: 1363-1370. J. Bacteriol. 104: 1223-1229. HETENYL, G. JR., AND K. H. NORWICH. 1974. Validity of the rates of production and utilization of metabolites SMITH, D. F., AND W. J. WIEBE. 1976. Constant release of photosynthate from marine phytoplankton. Appl. as determined by tracer methods in intact animals. Environ. Microbiol. 32: 75-79. Fed. Proc. Fed. Am. Soc. Exp. Biol, 33: 1841-1848. 1977. Rates of carbon fixation, organic carbon JACQUEZ, J. A. 1970. A global strategy for nonlinear release and translocation in a reef-building foraminifer, least squares. Math. Biosci. 7: 1-8. Marginopora vertebralis. Aust. J. Mar. Freshw. Res. JENNRICH, R. I. 1979. Fitting nonlinear models to data. 28: 311-319. Annu. Rev. Biophys. Bioeng. 8: 195-238. LEVY, G., M. GIBALID, AND W. J. JUSKO. 1969. Multi- STEEMANN NIELSEN, E. 1952. The use of radioactive compartment pharmacokinetic models and pharmaco- carbon ("C) for measuring organic production in the logic effects. Pharm. Sci. 58: 422-424. sea. J. Cons. Perm. Int. Explor. Mer. 18: 117-140. MAGUE, T. H., E. FRIBERG, D. J. HUGHES, AND I. THAKUR, A. K. 1972. On the stochastic theory of com- MORRIS. 1980. Extracellular release of carbon by partments: I. A single-compartment system. Bull. marine phytoplankton; a physiological approach. Math. Biophys. 34: 53-63. Limnol. Oceanogr. 25: 262-279. 1973. On the stochastic theory of compartments: MANN, J., AND E. GURPIDE. 1969a. Interpretation of II. Multi-compartment systems. Bull. Math. Biol. tracer data: significance of the number of terms in 35: 263-271. specific activity functions. Biophys. J. 9: 810-821. THAKUR, A. K., AND A. RESCIGNO. 1978. On the sto- 1969b. Interpretation of tracer data: some factors chastic theory of compartments: III. General time- which reduce the number of terms in the specific dependent reversible systems. Bull. Math. Biol. 40: activity functions in n-pool systems. Bull. Math. 237-246. Biophys. 31: 473-486. THOMASSON, W. M., AND J. W. CLARK JR. 1974. MYHILL, J. 1968. Some effects of data error in the analysis Analysis of exponential decay curves: a three-step of radiotracer data. Acta Radiol. Ther. Phys. Biol. scheme for computing exponents. Math. Biosci. 22: 7: 443-452. 179-195. 123 WIEBE, W. J., AND D. F. SMITH. I977a. Direct meas- The matrix to be assembled from the above equation is: urements of dissolved organic carbon release by phytoplankton and incorporation by microheterotrophs. sin(t) cos(t) cos(t +77131 Mar. Biol. 42: 213-223. A = cos(t) - sin(t) -sin(t +7T/3) I977b. "C-labelling of the compounds excreted sin(t) -cos(t) -cos(t + TT/3 ) by phytoplankton for employment as a realistic tracer in secondary productivity measurements. Microb. det A = [(sin(t) • sin(t) • cos(t + 11/3)) Ecol. 4: 1-8. + (cos(t)• sin(t + 1T/3 ) • sin(t)) - (cos(t + Trh) • cos(t) • cos(t))] - [(sin(t) • sin(t) • cos(t + 7T/3 )) - (cos(t) • cos(t) • cos(t + 7T/ 3 )) + (sin(t) cos(t) • sin(t + 111,1 ))] Appendix I Simplifying, After injecting label into a single compartment we det A = W - obtain tracer kinetic curves and these are expressed as [sin 2(r) • cos(t + 77/3 ) + sin(t) • cos(t) • sin(t + 11:1) sums of exponentials, Q ,(t), Q 2 (t) . . . Q, (t). - cos 2(t) • cos(t +71j:) - sin 2(t) • cos(t + eh) Take the first, second, and finally the (n - 1)th de- + cos 2(t) • cos(t + 7T/3 ) - sin(t) • cos(t) • sin(t + rivative of each Q(t) and order them as in the matrix . . . eh)] and co O.

[ Q t(t) Q 2 (t) d(Q 1 (t)) dQ 2 (t) dt di

d" - '(Q 1 (t)) d" -1 (Q2(1)) Appendix II d' 't d" -' t arising from tracer kinetic experi- Q Q n(t) Three sets of data ments have been included in this appendix to supply realistic d'21(t) d (Q n (t)) numerical values and associated errors for practice in peeling dt dt exponential ternis. One must be warned, however, that each of the three data sets contains a trap for the unwary. Some compart- mental radioactivity curves are described by a sum of dn-I (Qi(0) d" -1 (Q)1 (0) exponentials containing two terms whose exponents differ d" - 't .1 by less than a factor of two. Classic curve peeling tech- niques are not applicable in cases where the exponential The Wronskian, co, we seek is simply the number obtained sum contains a X, and a X9 such that by finding . . .

det A = co X2/X, < 2

As an example, suppose Y = sin(t), Y2 (t) = cos(t), Y 3(t) = cos (t + TT/3 ). Because we have three If a curve peeling procedure is to be done routinely functions we will need the first and second derivatives to one would, of course, write a simple program for calculator evaluate the Wronskian. or computer. None the less, the exercise of graphical The first derivatives are: analysis often provides insight into the experimental results that is impossible to obtain any other way. dl', d Y2 To illustrate the procedure by which we obtain initial = COS(i) = sin(t) dl dt estimates of the exponents present in the sum of expo- nentials we will employ the data of experiment 1 at the dY3 end of this appendix. — = - sin(t + 111) dt Inspection of the data from experiment 1 or the graphs in Fig. 5 indicate that tracer accumulates most rapidly in The second derivatives are:re: compartment E. The graph shows no inflection point and no discernible lag, which suggests that compartment E is D'Y, d2 y2 contiguous with the source of tracer supply, the iodide = - sin(t) = - cos(t) dt2 dt2 compartment. To test this assumption and to initiate the analysis we obtain an estimate of the constant term by d2 Y3 averaging the last four values of Q E (t) to obtain Q = = - COS(t el2) dt2 133.574.

124

EX PERI MENT 1. Tracer kinetid data:

Compartment A Compartment B Compartment C

Time Total radioactivity Time Total radioactivity Time Total radioactivity (min) (M cpm) (min) (M cpm) (min) (M cpm)

5.13 0.001 3485 0.7 0.000 338 4.85 0.007 556 8.57 0.003 166 2.47 0.014 98 8.30 0.068 02 13.37 0.060 588 4.64 0.12799 13.12 0.420 88 17.05 0.192 615 8.00 0.685 55 16.77 1.074 4 21.13 0.498 598 12.84 2.689 9 20.85 2.189 8 31.45 2.539 46 16.50 4.865 8 31.22 7.785 0 40.75 6.689 56 20.59 8.616 9 40.50 15.920 7 61.52 22.731 14 30.95 22Al2 61.23 38.5566 85.18 44.360 24 40.25 35.781 84.90 63.003 2 128.38 75.878 87 60.94 66.075 ' 128.1 88.326 9 166.89 88.524 24 84.62 .87.162 166.59 94.156 8 239.02 88.838 50 127.5 106.075 238.72 99.025 1 285.75 89.901 68 166.28 112.585 285.27 • 98.967 9 238.42 110.889 284.90 113.314

Compartment D Compartinent E Compartment F

Time Total radioactivity Time Total radioactivity Tirne Total radioactivity (min) (M cpm) (min) (M cpm) (min) (M clxn)

13.62 0.007 973 0.15 0.632 07 0.43 0.010 9 17.32 0.031 080 1.95 10.910 96 2.22 " 0.479 4 21.42 0.107 804 4.10 22.723 4 4.40 2.089 31.72 0.948 847 7.47 37.475 3 7.75 6.005 41.03 2.529 95 12.22 56.848 4 12.52 14.06 61.82 11.765 41 15.90 69.392 7 16.20 21.75 85.48 30.951 18 20.03 79.213 3 20.30 30.92 128.70 61.866 36 30.38 101.414 52.97 167.17 75.124 19 39.69 110.747 40.00 71.70 239.35 80.905 48 60.37 126.558 60.38 99.71 286.11 - 82.893 01 84.05 133.661 84.37 114.4 126.79 134.739 127.18 127.3 165.64 134.197 165.97 129.1 237.75 132.420 238.12 130.2 284.2 132.940 284.57 129.3

125

ËXPERIM.ENT 2. Tracer kinetic data.

Compartment A Compartment B tomPartment C. • Time Specific activity Time , Specific activity Time Specific activity (min) (cpm/gmol) (min) (cpm /p..mol) (min) (cpm/iLmol)

«0.12 3 595.52 0.43 97.90 0.77 1.94 96 626 2.27 4958.18 2.53 3.89 174 994 4.13 16 964.18 4.43 6.87 293 934 7.20 47 298 7.53 166.84 11.08 421 890 11.35 103 878 11.60 1 004.12 15.00 -- 555 810 15.32 178 846 15.62 3 117.18 19.10 629 634 19.38 247 934 19.72 7686.14 29.12 778 110 29.43 453 066 29.69 30 356 39.23 , 891 906 39.50 596 150 39.80 75 310 59.15 1004 978 59.42 836 998 59.70 222 758 80.00 1041 374 80.32 984 638 80.64 437 666 122.27 1055 662 122.60 1076 218 122.92 781 166 164.33 1074 026 164.60 1053 406 164.90 962 554 234.30 1064 774 234.65 1079 570 234.97 1064 478 273.95 1086 286 274.32 1036 828 274.67 1082 842

Compartment D Compartment E Compartment F

Tüne Specific activity Time Specific activity Time Specific activity (min) '• • (épm/gmol). (cpin/ p.mol) (min) (cPM//),

2.89 150.26 3.17 100.02 4.70 667.82 5.00 778.78 7.87 3 522.82 8.18 3619.06 8.53 237.88 11.87 12 691.80 12.15 12 384 12.47 1136.32 15.89 29 474 16.22 28 802 16.52 3402.16 20.00 ' 52 870 20.33 51 850 20.62 8 057.78 29.95 143 142 30.20 135 518 30.48 29 710 40.08 .„ 255 176 40.35 236 998 40.65 74 946 60.02 493 322 60.28 • 476 462 60.62 225 162 80.92 721 118 81.22 708 314 81.50 413 894 123.22 930 866 123.50 915 430 ' 123.82 751 106 165.20 1045 386 165.53 1050 050 165.90 « 950 998 235.27 1084 226 235.59 1085 798 235.92 1056 738 275.02 1123 734 275.30 1071 978 275.64 1071 754

126-

EXPERIMENT 3. Tracer kinetic data.

Compartment A 'Compartment B Compartment C

• Time' Specific activity Time Specific activity Tinte Specific activity (min) (cpm/gmol) (min) (cpm/grnol) (mm) (cpm/gmol)

6.25 636 1.27 22 438 9.58 2 510 2.50 47 748 14.47 1 440 14.73 8 894 6.57 126 176 19.65 4 744 19.95 20 572 9.92 182 352 \ 30.68 21 332 31.05 64 048 . 15.03 279 424 43.45 63 036 43.82 134 296 20.25 330 516 65.90 181 232 66.22 289 732 31.33 462 800 86.38 - 306 848 86.75 422 892 44.18 571 156 126.83 524 020 127.17 617 748 66.52 735 764 167.33 701 940 167.68 755 756 • 87.13 766 184 205,07 788 804 205.47 803 452 127.55 ' 839 976 233.92 799 900 . 234.28 849 048 168.03 832 372 205.80 871 476 ' 234.65 872 304

Compartment D Compartment E Compartment F

Time Spedific activity Time SpeCific activity Time Specific activity (min) (cPni/P-mol) (min) (Çpmairhol) (min) (cpm/gmol)

0.90 14 898 2.83 1 424 1.87 562 2.15 40 992 6.88 10 328 3.12 1 816- 3.42 65 740 10.25 22 894 7.23 10 680 7.58 .147 616 15.33 49 908 10.57 23 ro 10.92 . 197 552 20.62 82 256 15.63 47 540 15.98 282 228 31.62 166 380 20.92 81 052 21.27 349 432 44.50 , 263 756 31.92 164 768 32.32 463 612 66.87 ' 442 804 44.80 257 828 45.15 572 124 87.50 552 776 67.20 429 472 67.53 709 832 127.93 720 012 87.85 541 332 88.22 752 892 168.35 806 652 128.33 712 040 128.70 816 432 206.17 833 564 ' 168.67 780 908 169.05 847,640 235.03 830 288 205.52 820 124 206.87 , 851 928 235.37 830 604 235.70 876 028

127

TABLE AI. Initial estimates of exponents from data of Before considering the next compartment it is worth- experiment I. while to note again a relationship between all A„i 's. Because at t = 0 ail Q., (0) = 0, and because e 1, it follows that the A,,, must sum to zero, i.e. in the case of constant infusion (and primary productivity) experiments Companment E Compartment F Time Time Rr(t) I + A„, + A„2 + . . . + A „,„ - 0

(min) REM (min) Rb.(:) e - RPM] Considering the data set which was employed and that

0.15 0.995 0.43 0.9999 1.02 4.688 1 - A 1 = I - 1.003 = -0.003 1.95 0.918 2.22 0.9963 1.102 " 4.276 4.10 0.830 4.40 0.9839 1.2 3.7659 for a further exponential term would be naive. 7.47 0.719 7.75 0.9536 1.356 3.126 search 1 compartment F to be 12.22 0.574 12.52 0.8914 1.574 2.3949 Data from experiment show 15.90 0.481 16.20 0.8320 1.736 1.9478 accumulating tracer at the next greatest rate. Dividing 20.03 0.407 20.30 0.7612 1.913 1.5465 each data point by the average of the last three values, 30.38 0.241 30.70 0.5910 2.382 0.8482 129.5, we obtain R F(1) values given in Table Al. The 39.69 0.171 40.00 0.4463 2.744 0.4972 plot of R(t) in Fig. 6 shows why the larger values of 60.37 0.053 60.38 0.2300 3.566 0.1440 time, although most nearly associated with a single expo- 84.05 -0.001 84.37 0.1166 5.373 0.0093 nent, are not, in practice, of great value in estimating 126.79 -0.010 127.18 0.0170 5.47 0.0010 exponents. The scatter is simply greater than the infor- 165.64 -0.005 165.9'7 0.0031 5.80 0.0000 much to be avoided, however, 237.75 0.009 238.12 -0.0054 - mation content. Just as 284.20 0.005 284.57 0.0015 612.27 are those points obviously in the curvilinear region which obviously contain more than one exponential term. At this point we deviate from orthodoxy which would have us proceed in the same fashion as we did for the Each data point is divided by QmAN and the fraction _ curve Q E(r). Our rationale for the departure is based on subtracted from 1 to give R(t) values shown in Table Al. three facts. First, it can be shown in theory and in practice Values of R(t) obtained after t 84.05 are of no value that one cannot resolve two exponential terms if the ratio as they represent only statistical fluctuations (negative of the exponents is less than 2. Second, we lcnow the values have no physical meaning so they and all later values curve Q E (t) contains at least two exponents. Third, because are not employed). The remaining values of R E (t) plotted our system is one of interconnecting ccimpartments, the on sernilog paper give a straight line that passes through exponent associated with one compartrnent (i.e. 0.0454 even the earliest points indicating that R E(t) consists of with Q e(r)) may appear in the sum of exponentials asso- only a single exponential term and that our hypothesis ciated with another compartment (i.e. Q (I)). concerning compartment E's relationship to the iodide Having tabulated the values of R y (t), divide each compartment was correct. by e'.°4541 . If this term is present in R(r) we obtain a To obtain the initial estimate of the first exponent limiting value at large values of I approximately equal choose several points on the line and apply the relationship, to A Ir . This plateau value 5.8 (see Table Al) occurs be- given in equation (13). cause R(t) can be written as . . .

In(0.918 /0.995)/(0.15- 1.95) = 0.0447 k-I 1n(0.830 /0.918)/(l.95-4.10) = 0.0469 R(t) = = I /l • e + A k e - à*, In(0.719 /0.830)/ (4.10-7.47) = 0.0426 j=1 In (0.574 /0.719)/ (7.47- 12.22) 0.0474 Dividing R F(t) by e - Xki gives . . . Although this simple graphical analysis has given a k - 1 range of X estimates, the average, 7,, =0.0454, is quite RF(t) î A j e - (4 - hk» + A k "FTO adequate as an initial estimate for the subsemient parameter 1 =1 estimations. Next, using the average value of X , we find the coefficient A l by equation (14). The graph of this equation approaches the value A. at large values of time. 0.995e10°4541 (0.15) = 1.002 Having obtained A, we can take the difference R(t) ■ 0 .9 1 8, 0.0454) ( 4 .10) = 1. 003 and A and plot the difference as shown in Fig. Al. . 830,01°4541 (4.10) = 1.000 0 A straight line extending to the origin indicates that (7.47) = 1.009 0.719 e10°454 only a single terni remains. We find the exponent and coefficient of the term in the manner previously employed. The average value, A , = 1.003, is further evidence that our original hypothesis concerning the compartmental In (4.276 / 4.688)/ (0.43-2.22) = 0.0514 relationships was correct because for that to be true requires In (3.7659 / 4.276)/ (2.22-4.40) = 0.0583 QE(t) to be of the form In (3.126/ 3.7659)/(440.-7.75) = 0.0556 In (2.3939 / 3.126)/ (7.75-12.52) --- 0.0559 QE(t) = AE (1-le") - 71:2 = 0.0553

128 6 The remaining parameter A 2 is found as before . . .

4.688/e° 055 "° 42 = 4.8008 4.276/e -"." 53(2.22' = 4.8345 3.7659/e'."53' = 4.8033 5 3.126/e -°.°35"(7.") = 4.7986 A, = 4.8093

The sum describing the original Qp(t) is 4 Q(t) = 129.5 (1-5.8e-".°454' + 4.8e --.05531 )

Note that the signs associated with the coefficients of each o term are determined by 1 + A + A2 + + AN = 0. 3 This process is continued with the data from compart- i ment B whose radioactivity curve is found to be described o by . . .

Qll (t) = 113.30 (1-1.524e -'.° 54' + 0.524e -°.°5 °' cË- 2

1

0 0 40 80 120

Time (min)

FIG. Al. Graphs of functions containing one less exponen- tial than the original data. The difference between the smallest exponential term, A i e-xlf, and Re) is taken and plotted against time. This graph is then used to obtain the second smallest exponent present in the original data.

129 Cell-Cycle Events in Unicellular Algae

S. PUISEUX-DAO Laboratoire de Biologie cellulaire végétale, Tour 53, Université Paris VII, C.N.R.S. — ERA n°325. 2, Place Jussieu —75005 Paris, France

Introduction molecules. The process implies that at some previous time, DNA replication should have occurred. In nature, circadian rhythms of cell division By measuring the DNA content per cell and by have been observed in some marine dinoflagellates labeling cultures with radioactive DNA precursors, as early as 1958 (Sweeney and Hastings). However, preferably in synchronized cells, it has been shown in natural conditions, the percentage of cells dividing that DNA synthesis takes place during the interphase simultaneously is generally low, and various attempts between two divisions. In rapidly growing pro- have been made to increase it in culture populations. caryotes, the DNA replication lasts throughout the Apart from the selection of cells at a given time interval between two successive divisions. But stage in a population (separation of small cells or in eucaryotes, the duration of nuclear DNA synthesis of dividing cells that do not swim), most other (S period) is shorter than this time interval, and the techniques for synchronizing cell division are based periods between the mitosis (M) and the nuclear on the following principle: cells in an unsynchronized DNA replication have been called gaps: G 1 occurs population progress at different rates through inter- just after the cell division, G, just after DNA syn- phase and mitosis; if they could therefore be blocked thesis (Fig. 1). Such a cell cycle thus comprises at some definite stage of the progression, then all four stages: G 1 , S G2 , and M. In fact, cell meta- those which accumulate at this particular stage bolism is very active during the two gap periods, would recover when the applied blockage was lifted, but this activity does not concern the structure and and would resume their progression in parallel, at quantity of the nuclear DNA itself. least for some time. It is well-known that cell metabolism can decline Devices for accumulating cells at a given phys- when living conditions become unsuitable. The cells iological stage include temperature shocks, inhibitor enter dormancy and this step is called G,, . Most of treatments, deprivation of necessary nutrients, and, the occasions when this occurs, cells pass into G 11 for plant cells, dark periods. For unicellular algae, 1 , but G, cells are also known to become from G suitable light–dark (LD) cycles are most frequently dormant (G,, or G 2 ') (Fig. 1). As soon as the environ- used, sometimes associated with temperature cycles ment is favorable, cells recover and resume their (see in Zeuthen 1964; Cameron and Padilla 1966). normal course through the four usual stages. The best domesticated unicellular algae are Three events of the cycle have a particular either freshwater species like Chlatnydomonas, significance; they are the S phase, and, during mi- Euglena, and Chlorella, or brackish and marine tosis, the prophase and telophase. They indeed phytoplankton like Dunaliella (Volvocales), various correspond to irreversible transitions of cell life. dinoflagellates (Gonyattlax, Amphidinium), some Once DNA has replicated, it is inconceivable for Chrysophyceae like Olisthodiscus luteus, Prasino- the same cell to revert to a state where the quantity phyceae of the genus Platymonas, or diatoms like of DNA is reduced to that before replication; when Cylindrotheca fusiformis. chromosomes have divided, no return to a single This report describes the main findings resulting complement of chromosomes in that cell is possible; from an analysis of cell life in synchronized cultures when one cell has given rise to two daughter cells, of unicellular algae, with particular emphasis, where the new situation is normally irreversible (except possible, on marine species. under very particular experimental conditions, as in cell fusion). It appears to be very clear that all these irre- MAIN EVENTS OF THE CELL CYCLE versible processes are linked to the need for DNA IN EUCARYOTIC CELLS molecules to be equally distributed among the proge- In procaryotic as well as in eucaryotic cells, ny. On the contrary, G 1 and Go consist of reversible cell division is easily detectable and has always steps because all molecules, except DNA, can be been considered as the chief criterion of cell life. destroyed (in general enzymatically) and replaced This event gives rise to two daughter cells, each by new ones when signals are received for the tran- of which carries a normal genome issuing from a scription and translation of the relevant information semiconservative distribution of the parent DNA coded in the genome (Puiseux-Dao 1979).

130 small quantities exist in the mitochondria ( 1 -4%) and in plant cells, chloroplasts can contain up to 112 G 14% of the total DNA (in very peculiar algae like ( Acetabularia, the plastid DNA is even predominant 1 due to the fact that the cell has one nucleus but can 4 G2 /, possess several millions of chloroplasts). I The syntheses of nuclear and extranuclear DNAs f 01 • , .\ 4, would have to be examined separately. / Identification of the nuclear and extranuclear DNAs — DNA molecules are identified according to the physicochemical criteria of molecular weight,

le4 M••-••■••4•••-• configuration, and base composition from determi- nations of the buoyant density (p) after centrifugation FIG. 1. Diagram showing the irreversible transitions of in a CsCI gradient, thermal denaturation temperature, the cell cycle. TI 1 , S phase; TI„,, chromosome division; and so on. TT, , formation of the division membranes. During G, or Table 1 gives some data reported in the littrature G2 the cells can enter dormancy (G,,) or they can return of DNA density in some unicellular algae which to a previous stage (GB : wounded or intoxicated cells which can be used for their identification. digest part of their contents). Go and Gll are reversible stages, but the cells do not follow exactly the opposite path when In general (Fig. 2), in algae, nuclear DNA they progress through the division cycle and when they pass corresponds to 85-90% of the extracted DNA where- to Go or GI,. This gives hysteresis cycles, Go and GB corres- as the chloroplastic DNA represents 3-10%. Proof ponding to bifurcations in mathematical catastrophe theory. of the identity has been obtained by several means: During mitosis, another hysteresis cycle can appear when extracting DNA from isolated organelles (nucleus the cells are treated with colchicine or by low temperatures. or plastid fractions, Brawerman and Eisenstadt 1964); When the blockage is over, then the cells can progress labeling with tritiated thymidine that in some species normally to G, or they can become polyploid (G,„). I, (Euglena, Sagan 1965; Chlatnydomonas, Swinton interphase; M, mitosis (after Puiseux-Dao 1979). and Hanawalt 1972; Chlorella, Dalmon et al. 1975b; Dunaliella, Marano 1979) preferentially labels Main Biochemical Events of the Cell Cycle plastid DNA; and analyzing the DNA of bleached in Unicellular Eucaryotic Algae cells when possible (Schiff and Epstein 1966). One practical difficulty is immediately recognizable, namely that mitochondrial DNA is not easy to study DNA SYNTHESES as it can be easily confounded with plastid DNA. Most of the cellular DNA 85%; see Board- Results obtained for Euglena gracilis have not been man et al. 1971) is found in the nucleus; however, ambiguous in this respect because this alga loses its TABLE 1.

Nucleus Chloroplast chl DNA

%G.0 %G.0 Total DNA

Dunaliella bioculata 1.707 48% 1.696 36% 10-15%

Chlamydomonas reinhardtii 1.723 64% 1.695 36% 6-12%

Euglena gracilis var. bacillaris 1.702 43% 1.685 26% — var. Z 1.708 51% 1.684 25% 2-5%

Chlorella pyrenoidosa: — var. Emerson 1.715 56% 1.689 30% 12% — var. 211/86 1.710 51% 1.687 30% 10%

Acetabularia mediterranea 1.697 38% 1.705 45%

131 Do the peak 1.686 disappears when cells are cultured in darkness so that it can be attributed to plastid a DNA, whereas the shoulder at 1.691 that persists can be considered to correspond to mitochondrial DNA (see in Richards and Ryan 1974). In some species, the fractions of nuclear and chloroplastic genomes that code for their respective ribosomal RNAs have been detected by techniques of hybridization of DNA with the relevant labeled 0.t ribosomal RNA; they are OC rich in Euglena (Stutz and Vaudrey 1971). The nuclear ribosomal DNA could possibly be the AT rich DNA fraction in some other species like Dunedin (Marano 1980). Nuclear DNA synthesis — When the DNA content per cell is followed in samples of synchro- nized cultures collected at successive stages of the life cycle, an increase is observed during interphase and a decrease following mitosis. If similar samples are labeled for a definite t time interval with POI - , or a radioactive base or 1.696 1.707 1.742 nucleoside with the restriction reported above for FIG. 2. The two nuclear (a: 1.707) and chloroplastic DNA thymidine, one observes that the algal nuclear DNA (p: 1.696) of Dunaliella bioculata after analytic ultracentri- becomes significantly labeled only during the period fugation in a CsC1 gradient. (Marker, 49E DNA; 0, 1.742 of increasing DNA content (Fig. 3) that corresponds g/cm"). DO, optical density (after Marano 1979). as stated above to the S phase. chloroplasts when cultivated in the dark. In normal Frequently in algae, the S phase takes place dark-light cycles this species shows a main band just before cell division and the G2 period seems of bulk nuclear DNA (p = 1.707-1.709) with a to be absent. It can be short and difficult to detect shoulder at 1.691 and a well-defined peak at 1.686; for two reasons: first, even an efficient synchro- NB CE LL/ML .Q DNA cpm,,UNA PB/ML eeeeee

0,08

3.10 0,06

0,04

2.10

10 14 16 18 20 T(H) FIG. 3. The cell division cycle of Dunaliella bioculata. Abcissa: time in hours. Ordinates: e---111 number of cell per millilitre; 0-0 total DNA (ag/mL of culture); 13-13 specific activity of nuclear DNA (cpm/i.cg DNA); A__A specific activity of chloroplastic DNA. L, light period; DI , dark period (after Marano 1979).

132 OD CPM a 05 1000

6 12

Tin e at light (hours)

FIG. 4. Increase of DNA content and cell concentration in a synchronized Chlorella culture (after Wanka 1975). B d nization of >95% shows a certain variability; se- cond, many phytoplanktonic algae follow primitive 0.5 mechanisms of division that last longer than the classical mitosis of higher organisms and in which cell furrowing becomes visible before or at the very beginning of nuclear changes. Another peculiarity is that several unicellular algae can produce more than two daughter cells when dividing: for example, in Chlorella the number of daughter cells frequently reaches 16 and, in this case, during the resting phase the DNA content (Fig. 4) undergoes four successive doublings (Chiang and Sueoka 1967b; ,Wanka et al. 1970). At least in some dinoflagellates, the rate of DNA synthesis decreases in the cultures long before FIG. 5. DNA labeling of synchronized Dunaliella cul- the growth has reached a plateau (Galleron and tures. Identical samples have been labeled for 2 h with Durrand 1979) and a loss of DNA has been observed 32 1301- at different times of the cycle. (a) time 4 h of Fig. in stationary phase cells (Allen et al. 1975). 3; (b) time 6 h; (c) 8 h; (d) 10 h, and the extracted DNA (a, nuclear; 0, chloroplastic) has been ultracentrifugod. Chloroplast DNA synthesis — Various results OD, optical density at 254 nm; El, (CPM) specific radio- have been obtained showing either that extranuclear activity (after Marano 1979). DNA replication takes place during the S phase or and Sueoka 1967a; Chlorella , Dalmon et al. 1975a; that it is not concomitant with nuclear DNA syn- Fig. 6 and 7). thesis. In fact, it is possible that the correlation When several plastids are observable in the cells between the chloroplast (ct) and nuclear DNA repli- during at least the first half of the interphase, various cations depends on the number of plastids per cell results have been reported that could be linked to and the number of daughter cells formed at each the particular strains used and to the culture con- division. ditions. The best-known case of Euglena gracilis In the simplest case, where one cell contains strain Z, is complicated by the fact that the chlo- a unique chloroplast and gives two daughter cells, roplast content is influenced by the dark–light treat- for example, in Dunaliella bioculata or Amphidinium ments. Therefore, the ct DNA may vary from ---- 0 carterae , nuclear and extranuclear DNA syntheses (in bleached cells) to 15% of the total DNA, the occur simultaneously, with or without a slight shift most frequent value being about 3-5% of the cellular (Marano 1979; Galleron and Durrand 1979; Fig. 3 and DNA. Moreover, under usual growth conditions, 5). Manning and Richards (1972) have observed that In cells that contain one single plastid at the ct DNA may undergo 1 1/2 rounds of replication beginning of the cell cycle, but give in general four per round of nuclear DNA synthesis. Yet the mean (or more) daughter cells, ct DNA replication has amount of chloroplast DNA per cell remains stable. In been shown to occur twice before the two successive such conditions, the ct DNA content is the result of a nuclear DNA syntheses (Chlatnydomonas , Chiang balance between destructive mechanisms prepon-

133 Mitochondrial DNA synthesis — As reported 45 above, the case of Euglena is the only properly .3.40 studied one. In bleached cells, mitochondrial (mt) DNA is detectable and seems to replicate at the

3.04 same time as nuclear DNA. This replication is not inhibited by cycloheximide, an antibiotic intervening e 1 35 at the level of translation on 80S ribosomes, although 2.72 0 nuclear DNA replication is (Richards and Ryan 1974; Ledoigt and Calvayrac 1979; Fig. 8). 2.38

a 2.04 RNA SYNTHESES 25 o. Characterization of the different RNAs — The • 1.70 2 0 main bulk ( ,--, 80%) of cellular RNAs is comprised 0 of the ribosomal RNAs (rRNAs). In cells they are found in the 6 9 12 3 6 9 12 3 6 9 constituents of the ribosomes which are Noon Midnight cytoplasm, in chloroplasts , and in mitochondria. The different types of ribosomes are characterized by a FIG. 6. DNA synthesis in a vegetative division cycle of sedimentation coefficient (in Svedberg units) meas- Chlamydomonas reinhardtii. Abcissa: time in hours. Two in precise conditions of centrifugation and are periods of DNA increase are visible; the first was shown ured to correspond to the chloroplast DNA replication, the se- made of two subunits, each containing one long RNA cond to the nuclear DNA synthesis (after Chiang and Sueoka chain (Table 2). 1967a). The mitochondria' ribosomes have properties similar to chloroplastic ribosomes which themselves derant in the dark and a synthetic activity linked to resemble procaryotic ribosomes, but in general, in light. Moreover, in illuminated cells this balance is studies concerning algal cells, either they are not related to the growth. These considerations may differentiated from plastid ribosomes or they are explain why some authors (Richards and Manning not considered at all, the attention being focused 1972) have observed a ct DNA replication limited to on chloroplasts. the S phase while others (Cook 1966; Brandt 1975) Cytoplasmic rRNAs come from the nucleus; have reported two ct DNA synthetic periods, one except for the small 5S molecule, they are synthesized when light is given and the second concomitant with at the level of the nucleolus as larger precursors nuclear DNA replication. (45S-30S) and are split later on.

opmx10 4

35,

; 30.

1 3

25. 1 1 1

1 20. 3 1.30 :EI- Z 15 C.) 152 10 g. ------ --- - 4 o

012

0 3 6 9 12 15 18 21 24 27 29 Hours of the cycle

FIG. 7. Incorporation of labeled phosphate (pulse labeling) into the nuclear and chloroplast DNA in

synchronized cultures of Chlorella., 13— D cell number per millilitre x 106 (after Dalmon et al. 1975).

134 DO 6.:1CPM 255 55 3000

18S

35 16S

85 col 000 eé 2000 235

165 000

612Frac lions ?„1000

.E FIG. 9. (a) Ribosomal RNAs from Dunaliella bioculata. 2 Cytoplasmic rRNAs 255 and 18S; chloroplast rRNAs 23S ID and 165; (b) after 1-h labeling with "'POI - . DO, optical 1:1 density; P , (CPM) specific radioactivity (after Marano 100 1980). 0 6 12 Time (hours) RNA syntheses — Of course, for quantitative reasons, most data concern the ribosomal RNAs. Synchronous replication of DNA during the cell cycle Their syntheses have been followed by labeling suc- ---- cell growth cessive samples from synchronized cultures with - nuclear DNA radioactive P01 - , uracil, or uridine and then sepa- mitochondrial DNA rating the different rRNAs by centrifugation or elec- trophoresis (Fig. 9). FIG. 8. Replication of nuclear DNA ( ) and mito- chondria] DNA (_ . . .) in synchronized cultures of In general, the algal cultures are synchronized Euglena gracilis (pulse labeling with radioactive adenine) by dark-light cycles: rRNA content increases in the (after Ledoigt and Calvayrac 1979). light, mainly chloroplastic rRNAs, and this corre- sponds to an active precursor incorporation as is shown in Fig. 10 for Dunaliella . Similar data have TABLE 2. been described in Chlorella (Galling 1973, 1975), Euglena (Heizmann 1970), Chlatnydotnonas (Wilson and Chiang 1977), and Dunaliella (Marano 1980). Cytoplasmic Chloroplastic Both cytoplasmic and chloroplastic rRNAs are syn- ribosomes ribosomes thesized in parallel in illuminated cells, whereas in the dark rRNA labeling is lower (Galling 1973; Wilson and Chiang 1977), and may correspond to Sedimentation coefficient 80S 70S a simple turnover because there is no rRNA weight increase (Cattolico et al. 1973; Wilson and Chiang Two subunits 60S 50S 1977; Marano 1980). Moreover, in Dunaliella , rRNA synthesis is high during the first part of the S phase 40S 30S and declines during the second half. The replicating rRNAs of large subunit 25S-28S + 55 23S + 5S DNA is the lighter fraction during this second period, and might correspond to rDNA as in Chlorella; the small subunit 17-195 165 replicating activity at the level of rDNA possibly inhibits the ribosomal transcription (Marano 1980). Messenger RNAs, at least part of them, are The light dependence results in a cyclic activity recognized by the poly A sequence attached to them. of rRNA synthesis in algal cells which differs from They have a very variable molecular weight. that occurring in yeasts or animal cells. Yet in Transfer RNAs (-----5S) are detected among other Euglena gracilis , grown on lactate medium, RNA small RNAs (natural or resulting from the degra- synthesis was shown to be discontinuous in the light dation of larger molecules) by their capacity to bind as well as in darkness. Heavy RNAs (40S, 35S, a specific radioactive amino acid. 30S) are labeled during the cell division, and chase

135 ,s( r RNA cytop. • 215 • 185 -o 14 1- 4

g 2 20

AS AS 0.5 10 05 1.0

T(h) CELL- CYCLE STAGE FIG. 11. Extent of RNA—DNA hybrid fonnation with RNA samples taken at different times of the cell cycle. rRNA chlor. I I light; KI dark. Percent DNA in hybrid at satura- nuclear DNA (a) and chloroplast .23S tion was detennined for • 16 S DNA (s). % DNA in hybrid corresponds to the value observed at each stage compared with values obtained from asynchronous cultures (A.S) (after Howell 1975).

PROTEIN SYNTHESES In general, in exponentially growing cells, the total protein content doubles during interphase; for algae, this doubling (which can mask a higher rate of synthesis associated with a rapid turnover) takes place during the light period. Of course each protein should have its own synthesis progression and this has been studied by two methods. On the one hand, single types of proteins have been followed during all the cell division cycle: this is possible for enzymes, FIG. 10. Evolution of the synthetic activity of ribosomal cytochromes, or tubulins because these molecules in Dunaliella bioculata during the cell division RNAs easily detectable by their activity (see in Wanka cycle. Abcissa: time in hours; ordinates: specific activity are (A.S) of the ribosomal RNA extracted from samples pulse 1975), their absorption spectrum, or their migration labeled with "'PO. Light, 8 h; dark Ei , 16 h (after on one- or two-dimensional electrophoretic gel. On Marano 1980). the other hand, all the polypeptides of the cells have been studied at different stages of the cycle and the electrophoresis patterns have been compared (Howell experiments suggest that they could be precursors et al. 1977). A similar polypeptide analysis has also of cytoplasmic rRNAs (Ledoigt and Calvayrac 1975). been conducted on plastid fractions (Bourguignon Poly A mRNAs, which are synthesized in the and Palade 1976). nucleus as well as in plastids (Milner et al. 1979), For all proteins studied, as predicted, the data appear preferentially when light is given to the cell show a quantitative or labeling specific evolution (HoWell 1975). As might be expected, transcripts through the cell cycle: this is the case in Euglena of nuclear and chloroplast DNAs vary along the cell or Chlore/la for enzymes (see in Wanka 1975; Fig. cycle. This could be determined from RNA—DNA 12) or cytochromes (see in Ledoigt and Calvayrac reassociation at RNA excess, at least in Chlainy- 1979), for many of polypeptides in Chlamydomonas domonas reinhardtii (Howell 1975). Whereas only (Howell et al. 1977; Fig. 13), and for tubulin in about 12% of the nuclear genome seems to be ho- Chlamydotnonas (Weeks and Collis 1979). mologous to vegetative cell RNA, 60% of the single- Therefore, such a cyclic timing of the protein strand chloroplast DNA is homologous to vegetative synthesis implies coordinated regulation of processes cell RNA. In such conditions possibly the entire at the level of both transcription and translation. chloroplast DNA could be transcribed. But as RNA This regulation is influenced by some external signals, transcript complexity was observed to change during mainly light and dark, as shown for microbody and the cell cycle, the hypothesis of variations of genome mitochondria( enzymes in Euglena; for example, messages along the life cycle is likely to be valid fumarate and succinate dehydrogenase syntheses (Howell 1975; Fig. 11). are repressed by light acting on transcription at the

136 beginning of the light phase and exerting a post- transcriptional control at a later stage (Merrett 1975). Moreover, the coordination involves the three genomes and the different translating machineries of the cytoplasm, the chloroplasts, and the mitochondria which cooperate in synthesizing organelle 70S ri- bosomal proteins (Bogorad et al. 1975), thylakoid membranes (Hoober 1970; Apel and Schweiger 1972; Eytan and Ohad 1972), or enzymes like RUBP car- boxylase (Iwanij et al. 1975).

4 a 12 16 20 24 OTHER METABOLIC ACTIVITIES

FIG. 12. Phosphorylase (s— ) and amylase ( 0-0 ) There is no doubt that many of the physiological activities in synchronized cultures of Chlorella (after Wanka activities should behave with a typical periodicity 1975). during the division cycle as the cell enzymes them- selves show a periodic evolution. This is the case for photosynthesis, and all the associated meta- C S bolic pathways, including the mitochondrial or micro- body physiology. All these problems have been extensively studied for Euglena , Chlorella , Scene- desmus , and Chlamydomonas , and are discussed by others in this bulletin or elsewhere (see in Lefort- Tran and Valencia 1975; Edmunds 1978). This is also the case for nutrient uptake, mineral nutrients FR (POI - , Chisholm and Stross 1976a, b; Si(OH) 1 , Chisholm et al. 1978), as well as for organic nutrients (Pedersed et al. 1975) and for adenine nucleotide content (Weiler and Karl 1979).

Main Morphological Events of the Cell Division Cycle

The simplest morphological events are an in-

Cyc crease in size during interphase and, for swimming cells, a lower motility during the cell division with a flagellar regression observed in some species (Fig. m : membrane polypeptides 13). In Chlamydomonas , this regression is followed s soluble polypeptides by a significant tubulin synthesis (Weeks and Collis 0: 3H- Arginine labeling 1979) probably useful for building the microtubules D:35s -label in9 involved in mitosis, but also in the flagellar mor- 0:31-1-Arginine and 35S-labelin9 phogenesis of the daughter cells. When measuring the surface of the different organelles in electron micrographs, Marano (1980) FIG. 13. Cell-cycle map of labeling of some polypeptides could show in Dunaliella (Fig. 14) that except for in synchronous cultures of Chlamydomonas reinhardtii. the vacuoles, all other organelles followed a parallel Inner arcs show labeling periods for some membrane (m) size evolution. or soluble (s) polypeptides as suggested by experiments with different precursors. Cell-cycle events are indicated on outer circle. (L.D., 12-12; light-bar, light period; dark NUCLEAR EVOLUTION bar, dark period). PS II, Cy„ , and Cy„ , increasing photo- In plates I and II, one easily sees that in Duna- system II, cytochrome 553, and cytochrome 559; et S, liella (L-D: 8 h-16 h) the nuclear and nucleolar chloroplast DNA synthesis; CG, competent gametogenesis; Ca, increasing chlorophyll and carboxydismutase activity; structures change over the cycle during the non- Cy„ , increasing cytochrome 563; FR, flagellar regression; dividing period: at the end of the dark phase, the nu S, nuclear DNA synthesis; NCD, nuclear and chloro- chromatin and the nucleolus, fibrillar at that time, plast divisions; CD, cell divisions; CS, cell separation (after appear very compact (closed). When the light is Howell et al. 1977). given, those structures become less dense (open)

137 S. Cel.Ent.. CHLOROPLAST EVOLUTION L In algae grown in light—dark cycles, the thy- lakoid development is strongly influenced by light. At the end of the dark period, in the green algae studied so far (Chlorella , Atkinson et al. 1974; Euglena , Cook et al. 1976; Dunaliella , Marano 1980), the plastid lamellae are thin, short, and often form stacks similar to grana although the storage content is low. In Euglena (see in Leedale 1967; Buetow 1968) or in the Y-1 mutant of Chlainydo- moms reinhardtii (Ohad et al. 1967), when the dark period is long, the lamellae can disappear entirely (as can the chloroplasts themselves in bleached Euglena). As soon as the cells receive light, the thylakoids become longer and thicker; they associate in pairs; the storage grains increase in number and size (Pl. III). A decrease takes place again in the next dark phase. Such figures fit adequately with the biochemical data concerning the influence of light on the thylakoid chlorophyll and polypeptide syn- theses (Hoober 1970; Eytan and Ohad 1970, 1972; Bourguignon and Palade 1976). 16 24 T(h) O j 16 24 T(h) In algae which contain one single chloroplast, d its division proceeds in parallel with the cell division (Chiamydomonas , Osafume et al. 1972; Chlorella , FIG. 14. Evolution of the mean surface (arbitrary units) Atkinson et al. 1974; Dunaliella , Marano 1980). of organelles in Dunaliella bioculata during the cell cycle. On the contrary, when several chloroplasts are S.Cel Ent., surface of the entire cell; S. Pla., chloroplast observed in the cells, the chloroplast division is surface; S. Cyt., cytoplasm surface; S. Noy., nucleus sur- division. The organelles face; S. Vac., vacuole surface; S. Mit., mitochondrial not strictly linked to the cell surface; L, light; N, night; T(h), time in hours. Measure- divide before the cell division or may even do so ments have been made on micrographs of longitudinal axial without any cell division during the stationary culture sections (after Marano 1980). phase (Olisthodiscus , Cattolico et al. 1973). In Euglena , fusion of small plastids that takes place before their division has been described (Lefort-Tran and the number of preribosomal-type granules in- 1975). creases inside and around the nucleolus. The loose chromatin aspect corresponds to the S phase, whereas MITOCHONDRIAL EVOLUTION the formation of particles by the nucleolus is con- comitant with a high level of rRNA synthesis. Then In the electron micrographs, sections of mito- the structures return to the closed configuration of chondria are generally small. However, in Chlamy- the dark period (Marano 1980). doinonas (Osafume et al. 1972; Arnold et al. 1972), Chlorella (Atkinson et al. 1974), or Euglena (Cal- A similar evolution has been described in vayrac et al. 1974; Lefort-Tran 1975), observations Euglena (Frayssinet et al. 1975) as well as in Astasia of serial sections, under the light and the electron (Chaly et al. 1977). microscope, have suggested that a true mitochon- In most of the unicellular algae, mitosis does drial cycle exists. Small mitochondria observed not proceed exactly as in higher organisms. In after the cell division fuse forming a giant organelle general the mitoses are enclosed within a persistant which fragments or divides irregularly just prior to nuclear membrane, more or less complete, and dif- mitosis (Pl. IV). ferent types of mechanisms exist which are linked to various microtubule structures which are extra- nuclear in dinoflagellates (see Matthys-Rochon 1979), Regulation of the Cell Division Cycle intranuclear among volvocales (see Marano 1980), and of a very particular configuration in diatoms The regulation of the cell division cycle (cdc) (Pickett-Heaps et al. 1979a, b). An extensive study of has been extensively studied on animal cells in cancer the so-called "primitive mitoses" has been described research; much data have also been obtained on yeast by Pickett-Heaps' group. and on unicellular algae, for which light plays a

138 47000 After darkness

• 37000 After 4 hours of light 39 000

PLATE I. The nucleus of Dunaliella bioculata at different times of the day in synchronized cultures (L—D: 8-16 h). Nu, nucleolus; ch, chromatin; mb, nuclear membrane (after Marano 1980). Magnification indicated before print reduction to 46%.

139 49000 D

PLATE II. Closed aspect of the nucleolus in the nucleus of Duna/id/a in the dark (D); open aspect in the light (L) (after Marano 1980). Magnification indicated before print reduction to 70%.

140 . 28400

.23600 D PLATE III. The chloroplast of Dunaliella bioculata in synchronized cultures (L—D: 8-16 h). L, after 4-h light; D, at the end of the dark period; P, chloroplast; Py, pyrenoid; Mi, mitochondria (after Marano 1980). Magnification indicated before print reduction to 68%.

141 PLATE IV. The mitochondria of Euglena gracilis in synchroniz,ed cultures (A) during interphase and (B) during cell division. Pa, paramylon; F, flagellum; G, dictyosome; Mi, mitochondrion (by courtesy of Calvayrac). Magnification indicated before print reduction to 75%.

142 central role. Despite the apparent difficulties in which remained in the metabolically "dormant" mode discussing examples from such diverse sources, we are turned to a new physiological program by events find sufficient similarity for our purposes to justify occurring at the level of the external membrane. this approach. A similar situation has been described for some animal cells whose physiology is directed by peptidyl hormones (see Robison et al. 1971; Greengard et al. CELL-CYCLE PHENOMENA AND THE 1972; Koch and Leffert 1979; Rubin et al. 1979). But CELL DIVISION CYCLE in this case, the membrane modifications are under In exponentially growing Saccharomyces or the hormonal influence which activates at least adenyl Euglena many physiological functions have been cyclase, and have been shown to result in a higher shown to take place with a definite periodicity: for cAMP content inside the cells and, as a consequence, example, bud initiation in the yeast and photosyn- in a higher phosphorylating kinase activity which thetic capacity in algae. These rhythmic activities induces or inhibits precise metabolic pathways. persist even when the cells do not divide; this has been For microorganisms living in relatively simple observed in temperature-sensitive Saccharomyces media, nutrients likely play the same role as the spe- mutants transferred to a temperature that inhibits cific molecules mentioned above. Indeed changes in mitosis but not budding, which then initiates periodi- the physiological programs occur when the cells cally; this is typical of Euglena cultures in the sta- are transferred to depleted solutions: this is well- tionary phase where many metabolic pathways can known for bacterial sporulation in Bacillus, ascospore function with a periodicity, in general circadian (see formation in Saccharomyces, or cell sexualization Edmunds 1975, 1978). Though interrelations do exist in Chlamydotnonas. At least in Bacillus subtilis between the network of metabolic pathways leading and Saccharomyces, the new metabolic activities to mitosis rhythms and the network responsible for seem to be linked to a new type of phosphorylation other periodic cellular events, we shall try to limit pattern (see in Rhaese et al. 1979). Moreover, simple this discussion to the cell division cycle consisting modifications of the external ion concentration can of G I , S, G, and M in eucaryotes, plus G,,. induce cells to choose a new metabolic orientation. Two properties are to be emphasized: (1) the For example, in Blastocladiella motile zoospores cdc is sensitive to environmental conditions (nutri- transform into a sessile growing germling with a ents, temperature, light for plant cells) that are able high degree of synchrony when IC± is added to the to impose their own periodicity (entraining con- medium; the transformation involves retraction and ditions); yet (2) the cdc can persist, at least for a disassembly of the flagellum, fragmentation of the certain time, when the environmental conditions single mitochondrion , dispersal of the ribosomes from become constant (free-running conditions). the nuclear cap through the cytoplasm, and de novo The easiest way to attack the problem is to synthesis of a chitinous cell wall, changes that take analyze what occurs when resting cells are engaged place in 1 min and do not require protein synthesis in a new run of successive division cycles. This (Soll arid Sonnebom 1972; Van Brunt and Harold takes place when nutrients are added to the medium 1980). of stationary cultures. This is also observed after Another fact of interest is that all cells are able fertilization of dormant female gametes or when to modify their movements (animal cells) or cyto- lymphocyte cultures are given a lectin. In this last plasmic streaming (plant cells) very rapidly. This is case, it has been shown that the attachment of the due to the following property: the structures involved lectin to the lymphocyte plasma membrane induces in the cytoplasm hydrodynamics, such as micro- several very rapid reactions: ion transport change tubules, can appear and disappear in seconds when and acceleration of phospholipid turnover conco- the environment (ion content, temperature) is mitant with an increase in the membrane lipid fluidity changed. Each type of structure has its own envi- (Whitney and Sutherland 1972; Hui et al. 1979; ronmental reactions, but all need ATP or GTP and Holian et al. 1979; Segel et al. 1979). Various phosphorylations for building and functioning. More- membrane enzymatic activities are then enhanced over, in Acetabularia, one could observe that cellular (cyclases, ATPases, methyltransferases, phospho- streaming, when arrested by darkness, recovers after lipases) (Hadden et al. 1974; Hirata et al. 1980). a few seconds of light, blue light being much more On the other hand, fertilization has been demonstrated efficient than red light; this response controls RNA to trigger a very rapid and precise program of ion transport from the nuclear basal part towards the exchanges between the cells and the surrounding apical growth zone (Puiseux-Dao et al. 1980; Dazy medium (Epel 1977). Therefore, inducing mitotic et al. 1981) and chloroplast transport which shows activity seems to be linked to changes in the plasma a circadian rhythm correlated with the diurnal perio- membranes with consequences at this level on lipid dicity of electrophysiological activity (Broda et al. fluidity, ion transport, and enzyme function. Cells 1979; Schweiger et al. 1981; Dazy et al. 1980).

143 In Euglena , Schiff (1975) described two photo- controls of chloroplast development: one is blue-light dependent and influences nuclear DNA transcription, in particular to form cytoplasmic ribosomal RNA necessary for the translation of some plastid proteins; on the other hand chloroplast ribosomal RNA syn- thesis is regulated by both blue and red light, impli- cating another type of receptor (protochlorophyll or or the like). A similar double photocontrol has also been proposed for the regulation of chloroplast DNA content in Euglena (Srinivas G, ; M11;. G, and Lyman 1980). A M, M, When analyzing the data on Euglena and Ace- Environment tabularia together, it seems possible to propose at A variable duration least in green algae that two photocontrols can co- Fin. 15. Diagram showing our model for the cell-cycle exist: one is very sensitive to blue light and respon- regulation. In G 1 , external conditions influence the syn- sible for cytoplasmic reactions (among the more rapid thesis (at least) of a plasma membrane protein P, whose responses, one finds the acceleration of cytoplasmic level modifies the cell permeability to ions (and water). streaming which in turn can activate the metabolism At a given level of P1 , the ion pattern ("ionic spectrum") by the way of ion, molecule, and organelle transport); inside the cell that influences phosphorylation and other the second takes place inside chloroplasts and directs biochemical processes becomes suitable for DNA syn- the plastid physiology with secondary consequences thesis and at a correlated time for the synthesis of another for the cytoplasm. protein (P2 ) of the plasma membrane. P, synthesis does Turning back to animal cells or Blastocladiella , not necessarily begin exactly when DNA replication starts. When the level of P2 is high enough (associated with the changes of the metabolic program are observed when decrease of P, due to the depletion of the specific mRNAs), information is received at the level of the plasma the new ion balance inside the cell favors tubulin synthesis membrane from ions, nutrients, or other external (TuS) and progressively but rapidly the evolution of mi- molecules. Then enzymes (cyclases, ATPases) of tosis. At the end of mitosis, mRNAs for P2 are depleted, the plasmalemma modify their activity which, in turn, so the level of this protein decreases while synthesis of influences ion transport and the phosphorylating new membrane proceeds. This model is based on modi- pattern of the cells. This should, of course, be enough fications of the plasma membrane permeability to ions to transform by degrees all the cellular biochemical (and water), which change the internal range of ion content. reactions which depend upon ions and phosphoryla- Driven by the ion content, the cell machinery works, at least for each of the two considered proteins (minimal tion. But the architectural components, such as hypothesis), with successive waves of mRNA and protein microfilaments and microtubules, are also very sen- syntheses. When external signals are given to the cell sitive to ions and phosphorylation; they may respond (light, hormones), the internal ion pattern is adjusted in (cf. the rapid loss of the flagellum in Blastocladiella the convenient range (entrained rhythm in the case of after the addition of K+ to the medium) from nucle- plant cells). When the cell does not receive signals from ation centers on the plasma membrane itself, accel- outside, the periodic functioning of the cell machinery can erating or inhibiting cellular streaming. This control go on until the progressive shift of the internal "ion spec- on cytoplasmic streaming amplifies the cellular trum" (linked to a slow down of protein synthesis) has rhythm). response, which would be slow if diffusion alone become too large (decreasing free-running were responsible for transporting biochemical ma- terials. tion. But this type of control can also work for the Therefore, we have proposed (Puiseux-Dao regulation of the cell cycle as proposed in Fig. 15. 1979) that the physiological signals acting on the This proposal is based on the experimental evidence plasmalemma (on proteins or lipids) are transformed that practically all molecules are always present in into ionic messages and phosphorylating capabilities the cell, but they do or do not always function. with consequences first on the membrane enzymes, In our hypothesis (the most economical one), then on the cell hydrodynamics, and then on all the after mitosis and under usual conditions the cells internal biochemical reactions. This is also valuable enter a resting period (A) that lasts until a start signal for the membranes containing at least one photo- is given by the environment, which in general for receptor, which is the case of the thylakoids, which algae is light. This information induces a new definite are known to produce ion exchanges and ATP in light. ion composition ("ionic spectrum") inside the cell Under such conditions it is clear that external which in turn can achieve the possibility of synthe- information can drive the cell from one physiological sizing a short-lived protein (P, ) of the plasma mem- program to another, for example, during differentia- brane. As the P, level increases, the plasmalemma

144 permeability may change and at a given time becomes Signal available for nuclear DNA replication and at a cor- A I GI 5G2 M related time for the synthesis of another short-lived G O protein (P.,) of the plasma membrane. The cessation Al ter mitosis 1 of P, synthesis, probably due to the depletion of the Signal specific mRNA, occurs more or less simultaneously. GI — S When the plasma membrane content in of P, and P., Cycle 1 reaches certain precise values, then the "ionic spec- S G2

trum" can allow prophase (M, ) to begin anfl for • G2 — M2

other values, telophase (M2 ) can take place. Then After M2 the mRNAs for P, may be used up or at least at too Signet —sG,

low a level and P., synthesis ceases. The cells which Cycle 2 r have accumulated mRNAs for P, during the second GI S part of the described cell cycle can wait in A for a Last signal new environmental signal. A GI 5 G2 M The existence of a short-lived protein involved in the regulation of G, traverse has been recently GI after MO suggested from pulse experiments with cyclohexi- Mt mide (see Shilo et al. 1979). Alter MI M2 r--

In fact our hypothesis shall be modified to Alter M2

include the various experimental situations described 143canrat achieve in the literature. First, it is necessary to explain why part of the rhythmic activity of the cells (see 02 cannot seineve budding in Saccharomyces, photosynthesis in Eu- glena) can be disconnected from the cell division. cannot edge. This part might be linked to the environmental signal 3 GI cannot achle. GO and to P, synthesis, but not to P2 synthesis that could not take place when the P, level does not FIG. 16. (A) Hypothetical evolution of the cell internal reach a certain value (depleted medium) or when "ionic spectrum" E21 during the cell division cycle. The the temperature is not suitable (temperature-sensitive external signal (light for example) sets the shifted "ionic mutant). The periodic metabolism would be due to spectrum" to the right position after each cycle and G, the use of the mRNA stock for P, and its replace- starts progressing normally. (B) Free-running conditions. ment. After the first mitosis, the ionic spectrum has shifted a little from the optimum range. However, G, can still start The hypothesis must also account for circadian and the cell cycle can progress. At each round, in the rhythms which persist though declining for about absence of resetting by an external signal, the position of 2 wk in free-running conditions. Our model is based the ionic spectrum is more and more shifted, the traverse on the alternation of two opposite waves of translation through the cell cycle becomes progressively slowed, and and transcription concerning P 1 , which influence the even the last events (mitosis, then G2 then S) disappears cell permeability and in turn may induce delayed but one after the other. Then G, itself cannot take place and the cells enter G,. Of course, the rate of disappearance similar waves of translation and transcription for P2. of events occurring during the cell cycle might depend on The starting signal, which comes from the environ- the cell type and the external conditions. However, it ment and functions in G , is necessary for resting is clear from experimental data that the cell division stops cells, but once given it probably serves not only for before other processes like photosynthesis which normally one unique cycle but for several because the oscil- increases in G, and that except in precise conditions (in- lating machinery can function in a range of ionic toxications for example), the cells enter G„ from 0 1 . composition and has its own inertia; in the absence of a new impulse from the outside, it will cease working from 0 1 . Of course, the shift of the internal "ionic after some cycles (whose number may depend upon spectrum" might be variable with the cell type and the cell type and external conditions) after the "ionic environmental factors. From the experimental data, spectrum" is shifted, by degrees, out of the con- which mainly concern the cell division and the venient range. In Fig. 16, the model predicts that in photos9nthetic activity, we can be assured that the free-running conditions, the cell division should be second part of the cell cycle disappears before the the first event to cease, then the G, period, then the S first, that is G, . However, the possibility of obtaining phase, then G I . This could explain why in stationary polyploid cells has to be considered because the S cultures, no mitosis takes place while rhythms of the phase can occur without any cell division. photosynthetic activity are still observable. This also In our hypothesis, the biological clock is linked fits with the fact that in general, cells enter dormancy to alternate waves of transcription and translation

145 of at least two short-lived membrane proteins. But the (leading to G 11) to new conditions suitable for growth model must account for some specific properties of recovery. In fact we must remember that in the model, the biological clock: (1) it may be temperature inde- the progression through G, is linked to the synthesis pendent; (2) at least for plant cells, it may not func- of a protein P, from a stock of mRNA accumulated tion under bright illumination. As the plasma mem- in the preceding cell division cycle. We can assume brane is the fundamental structure implied in our that during the resting stage or at least the stationary model (coupled with chloroplast membranes in plant period, this stock of specific mRNA slowly decreases cells), we have to look for properties which are in such a way that in the cell population when the rapidly modified at that level when the temperature transfer to good conditions takes place, some cells is changed. In fact, in animal cells the lipid compo- would be unable to recover, others would recover sition of membranes can adapt almost instantaneously very slowly, and others more rapidly depending upon with an increase of unsaturated fatty acid chains the rest of mRNA still present. at low temperatures and on the contrary a high content Of course, this model is the simplest we could of saturated fatty acids at high temperatures. Under propose based on our hypothesis, which empha- such conditions, rapid modifications of this type, sizes the rapid response of the plasma membrane similar to the changes observed when a lectin attaches or membrane containing photoreceptors to environ- to the lymphocyte external membrane, maintain mental messages. It will probably be modified, in the fluidity of the membranes at a constant level. the future, but it calls attention to the regulatory Such an explanation has already been proposed by role of ions which can influence all macromolecule Njus et al. (1974) for explaining the temperature configuration and functioning. In nature, pollutants independence of the biological clock. that are lipophilic molecules or metals may interfere The loss of rhythm in bright illumination ob- with this type of regulation. served in plant cells is probably due to the effects of high light intensities on the cell permeability and References the behavior of the cytoskeleton. In Acembularia, we have seen that under such conditions, mainly J. R., T. M. ROBERTS, A. R. LOEBLICH HI, dark-treated algae transferred to light, large move- ALLEN, in AND L. C. KLOTZ. 1975. Characterisation of the ments of water, and probably mineral ions, occur DNA from the dinoflagellate Coptothecodinium which induce plasmolysis and even rupture of the cohnii and its implication for nuclear organization. cytoplasm and the vacuole; but also the thin strands Cell 6: 161. of cytoplasm break one after the other (Puiseux-Dao ALTKINSON, A. W., P. C. L. JOHN, AND B. E. S. et al. 1980). GUNNING. 1974. The growth and division of the Another fact has to be explained: the phased single mitochondrion and other organelles during the growth described at least for some dinoflagellates cell cycle of Chlorella. Protoplasma 81: 77. H. G. SCHWEIGER. 1972. Nuclear (see Chisholm 1981). This type of growth concerns APEL, K., AND dependency of chloroplast proteins in Acetabularia. cultures in which, every day, a portion of the cells Eur. J. Biochem. 25: 229. divide at a precise time, which is always the same; ARNOLD, C. C., O. SCHIMMER, F. SCHOTZ, AND M. therefore the cell division is synchronized in relation BATHELT. 1972. The mitochondria of Chlamydo- to the time of day, but does not take place in all cells. monas reinhardtii . Arch. Mikrobiol. 81: 50. It seems that at the beginning of each light period, the BOARDMAN, N. K., A. W. LINNANE, AND R. M. SMIL- algae have been reset to a "start position," even when LIE [ed.]. 1971. Autonomy and biogenesis of mito- they have -not achieved the preceding cell division chondria and chloroplasts. North-Holland. Publ. cycle. In our hypothesis such resetting could be Co., Amsterdam and London. M. R. HANSON, AND internal "ion spec- BOGORAD, L., J. N. DAVIDSON, simply due to a shift back of the L. J. METS. 1975. Genes for proteins of chloroplast trum" of the cells. This could be induced possibly by ribosomes and evolution of eucaryotic genoines, p. the light—dark transition. In Acetabularia , each time 111. In S. Puiseux-Dao [ed.] Molecular biology of the light is turned off, an action potential takes place nucleocytoplasmic relationships. Elsevier, Amsterdam involving transmembrane ion movements (Rogatykh and New York. et al. 1979). Under such conditions, at least for some BOURGUIGNON, L. Y. W., AND G. E. PALADE. 1976. dinoflagellates, both light—dark and dark—light tran- Incorporation of polypeptides into thylakoid mem- sitions would be able to modify the "ionic spectrum" brane of Chlamydomonas reinhardtii . J. Cell Biol. which could explain why Weiler and Eppley (1979) 69: 327. BRANDT, P. 1975. Zwei Maxima plastidârer DNA-Syn- could relate the time division of Cm-anion both to the these mit unterschiedlicher Lichtabhângigkeit im onset of dark and of light. Zellcyclus von Euglena gracilis. Planta 124: 105. Another problem should be discussed: the lag BRAWERMAN, G., AND J. M. EISENSTADT. 1964. period observed in growth curves when, for example, Deoxyribonucleic acid from the chloroplasts of algae are transferred from stationary conditions Euglena gracilis. Biochim. Biophys. Acta 91: 477.

146 BRODA, H., G. SCHWEIGER, H. U. KOOP, R. SCHMID, DAZY, A. C., H. BORGHI, AND Y. PICHON. 1980. AND H. G. SCHWEIGER. 1979. Chloroplast migration: Activité bioélectrique spontanée chez l'algue uni- a method for continuously monitoring a circadian cellulaire Acetabularia mediterranea C.R. Acad. rhythm in a single cell of Acetabularia, p. 163. In S. Sci. 291: 637. Bonotto, V. Kefeli, and S. Puiseux-Dao [cd.] Develop- DAZY, A. C., S. PuisEux-DAO, M. DURAND, AND mental biology of Acetabularia. Elsevier-North Hol- A. SANTA MARIA. 1981. The effects of blue and land, Amsterdam, New York, and Oxford. red light on streaming recovery and RNA transport BUETOW, D. E. [ed.]. 1968. The biology of Euglena after dark treatment in Acetabularia mediterratzea. Academic Press, New York and London. Protoplasma 105: 356. EDMUNDS, L N. JR. CALVAYRAC, R., O. BERTAUX, M. LEFORT- TRAN, 1975. Temporal differentiation in AND R. VALENCIA. 1974. Généralisation du cycle Euglena: Circadian phenomena in non-dividing mitochondrial chez Euglena gracilis (Z) en cultures populations and in synchronously dividing cells. synchrones, hétérotrophe et phototrophe. Protoplasma In Les cycles cellulaires et leur blocage chez quelques 80: 355. Protistes. Colloques CNRS, France 240. p. 53. 1978. CAMERON, I. L., AND G. M. PADILLA [cd.]. 1966. Clocked cell cycle clocks: implications Cell synchrony: studies in biosynthetic regulation. towards chronopharmacology and aging, p. 125. In Academic Press, New York and London. H. V. Samis and S. Capobianca [cd.] Aging and biological rhythms. Plenum Publishing Corp., New CATTOLICO, R. A., J. W. SENNER, AND R. F. JONES. York, NY. 1973. Changes in cytoplasmic and chloroplast ribo- EPEL, D. 1977. The somal ribonucleic acid during the cell cycle of Chlamy- program of fertilization. Sci. Am. 237: 128. domonas reinhardtii. Arch. Biochem. Biophys. 156: 58. EYTAN, G., AND I. OHAD. 1970. Biogenesis of chloro- plast membranes. J. Biol. Chem. 245: 4297. CHALY, N., A. LORD, AND J. G. LAFONTAINE. 1977. 1972. A light and electron microscope study of nuclear Biogenesis of chloroplast membranes. J. Biol. Chem. 247: 112. structure throughout the cell cycle in the euglenoid Astasia longa. J. Cell Sci. 27: 23. FRAYSSINET, C., O. BERTAUX, AND R. VALENCIA. 1975. Noyau et nucléole chez Euglena au cours du CHIANG, K. S., AND N. SUEOKA. 1967a. Replication cycle cellulaire. In Les cycles cellulaires et leur blocage of chloroplast DNA in Chlamydomonas reinhardtii chez quelques Protistes. Colloques CNRS, France during vegetative cell cycle: its mode and regulation. 240. p. 291. Proc. Natl. Acad. Sci. 57: 1506. I967b. Replication of chromosomal and cyto- GALLERON, C., AND A. M. DURRAND. 1979. Cell cycle plasmic DNA during mitosis and meiosis in the and DNA synthesis in a marine dinoflagellate Am- eucaryote Chlatnydomonas reinhardtii. J. Cell. phidiniwn carterae. Protoplasma 100: 155. Physiol. 70 Suppl. 1: 89. GALLING, G. 1973. Synthesis of chloroplast ribosomes CHISHOLM, S. W. 1981. Temporal patterns of cell division during the development cycle of the Chlorella cell. in unicellular algae. Can. Bull. Fish Aquat. Sci. 210: Biochem. Physiol. Pflanz. 164: 575. 150-181. 1975. RNA synthesis in synchronous cultures CHISHOLM, S. W., F. AZAM, AND R. W. EPPLEY. 1978. of unicellular algae. In Les cycles cellulaires et le& Silicic acid incorporation in marine diatoms on blocage chez quelques Protistes. Colloques CNRS, light:dark cycles: use as an assay for phased cell France 240. p. 225. division. Limnol. Oceanogr. 23: 518. GREENGARD, P., R. PAOLETTI, AND G. A. ROBISON CHISHOLM, S. W., AND R. G. STROSS. 1976a. Phosphate [cd.]. 1972. Physiology and pharmacology of cyclic uptake kinetics in Euglena gracilis (Z) (Eugleno- AMP. Raven Press, New York, NY. phyceae) grown in light-dark cycles. I. Synchronized HADDEN, J. W., E. M. HADDEN, AND N. D. GOLDBERG. batch cultures. J. Phycol. 12: 210. 1974. Cyclic GMP and cyclic AMP in lymphocyte 1976b. Phosphate uptake kinetics in Euglena metabolism and proliferation, p. 237. In W. Braun, gracilis (Z) (Euglenophyceae) grown in light-dark L. M. Lichtenstein, and C. W. Parker [cd.] Cell cycles. II. Phased PO 1-limited cultures. J. Phycol. growth and the immune response. Springer Verlag, 12: 217. New York, NY. COOK, J. R. 1966. The synthesis of cytoplasmic DNA HEIZMANN, P. 1970. Propriétés des ribosomes et des in synchronized Euglena. J. Cell Biol. 29: 369. RNA ribosomiques d'Euglena gracilis. Biochim. COOK, J. R., H. SMITH, AND P. HARRIS. 1976. Cyclic Biophys. Acta 224: 144. changes in chloroplast structures in synchronized HIRATA, F., S. TOYOSHIMA, J. AXELROD, AND M. Euglena gracilis . J. Protozool. 23: 368. WAXDAL. 1980. Phospholipid methylation: a bio- DALMON, J., M. BAYEN, AND R. GILET. 1975a. Periodic chemical signal modulating lymphocyte mitogenesis. synthesis of DNA from Chlorella pyrettoidosa Proc. Natl. Acad. Sci. 77: 862. (strain 21 1/8B and strain Emerson). In Les cycles HOLIAN, A., C. J. DEUTSFI, S. K. HOLIAN, R. P. cellulaires et leur blocage chez quelques Protistes. DANIELE, AND D. F. WILSON. 1979. Lymphocyte Colloques CNRS, France 240. p. 179. response to phytohaemagglutinin: intracellular volume DALMON, J., M. BAYEN, AND A. RODE. 19756. In- and intracellular (K+). J. Cell. Physiol. 98: 137. corporation of radioactive precursor into the desoxy- HOOBER, J. K. 1970. Sites of synthesis of chloroplast ribonucleic acid of Chlorella pyrenoidosa, Strain membrane polypeptides in Chlamydomonas reinhard- Emerson. Z. Pflanzen Physiol. 74: 189. tit y-1. J. Biol. Chem. 254: 4327.

147 HOWELL, S. H. 1975. Changing expression of the nuclear cellular cycle of Chlamydomonas reinhardtii in and chloroplast genomes during the cell cycle in synchronous culture. Plant Cell Physiol. 13: 211. Chlamydomonas reinhardtii . In Les cycles cellulaires PEDERSED, A. G., R. PETTERSEN, AND G. KNUTSEN. et leur blocage chez quelques Protistes. Colloques 1975. Uptake and metabolism of uracil, guanine and CNRS, France 240. P. 159. phenylanine by synchronized Chlore/la fusca . In HOWELL, S. H., J. W. POSAKONY, AND K. R. HILL. Les cycles cellulaires et leur blocage chez quelques 1977. The cell cycle program of polypeptide labelling Protistes. Colloques CNRS, France 240. p. 203. in Chlatnydomonas reinhardtii . J. Cell Biol, 72: 223. PuISEUX-DAO, S. 1979. Mouvements cytoplasmiques Hui, D. Y., G. L. BEREBITSKY, AND J. A. K. HARMONY. et morphogenèse chez l'Acetabularia mediterranea 1979. Mitogen-stimulated calcium ion accumulation Biol. Cell. 34: 83. by lymphocytes. J. Biol. Chem. 254: 4666. PuisEux-DAO, S., A. C. DAZY, M. DURAND, F. IWANIJ, V., N. H. CHUA, AND P. SIEKEVITZ. 1975. MARANO, C. BRUN, D. HOURSIANGOU - NEUBRUN,

Synthesis and turnover of ribulose biphosphate carbo- AND A. SANTA - MARIA. 1980. Régulation par la xylase and its subunits during the cell cycle of Chlamy- lumière et la teneur en Ca" du milieu des mouve- domonas reinhardtii. J. Cell Biol. 64: 572. ments cytoplasmiques et du transport des RNA chez KOCH, K. S., AND H. L. LEFFERT. 1979. Increased l'Acetabularia. Biol. Cell. 39: I la.

sodium ion influx is necessary to initiate rat hepatocyte PICKETT - HEAPS, J. D., D. H. TIPPIT, AND J. A. AN- proliferation. Cell 18: 153. DREOZZI. 1979a. Cell division in the pennate diatom LEDOIGT, G., AND R. CALVAYRAC. 1975. Synthèses Pinnularia, I and II. Biol. Cell. 33: 71 and 79. périodiques, d'ARN chez Euglena gracilis en culture 1979b. Cell division in the pennate diatom synchrone. In Les cycles cellulaires et leur blocage Pinnularin , III and IV. Biol. Cell. 35: 195 and 199. chez quelques Protistes. Colloques CNRS, France 240. RHAESE, H. J., R. SCHECKEL, R. GROSCURTH, AND p. 241. G. STAMMINGER. 1979. Studies on the control of 1979. Phénomènes périodiques, métaboliques et development. Highly phosphorylated nucleotides structuraux chez un Protiste, Euglena gracilis. J. (HPN) are correlated with ascopore formation in Protozool. 26: 632. Saccharomyces cerevisiae . Mol. Gen. Genet. 170: LEEDALE, G. F. [ed.]. 1967. Euglenoid tlagellates. 57. Prentice-Hall, Inc., Englewood Cliffs, NJ. RICHARDS, O. E., AND J. E. MANNING. 1975. Replication LEFORT -TRAN, M. 1975. Mitochondries et chloroplastes of chloroplast DNA in Euglena gracilis . In Les cycles chez Euglena en culture synchrone. In Les cycles cellulaires et leur blocage chez quelques Protistes. cellulaires et leur blocage chez quelques Protistes. Colloques CNRS, France 240. p. 25. Colloques CNRS, France 240. p. 297. RICHARDS, O. C., AND R. S. RYAN. 1974. Synthesis LEFORT -TRAN, M., AND R. VALENCIA [cd.]. 1975. and turnover of Euglena gracilis mitochondrial DNA. Les cycles cellulaires et leur blocage chez quelques J. Mol. Biol. 82: 57. Protistes. Colloques CNRS, France 240. ROBISON, G. A., G. G. NAHAS, AND L. TRINER [cd.]. MANNING, J. E., AND O. C. RICHARDS. 1972. Synthesis and cell function. Ann. N.Y. 1971. Cyclic AMP and turnover of Euglena gracilis nuclear and chlo- Acad. Sci. 185. roplast desoxyribonucleic acid. Biochemistry 11:2036. ROGATYKH, N. P., V. G. MELKUMYAN, AND T. N. MARANO, F. 1979. Biosynthèse des DNA nucléaire et chloroplastique chez une Volvocale sans paroi. Biol. ZUBAREV. 1979. Action potential signals and cell Cell. 36: 65. moiphogenesis in Acetabularia. In S. Bonotto, V. 1980. Synchronisation et cycle cellulaire du Kefeli, and S. Puiseux-Dao [cd.] Developmental Dunaliella bioculata, Utilisation de ce modèle expé- biology of Acetabularia. Elsevier-North Holland, rimental pour l'étude du mode d'action de l'acroléine. Amsterdam, New York, and Oxford. Thèse d'état, Université Paris 'VII. RUBIN, A. H., M. TERASAKI, AND H. SANUI. 1979., Major intracellular cations and growth control: cor- MATTHYS -ROCHON, E. 1979. Évolution d'un dinoflagellé libre au cours du cycle cellulaire. Biol. Cell. 35: 313. respondence among magnesium content, protein MERRETT, M. J. 1975. The temporal expression of micro- synthesis and the onset of DNA synthesis in BALB/ body and mitochondrial enzymes in division-syn- C3T3 cells. Proc. Nat!. Acad. Sci. 76: 3917. chronized Euglena cultures. In Les cycles cellulaires SAGAN, L. 1965. Unusual pattern of tritiated thymidine et leur blocage chez quelques Protistes. Colloques incorporation in Euglena. J. Protozool. 12: 105. CNRS, France 240. p. 123. SCHIFF, J. A. 1975. The control of differentiation in MILNER, J. J., C. L. HERSHBERGER, AND D. E. BUETOW. Euglenn . In Les cycles cellulaires et leur blocage 1979. Euglena gracilis chloroplast DNA codes for chez quelques Protistes. Colloques CNRS, France polyadenylated RNA. Plant Physiol. 64: 818. 240. p. 79. NJUS, D., F. SULZMAN, AND J. W. HASTINGS. 1974. SCHIFF, J. A., AND H. T. EPSTEIN. 1966. The continuity Membrane model for the circadian clock. Nature of the chloroplast DNA in Englena, p. 131. In M. (London) 248: 116. Locke [cd.] Reproduction: cellular, subcellular and OHAD, I., P. SIEKEVITZ, AND G. E. PALADE. 1967. molecular. Academic Press, New York, NY. Biogenesis of chloroplast membranes. I. Plastid SCHWEIGER, H. G., H. BRODA, AND D. WOLFF. 1981. dedifferentiation in a dark-grown algal mutant (Chla- Simultaneous recording of two circadian rhythms in mydomonas reinhardtii). J. Cell Biol. 35: 521. an individual cell of Acetabularia. International work- OSAFUME, T., S. MIHARA, E. HASE, AND I. OHKURO. shop on morphogenesis in Acetabularia Proto- 1972. Electron microscope studies on the vegetative plasma 105: 357.

148 SEGEL, G. B., W. SIMON, AND M. A. LICHTMAN. In Les cycles cellulaires et leur blocage chez quelques 1979. Regulation of sodium and potassium transport Protistes. Colloques CNRS, France 240. p. 131. in phytohemagglutin-stimulated human blood lym- WANKA, F., H. F. P. JOOSTEN, AND W. J. DEGRIP. phocytes. J. Clin. Invest. 64: 834. 1970. Composition and synthesis of DNA in syn- SHILO, B., V. G. H. RIDDLE, AND A. B. PARDEE. 1979. chronously growing cells of Chlorella pyrenoidosa. Protein turnover and cell-cycle initiation in yeast. Exp. Arch. Mikrobiol. 75: 25. Cell Res. 123: 221. WEEKS, D. P., AND P. S. COLLIS. 1979. Induction and SOLL, D. R., AND D. R. SONNEBORN. 1972. Zoospore synthesis of tubulin during the cell cycle and life of germination in the water mold. Blastocladiella emer- Chlatnydomonas reinhardtii . Dey. Biol. 69: 400. sonii . IV. Ion control over cell differentiation. J. WEILER, C. S., AND R. W. EPPLEY. 1979. Temporal Cell Sci. 10: 315. pattern of division in the dinoflagellate genus Ceratiutn SRINI VAS, U., AND H. LYMAN. 1980. Photomorphogenic and its application to the determination of growth regulation of chloroplast replication in Euglena rate. J. Exp. Mar. Biol. Ecol. 39: 1. Plant Physiol. 66: 295. WEILER, C. S., AND D. M. KARL. 1979. Diel changes STUTZ, E., AND J. P. VAUDREY. 1971. Ribosomal DNA in phased-dividing cultures of Ceratiwn furca (Dino- satellite of Euglena gracilis chloroplast DNA. FEBS phyceae): nucleotide triphosphates, adenylate energy Lett. 17: 277. charge, cell carbon, and patterns of vertical migration. SWEENEY, B. M., AND J. W. HASTINGS. 1958. Rhythmic J. Phycol. 15: 384. cell divisions in populations of Gonyaulax polyedra WHITNEY, R. B., AND R. M. SUTHERLAND. 1972. J. Protozool. 5: 217. Enhanced uptake of calcium by transforming lympho- SWINTON, D. C., AND P. C. HANAWALT. 1972. In vivo cytes. Cell Immunol. 5: 137. labelling of Chlatnydomonas chloroplast DNA. J. WILSON, R., AND K. S. CHIANG. 1977. Temporal pro- Cell Biol. 54: 592. gramming of chloroplast and cytoplasmic ribosomal VAN BRUNT, J., AND F. M. HAROLD. 1980. Ionic control RNA transcription in the synchronous cell cycles of germination of Blastocladiella emersonii zoospores. of Chlatnydomonas reinhardtii . J. Cell Biol. 72: 470. J. Bacteriol. 130: 249. ZEUTHEN, E. [ed.]. 1964. Synchrony in cell division WANKA, F. 1975. Possible role of the pyrenoid in the and growth. Wiley-Intersciences, New York, Sydney, reproductional phase of the cell cycle of Chlorella . and London.

149 Temporal Patterns of Cell Division in Unicellular Algae

SALLIE W. CHISHOLM Ralph M. Parsons Laboratory Massachusetts Institute of Technology Cambridge, MA 02139

Introduction by manipulation of culture conditions (induction synchrony). Strictly speaking, cell division synchrony The study of the growth of unicellular algae achieved by the selection method is termed "syn- on light/dark cycles is "housed" in three separate chronous," whereas that achieved by induction is subdisciplines of biology with distinctly different called "synchronized" (Abbo and Pardee 1960; James motivations for interest in the subject. These sub- 1966). The distinction between these two terms will disciplines include cell biology (where interest lies not be strictly observed in this review. The use of in the mechanisms of cell cycle control), chrono- induction synchrony for the study of cell cycle events biology (where interest is in revealing the mechanism has often been criticized (see Zeuthen 1974) as of circadian clocks), and ecology (where the goal is distorting the natural timing of events, because cells to understand the temporal interactions between are most often subjected to periodic stresses such as organisms and their environment. The forum for this temperature shocks, amino acid deprivation, and review, and the motivation behind it, involves the thymidine blocks to bring them into synchrony. third goal, but much of the material covered will be Cultures oi unicellular algae grown on 24-h drawn from the cell cycle and clock literature. It is light/dark cycles express a special type of induction hoped that this synthesis will provide a new perspec- synchrony. What distinguishes this group of organ- tive for future studies of the physiological responses isms from all heterotrophic microorganisms is that and adaptations of phytoplankton to the most pre- they have evolved in an environment where the supply dictable selective force in their environment — the of energy for cellular growth is periodic. The regu- 24-h light/dark cycle. larity and inevitability of the natural photocycle must have, through natural selection, imposed a need for temporal order among synthetic and BACKGROUND coordination and reproductive events in the cells. The result, for most The methodologies and the terminology dealing species at least, is a "preferred" alignment (phase with cell division synchrony originated in the cell relationship) between the biosynthetic processes, biology school, where the precise goal is to create the cell division cycle, and the environmental light/ populations of cells that express the behavior of a dark cycle. When cells in a population of unicellular single cell. To this end, culture systems have been algae assume this optimal or preferred relationship, developed in which the generation times of the indi- the cell cycles in the population become temporally vidual cells in populations are aligned as tightly as phased by the photocycle. When the cell division possible to result in simultaneous division. cycles of all the cells in a population are aligned There are several ways this synchrony in cell such that all cells divide each day, the population is division can be achieved (Prescott 1976). "Natural" termed "synchronous." When only a portion of the synchrony (termed "natural" because it is sponta- cells divides each day, but does so in a restricted neous) exists in many cell types from plants (Erikson gate, the population is "phased." This is a natural 1964) and animals (Agrell 1964). The initial cell and fundamental property of growth in phytoplankton divisions of fertilized sea urchin eggs (Mazia and populations, and should not be viewed as induction Dan 1952) are synchronous, for example, as are synchrony in the artificial sense. We can, and must, nuclear divisions in slime molds (Rusch 1969), and assume that the temporal sequence of cell cycle in the anthers of liliaceous plant genera (Hotta et al. processes and cell division patterns expressed by 1966). phytoplankton populations grown on 24-h photocycles In experimentally derived synchrony, simul- are adaptive, and that they reflect the optimized taneity of events is achieved by mechanical selection alignment of cellular processes to a naturally periodic of cells at the same stage of the cell cycle (selection energy supply. The mechanisms by which this align- synchrony) or by bringing all the cells of an asyn- ment is achieved and maintained and the extent to chronous culture to a single point in the cell cycle which it varies among and between various phyto- 150 plankton species are critical aspects of the adaptive known about the G, state in algae, but it probably physiology of phytoplankton (Soumia 1974). exists in dormant cysts and conceivably plays a role in the maintenance of vegetative cells in resource- limited "stationary phase." The Eukaryotic CèII Cycle The traverse time of the cell cycle in genetically identical cells in a uniform environment is not a The generalized cell division cycle for eukaryotic constant. The distribution of generation times in most cells is usually viewed as 4 discrete sequential inter- clonal populations of cells is described by a normal vals: G 1 —S—G,—D (Fig. 1). G I and G2 are the gaps distribution skewed in favor of the shorter generation separating DNA synthesis (S) and cell division (D, times (e.g. Prescott 1959; Cook and Cook 1962). or M for mitosis). S and D periods are well defined This distribution dictates the degree of synchrony and identifiable for most cell types. The Rrecise (the duration of the division burst) expressed by a biochemical events that comprise and define the two population of cells after one generation, beginning gap phases are not at all well understood (Prescott with cells aligned at the beginning of G 1 . It also 1976). dictates the rate at which a synchronized population The time it takes to traverse one cell division will become asynchronous when released from the cycle is the generation time of the cell. The duration synchronizing force (Peterson and Anderson 1964; of the various stages varies with cell type. For exam- Spudich and Koshland 1976). Early attempts to ple, a mammalian cell with a generation time of explain this variability in generation times main- 16 h will typically spend 5 h in G 1 , 7 h in S, 3 h tained that the G I —S—G,—M sequence was suffi- in G2 and 1 h in division (Prescott 1976). The dino- ciently complex that small biochemical variabilities flagellate Amphidinium carteri has a very short along the way would be magnified to result in a rather (usually unmeasurable) G I phase, a 6- to 9-h S phase, large difference in intermitotic times. More recently, and a 5- to 6-h G2 phase (Galleron and Durrand Smith and Martin (1973) attempted to explain the 1979). In Chlorella grown with a 20-h generation distribution with a cell cycle model in which the exit time, the G 1 is about 10 h long, G, is absent, Sis 6h in from G, is probabilistic (the probability of the tran- duration, and the division process takes about 5 h sition from G, to S is a constant, regardless of how (John et al. 1973). long a cell has been in G, ), whereas traversing the The G. state (Fig. 1) is one in which the cell sequence S—G 2—M is deterministic. cycle is arrested in G 1 , i.e. withdrawn from the cell Klevecz (1976) observed that the distribution of division cycle (Prescott 1976). This arrest is reversible generation times in mammalian cell lines appears for most cell types in which the G„ state has been to be "quantized," such that cell generation times identified, which include plant embryos (dormant tend to occur as integer multiples of 3-4 h. To ex- seeds), cells in various tissues (such as kidney, liver, plain this distribution he has proposed a cell cycle and pancreas), and nonrenewing tissues (neurons model in which the G, interval has been replaced and skeletal muscle cells). Little, if anything, is by a loop (G„) whose traverse time is equal to the interval between the quantized peaks in generation times. Here, G„ is viewed as a subcycle from which State the cells may make a gated exit into the remainder 0 of the cell cycle. The exit of a particular cell from G„ is probabilistic in the sense that it depends on environmental factors, but the cell can only exit from G,, after completely traversing the subcycle one or more times. This" gated" entry into the division process is analogous to the gating mechanism pro- posed for the circadian clock, which is known to play a key role in regulating the timing of division in certain phytoplankton species.

The Circadian Clock

The FIG. 1. Diagrammatic view of the cell cycle. The cycle underlying biochemical nature of the endo- begins after division, D, with the G (gap) phase. It proceeds genous biological clock is not known, but this time- through DNA synthesis, S, through another gap (G,), after keeping mechanism appears to be essential to the which division occurs. Some cells can go into a reversible creation and maintenance of temporal order in euka- cell cycle arrest (G o) while in G 1 . (After Prescott 1976.) ryotic organisms and in the ecological systems which

151 they comprise (for books and symposia treating the observed rhythm in these processes. The overt rhythm subject see Aschoff 1965; Biinning 1967; Sweeney (i.e. the one we can measure) is often viewed as the 1969; Menaker 1971; Hastings and Schweiger 1975; "hands of the clock," because it reflects the properties Palmer 1976; Suda et al. 1979; Scheving and Halberg of the clock, but is independent of, and distinct from, 1980; and references therein). Many descriptors the mechanism itself. have been used to refer to "the clock," and they Clock-controlled rhythms are entrainable. That will be used interchangeably here: endogenous clock, is, they will assume the period length of an exogenous biological clock, circadian clock, innate o«Scillator, oscillator (the "Zeitgeber") and maintain a fixed and circadian pacemaker. In all eukaryotic organisms phase relationship to it. The Zeitgeber of interest including unicellular algae, the clock can function to ecologists (which, no doubt, molded the evolution (1) to create internal temporal order in cellular pro- of the clock) is the natural light/dark cycle. It entrains cesses, (2) to properly phase these processes to natural the clock, and in turn the overt rhythm, to a period (external) periodicities, and, in some organisms, (3) length of exactly 24h. It is most important to recognize to measure the passage of time (as in sun compass that the role of the light/dark cycle is not to drive or behavior or photoperiodic time measurement). create rhythmicity but to couple with the biological Various physiological processes that are known to rhythm, i.e. set its phase relative to local time. be under the control of circadian clocks in algae When an organism is released from entrainment, are shown in Table 1. for example by placing it under conditions of constant My purpose here is not to document the evidence darkness or constant dim light, endogenous rhythms that circadian clocks exist, but rather to clearly define will "free-run" with a period length close to 24 h. the general properties of the clock and clock-con- This free-running rhythm will persist indefinitely trolled processes, to provide a framework for the with remarkable precision in individual organisms. interpretation of clock-coupled cell division patterns Although the rhythm may damp out after some time . in unicellular algae. Experimental documentation in populations of unicellular organisms, this must of the clock properties described below will be pre- be recognized as reflecting slight differences in the sented for specific phytoplankton species in sub- free-running period lengths in individual cells. This sequent sections. results in eventual loss of synchrony between the indi- The clock can be viewed as an endogenous vidual cells and should not be interpreted as a loss self-sustaining oscillator whose natural period length of clock control in the individual cells under the is close to (but not equal to) 24 h. The innate "pace- free-running condition. maker" can be coupled to a multitude of physiological Although temperature changes can entrain an processes (see Table 1) and results in an overt or endogenous rhythm and ambient temperatures can

TABLE 1. Circadian rhythms in algae.

Species Rhythm Reference

Acetabularia sp. Photosynthetic capacity Sweeney and Haxo 1961 Chlamydonzonas reinhardtii Cell division Bruce 1970 Phototaxis Gonyaulax polyedra Cell division Sweeney and Hastings 1958 Luminescence Hastings and Sweeney 1958 Glow Hastings 1960 Photosynthetic capacity Hastings et al. 1960 Ceratium furca Cell division Meeson pers. comm. Photosynthetic capacity Prezelin et al. 1977 Euglena gracilis Alanine dehydrogenase activity Sulzman and Edmunds 1972 Cell division Edmunds 1966 Amino acid incorporation Feldman 1968 Settling Terry and Edmunds 1970 Photosynthetic capacity Lonergan and Sargent 1978 Euglena obtusa Vertical migration Palmer and Round 1965 Oedogonium cardiacium Sporulation Bühnemann 1955 Phaeodactyhan tricornutum Photosynthetic capacity Palmer et al. 1964 Hantzschia virgata Vertical migration Palmer and Round 1967

152 affect the amplitude of the rhythm, the period length This discussion would not be complete without of the rhythm is not dependent on temperature. This at least addressing the question of the underlying "temperature-compensation" is one of the most re- physical mechanism of the clock, and whether the markable and essential characteristics of the clock hypothetical mechanism is indeed the same in all mechanism. It is remarkable in that most known organisms. Several recent symposia (e.g. Hastings physiological processes of cells are quite temperature and Schweiger 1976; Suda et al. 1979) have been dependent (the most fundamental being growth rate), devoted to these questions, as is most of the current and it is essential in that it is difficult to conceive research in this field. The general feeling from these of a "chronometer" that could function properly recent accounts is that a common molecular basis unless its time-keeping property were insensitive in all organisms is unlikely — i.e. that the universal to temperature. features of circadian rhythms could easily result An entrained endogenous rhythm can be viewed from convergent evolution under the strong selective as three oscillators, the Zeitgeber, the clock, and pressure of the same external cycles (Hastings 1975). the overt rhythm, which, under the proper conditions Nonetheless, many features of circadian rhythms (see the section that follows for exceptions), are from different organisms may be accommodated by tightly coupled. One of the few ways to dissect the one or parts of three classes of models (Edmunds phase dependence of the three oscillators, and expose 1976). (1) Metabolic feedback loop models in which characteristics of the pacemaker itself, is to allow the oscillatory behavior is maintained via biochemical the rhythm to free run in constant conditions and feedback systems analogous to short-period glycol ytic then subject it to a short stimulus (e.g. light pulse oscillations well documented in yeast (see Chance in constant darkness) at various times during the et al. 1973). (2) "Tape reading" transcription models cycle. Depending on when the pulse is given, a that propose that the distance between genes and, characteristic phase shift (either an advance or a thus, the time it takes to transcribe the DNA "tape," delay) will occur in the overt rhythm. When the could be used to measure biological time (Ehret results of such an experiment are plotted as the and Trucco 1967; Ehret 1974). (3) Membrane difference between the phase of the shifted rhythm models, in which the oscillatory mechanism is derived and its original phase, as a function of the phase of from the interaction between membrane configu- the oscillation when the pulse was given, a phase rations, ion distribution, and membrane bound ion response curve (PRC) results. The PRC can be viewed transport channels (Njus et al. 1974). as a plot of the changing sensitivity of the endo- One of the most promising developments in the genous oscillator to the light phase, which is the efforts to reveal the clock mechanism has been the essential property underlying the mechanism of isolation of clock mutants in Drosophila (Konopka entrainment (detailed discussion and analysis of and Benzer 1971), Neurospora (Feldman and Hoyle PRCs and their utility in revealing properties of the 1973), and Chlamydomonas (Bruce 1972). In Neuro- clock can be found in Enright 1965; Aschoft 1965; spora , for example, several mutants have been iso- Pittendrigh 1975, 1979). lated that have dramatically altered free-running The phase response curve is a "fingerprint" period lengths. Each mutant segregates as a single of pacemaker properties for a given species in that its nuclear gene; thus it is likely that the mutations are in shape is the same regardless of the overt rhythm that the basic timing mechanism itself (Edmunds 1976). was monitored to construct it (Pittendrigh 1975). For example, although the rhythms in cell division, stim- Infradian and Ultradian Growth ulated luminescence, glow, and photosynthesis in Gonyaulax polyedra have peaks at different times The premise of circadian regulation of cellular relative to each other on a light/dark cycle and when events runs into difficulty in populations that are released from entrainment (McMurry and Hastings doubling more than once each day, because in such 1972), the phase response curve for each of these cases the average cell division cycle is shorter than rhythms is identical (Hastings 1960; Sweeney 1969; the circadian period length. Wille and Ehret (1968) McMurry and Hastings 1972; Pittendrigh 1975). The were the first to formalize this problem, and Ehret conservative detailed form of the PRC serves as the and Wille (1970), Ehret (1974), Ehret and Dobra clearest evidence that the clock is independent of the (1977), and Ehret et al. (1977) have developed a processes it drives, and uniquely responsive to the unified theory (complete with an entirely new voca- environment that entrains it (Pittendrigh 1975). The bulary) to describe this apparently schizophrenic fact that any of the overt rhythms can be destroyed by existence in eukaryotic microorganisms. inhibitors (e.g. DCMU on photosynthesis) and the They first define the two fundamental growth coupling between the pacemaker and the other modes. One is the fast exponential or "ultradian" rhythms is not disturbed (Hastings 1960) confirms the growth mode in which the generation times of the autonomy of the clock. cells are shorter than 1 d, and the other is the slow

153 or "infradian" growth mode in which the average 100 cell generation times are much longer than I d. The basic axiom of their formalization is the "eukaryotic- _c 1 circadian principle" (that all eukaryotic organisms ■ have the capacity for circadian timekeeping), and its corollary is the "circadian—infradian rule" (that lnfradian a cell must be in the infradian mode to have circadian F- 50H outputs). This rule was dubbed the G—E—T effect a) GT? 24h CT (Gonya;dax—Euglena—Tetrahymena) because these 7.3 three organisms were thought to best demonstrate C.) phenomenon of light-synchronizable division

the

patterns with circadian period lengths in cultures or growing in infradian mode (Ehret and Wille 1970). It should be noted here that Gonyaufax polyedra has never been reported to exhibit generation times less Time Ultradian

- I CT than 1 d; thus it is not an appropriate species from n GT=CT< 24h io which to launch such a theory. t a Although the formalism and terminology sur- rounding the ultradian/infradian growth mode phe- ner nomenon are perhaps excessive, the concept and Ge distinction are very important, particularly for fast- 10 growing phytoplankton species that must frequently 10 20 30 40 switch between these two modes. What is emphasized Limiting Factor (relative) in recent discussions (Ehret et al. 1977) is that cell cycle time and cell generation time are not equivalent. Although this has been stressed repeatedly in a variety FIG. 2. Diagrammatic view of the relationship between of contexts (e.g. Hartwell 1971, 1974; Gotham cell generation time (GT) and cell cycle time (CT) in 1977; Edmunds 1978; Mitchison 1971, 1973, 1974) infradian and ultradian growth modes controlled by , a it becomes particularly significant in the context of growth-limiting factor. In ultradian mode, CT is a function circadian control. According to the infradian/cir- of GT. They are uncoupled in infradian mode, where CT cadian rule, when cells are growing in the fast (ultra- becomes circadian and GT assumes values that are integral the dian) mode, the cell cycle is tightly coupled to the multiples of 24 h. Only in lethal environments (to left of the broken vertical line) does the period of CT cell division cycle, so that the period lengths of the exceed circadian values. (Modified from Ehret and Dobra two processes are functionally related (Fig. 2) and 1977; Gotham 1977.) they are both temperature dependent. When cell generation times are slowed down to infradian mode by some limiting resource, the cell cycle becomes tions of cells into infradian growth mode and en- uncoupled from the cell division cycle, assumes a training their cell cycles with a Zeitgeber such as circadian period length (Fig. 2), and becomes tem- light/dark cycles or temperature cycles. This en- perature independent (Ehret et al. 1977). At the limit trainment serves to bring all the cell cycles in phase of infradian mode is the "stationaiy-phase" of growth, so the population of cells will express chronotypic which is classically viewed as a population of non- patterns. dividing cells, or cells that are "arrested" in G I (or Chronotypes for Euglena and Gonyaulax have possibly G„). Ehret stresses that such a view, un- been assembled in Fig. 3. Unfortunately, the various fortunately, projects the image that when cells do not activities measured for these two organisms show divide they do not cycle, which is not the case. He very little overlap; thus comparisons are limited. further suggests that the "absolutist" concept of sta- What is important here is to recognize that the chro- tionary-phase cells as nondividing cells should be notype is a conservative property of a species. In replaced by one in which the cells are viewed as having Gonyaufax, for example, the phase relationship a finite (though small) probability of division. between the rhythms in glow, stimulated lumines- The temporal order of cellular processes that cence, photosynthetic capacity, and cell division is cycle independently of the cell division cycle, the preserved throughout a phase shift induced by inter- "circadian chronotype" (Ehret et al. 1977; Ehret and rupting the continuous light regime with a dark inter- Dobra, 1977), can be considered the temporal ana- val (McMurry and Hastings 1972). Of equal impor- logue of the phenotype, and is useful for comparing tance is recognition that many cellular processes physiological aspects of various phyla. The chrono- are rhythmic regardless of whether or not the cell type of a species can be elucidated by forcing popula- division cycle is proceeding toward mitosis.

154 EUGLENA

LDH Activity 11, • GONYAULAX ADH Activity Division G -6- P DH Stimulated Luminescence C A Chl-a, Chl-b, peridinin L-TD L - SD Glow —111— Photosynthetic Capacity Photosynthetic Capacity •

DN A Ru BP - Carboxylase Chlorophyll/cell • -0- Cell Division • PO4 Uptake • ---Photo t axis

Light Dark Light Dark FIG. 3. Chronotypes for Euglena gracilis and Gonyaulax polyedra . Solid circles mark the peak in activity and lines mark the interval of the peak in the rhythm. All rhythms except DNA synthesis and cell division were measured in infradian cells. Euglena ADH = alanine dehydrogenase activity Sulzman and Edmunds 1972 LDH = lactic dehydrogenase activity Edmunds 1974 L-TD = L-threonine deaminase activity Edmunds 1974 L-SD = L-serene deaminase activity Edmunds 1974 G-6-P = glucose-6-phosphate dehydrogenase activity Lonergan and Sargent 1978 CA = carbonic anhydrase activity Lonergan and Sargent 1978 Photoaxis Bruce and Pittendrigh 1956 PO, uptake Chisholm and Stross 1976 Chlorophyll/cell Laval-Martin et al. 1979 Photosynthetic capacity Laval-Martin et al. 1979 DNA synthesis Edmunds 1965b Cell division Cook 1961a Gonyaulax Stimulated luminescence Hastings and Sweeney 1958 Glow Hastings 1960 Photosynthetic capacity Hastings et al. 1960 Cell division Sweeney and Hastings 1958 Chlorophyll a, c, peridinin Prezelin and Sweeney 1977 RuBP carboxylase activity Bush and Sweeney 1972

The question of the behavior of the cell cycle times that can result from such control must then when the cell division cycle is in ultradian mode be "quantized" and (assuming G„ < 12 h) allow remains. Ehret and Dobra (1977) claim that the cell for ultradian or infradian growth. In mammalian cycle (and the circadian clock) assumes the frequency cells, where Gi„ appears to have about a 4-h duration, of the cell division cycle (Fig. 2), which represents various enzyme activities have been observed to a fundamental change in temporal order within the oscillate at a corresponding frequency regardless of cell and reflects very "unclocklike" properties. the GT values of the population, and the rhythms Klevecz (1976) offers an alternative, with a model are temperature compensated (Klevecz 1969, 1975). that allows for strict temporal control in ultradian From this, the inference is that 0„ is a clock of sorts, and infradian growth modes. As discussed previously, but with a shorter period length than the circadian the G„ subcycle in his model has a traverse time clock. The latter could be a special case (G„ = 24 h) that is necessarily shorter than the generation time or the circadian frequency might reflect several passes and must be cycled through at least twice before through the high frequency G„ loop. This, however, mitosis can occur. The possible values of generation is all in the realm of speculation.

155 Euglena: An Example of Clock-Coupled When cultured in continuous light under opti- Division mum conditions (25°C; 7000 lx) Euglena grows with an average generation time of - 11 h (Cook 1963; Of all autotrophic unicellular algae, Euglena Edmunds 1965a). In an exponentially growing culture is the best understood in terms of the characteristics with an average generation time of 13.4 h, the genera- and mechanisms of phasing the cell cycle to light/dark tion times of individual cells range from 9.5 to 23.7 cycles. Work on synchrony in this genus was pioneered (Cook and Cook 1962). The minimum number of by Leedale (1959), Cook (1961a, b, 1963, 1971), hours a cell must be exposed to saturating light inten- Cook and Cook (1962), and Cook and James (1960) sities to undergo division has been estimated to range and systematically pursued by Edmunds and co- from 10h (Edmunds and Funch 1969b) to 14h (Padilla workers (see Edmunds 1974, 1978) who have firmly and Cook 1964; Edmunds 1965a). established that the cell cycle in this species is coupled to a circadian clock. Although this genus is not ideal When grown on an appropriate light/dark cycle, for ecological extrapolation, it is presented here as a populations of Euglena become synchronized so that "model system" exemplifying the basic characteristics all the cells divide during a restricted time interval of clock-coupled cell cycles in unicellular algae. These (the division gate) during the dark period (Table 2). characteristics should, in principle, be manifested in For example, on L:D 10:14 (Fig. 4a) the division all species with the capacity for circadian control of burst begins 12-13 h after the onset of the light cell division. period, continues for about 10-11 h, and results

TABLE 2. Summary of the growth and periodicity parameters for Euglena gracilis Z grown on various photoperiodic regimes.

Beginning of division L:D burst Period Photoperiodic photo- Average Equivalent" (hours after length regime Light I period step size photoperiod lights on) (h) Reference

24-h periods 3500 lx 16:8 2.03 11-14 h 24 Edmunds I965a 800 lx 1.14 14 h 24 Edmunds 1966 3500 lx 14:10 2.02 13-14 h 24 Edmunds 1966 8000 lx 12:12 2.00 24 Edmunds and Funch 1969a 8000 lx 10: 14 1.97 12-13 h 24 Edmunds and Funch 1969b 8000 lx 16 1.68 12 h 24 Edmunds and Funch 1969b Skeleton 8000 lx 4,4,4:12 1.50 8:16 12h 24 Edmunds and Funch 1969a photo- 8000 lx 3,6,3:12 1.31 6:18 13 h 24 Edmunds and Funch I969a periods Constant 8000 lx 10:10 1.81 20 Edmunds and Funch 1969a L:D ratio 8000 lx 8:8 2.02 33 8000 lx 6:6 (Random division) 8000 lx 5:5 2.4 32 8000 lx >2 30-40 High 8000 lx 0.25:0.50 1.68 8.16 30 Edmunds and Funch 1969a frequency 8000 lx 0.50:1 1.22 8:16 26 8000 lx 1:2 1.48 8:16 26 8000 lx 1:3 1.20 6:18 24 8000 lx 1.47 8:16 27 8000 lx 2:6 1.41 6:18 28 Random 8000 lx Random 1.67 8:16 27 Edmunds and Punch 1969a Continuous 800 lx Continuous 1.17 24.2 Edmunds 1966 low light (following 12:12) Low 8000 lx 12:24 1.74 36 Edmunds 1971 frequency 8000 lx 12 :36 1.79 48 Edmunds 1971

"Total number of hours of light and dark per 24-h period.

156 RIM I I A B

LD.I0 14 L /m

L 4 LS I 0 /m CEL I 04 OF CELLS OF BER 3 1 I 0 NUM

BER = 1.97 NUM

10 3 4 5 6 2 3 4 5 TIME (DAYS) TIME (DAYS)

11•13 ••MBIZ 1•11 1117 • Ma L

D:12,14E-- LID:4,4,4:12 /m LS CEL L 4

/m Io

I 0 OF S L

CEL SS I.50 NUMBER OF

BER NUM 3 3 I 0 I 0 I 2 3 4 5 6 7 I 2 3 4 5 6 7 TIME (DAYS) TIME (DAYS)

FIG. 4. Patterns of entrained cell division in Euglena gracilis grown on various photocycles. The average step size (Ts-) is the relative increase in the number of cells during each 24-h period. In all photocycles shown, the period length of the division rhythm is 24 h. (Adapted from Edmunds 1978; see also Table 2.) in a doubling of cell numbers (step size = 1.97). gate (6-10 h) but all the cells do not divide every The distribution of generation times reflected by the 24 h (step size = 1.68). Here, the doubling time of duration of the division burst is consistent with the the population is 36 h as is the average generation distribution of generation times observed in exponen- time of the cells. Because of the restriction imposed tially growing cultures under optimal conditions by the division gate, however, the generation times (Cook and Cook 1962). of individual cells must be discontinuously distributed If the photoperiod is shortened to L:D 8:16 with modes at 24 and 48 h (Edmunds 1978). This (Fig. 4b) cell division in the population becomes type of population response is suggestive of clock- phased so that division is still restricted to the division controlled phasing.

157 Lengthening the photoperiod beyond 10 h does in the division rhythm (Table 2). On L:D 10:10, not significantly affect the synchrony in the popu- the populations entrain directly to the 20-h period lation, provided light intensities are reduced for the length of the light/dark cycle. This appears to be the longer photoperiods (Table 2). Note that L:D 10:14 lower limit of entrainment for this organism as and 12:12 (8000 lx) and L:D 14:10 and 16:8 further reduction of the entraining period to L:D (3500 lx) yield approximately the same synchrony 8:8 results in an uncoupled division rhythm with an patterns, with step sizes close to 2 and the onset of average period length of 33 h. The fact that this period division beginning between 11 and 14 h after the length is about twice the duration of the entraining onset of the light period. If photoperiod is lengthened cycle suggests that "frequency demultiplication" has beyond 12 h (e.g. L:D 14:10, 16:8, 18:10) while occurred. Curiously enough, all attempts to entrain maintaining high light intensities (8000 lx), the Euglena to L:D 6:6 have failed (Edmunds and Punch step size increases above 2 indicating that some cells 1969b). On this regime, the population ignores the are dividing twice during each division burst (Ed- L:D cycle completely and increases exponentially munds and Funch 1969b). Although the division with a generation time of 28 h. Higher frequency gate is apparently preserved to some extent in this cycles (L: D 5:5 and 4:4) result in long period division ultradian growth mode, there is a tendency towards bursts with step sizes greater than 2 (Table 2). The asynchrony, which becomes complete in LL. pattern on L:D 4:4 is particularly interesting in that Continuous illumination during the photoperiod the period lengths are reportedly cyclical (Edmunds is not essential to maintain the phasing of cell division and Funch 1969b). Duplicate experiments have in Euglena (Fig. 4c, d). When the photoperiod of shown successive periods of 28, 38, 40, and 32 h an L:D 12:12 regime is interrupted by 4 (Fig. 4c) in one culture and 30, 36, 40, and 30 h in another. or 6 (L:D 3,6,3:12) h of darkness, gated cell divi- This behavior, in conjunction with step sizes greater sion continues, although the step size is reduced than 2, could be interpreted as rhythmicity in ultradian because of the reduced energy input (Table 2). growth mode, and is reminiscent of patterns seen in All light/dark regimes discussed thus far have marine diatoms (see below). resulted in synchronized or phased division patterns The ultimate "exotic" light/dark cycle em- that have period lengths equal to that of the entraining ployed by Edmunds and Funch (1969a, b) was a cycle, i. e. 24 h. To reveal the free-running period "random" one, in which (i) the total duration of of the putative clock mechanism, many light regimes light was 8 h per 24 h, (ii) the total duration of dark- may be employed. The most straightforward demon- ness during that inerval was 16h, and (iii) the lengths stration that the cell cycle in Euglena can be coupled of the light periods were between 0.25 and 1.0 h, to an endogenous clock is to entrain the population and the dark periods between 0.5 and 1.5 h. The to L:D 12:12 and then place it in continuous dim division pattern resulting from this regime appears to (800 lx) light (Edmunds 1966). Under such condi- be truly "free-running" with an average period length tions rhythmic cell division will persist for at least of 27.5 h and step size of 1.67 (Fig. 51)). 10 d (for weeks in continuous culture) with a period Various low-frequency light/dark regimes have length between bursts of 24.2 h, and a step size of also been employed to examine the mechanism of about 1.17 (Fig. 5a). Although the average individual synchronization in Euglena (Edmunds 1971). When cell cycle in this experiment has a duration of 134 h, grown on L:D 12:24 or L:D 12:36, the cells directly a population rhythmicity is expressed with a period entrain to the period of the light/dark cycle (Table 2). length of 24.2 h, reflecting the gated division time. Step sizes of 2 are not achieved (even though 12 h Various "exotic" (Edmunds 1974) photocycles light should be sufficient) presumably because of also reveal the underlying clock mechanism. High- the metabolic "tax" imposed during the extended frequency light/dark cycles with periods that are dark period. integral submultiples of 24 h all invoke free-running It is somewhat puzzling that the step sizes cell division patterns with period lengths ranging resulting from the various photoperiods show little from 24 to 30 h and step sizes ranging from 1.20 to relationship to the total amount of light seen by the 1.68 (Table 2; Fig. 4d). It should be noted that popu- cells during each 24-h period (Table 2). Note, for lations taken from continuous bright light (7500 example, that even though L:D 1/4:1/2 and 1/2:1 lx) and placed in similar high-frequency light/dark both receive the equivalent of an 8:16 photocycle, cycles will express the same rhythmicity as those the step sizes are dramatically different. Similarly, taken from L:D 12:12; thus prior entrainment to a cells on L:D 1:3 and 2:6 receive 6h of light per 24 h 24-h period is not necessary for the free-running yet the population receiving 2-h "parcels" of light rhythm to be expressed. seems to grow more efficiently. In fact this popu- Light/dark regimes of somewhat longer period lation grows faster than one maintained on L:D lengths that are not integral submultiples of 24 h 1/2:1 which receives 2-h more of light per 24-h (with an L:D ratio of 1) invoke a variety of responses period (Table 2). In contrast, the "random" regime,

158 L:D 1/4:1/2, more why cellscan step sizes amount dark period lengthis FIG. 0 -»-LbD cycle therhythm 5. CYCLE efficiently thanothersisnotatall 5 I ■ (3500 of Free-running, clock-coupled,cell (albeit IN , light 14,10 CELL INCREASE Id 24.0 "process" certain j and L:D8:16, and different periodlengths).Precisely NUMBER - NUM B ER OF CELLS /m L v/.e h, - dark has step sizes 104 I 0 . / ( 7.9) 3 a per 24h, LL period length ■ LI>10,14 d / // (8001x) are ' (6.5 2 all with light/dark regimes extremely small, 1 yielded identical DAYS AFTER division of 9 j 1 1 '14.0

clear. 3 27.5 2 the I rhythms h andanaverage TIME (DAYS) / 10 same a ■ 1, ;14.9 "Random" and theaverage 4 ' 3 // I 1 in division must the grown II ' 116.6' Euglena gracilis. A' LD 4 INOCULATION mechanism cycle The data involve Euglena step size generation timeis 12 to 109 Amme7 Amgmv 5 a certain an presented thus of strongly 13 of (A) In endogenous clockthatrestrictscell 2 12.8 7 1.67. 6 synchronization "gate" continuous dimlight, (Adapted fromEdmunds 5.5 .G.7.• 134hF5 B support thecontention 11.9

d. in the far forautotrophically (13) On , 12.6 9 entraining >2 a "random" in days) this species 17.3! % the average to IA ze//4 cycle. 1978.)

light/ that 159 Ice

It would be difficult to invoke light- or dark- The rhythm in cell division in G. polyedra is inhibited cell cycle block points (see Spudich and known to be endogenous (Sweeney and Hastings Sager 1980 and the section on green algae below) 1958). As mentioned above, cell division in this to explain all the observed patterns, particularly the species is restricted to a 5-h period during the light/ free-running rhythm. The interpretation of these types dark transition. When an entrained culture is trans- of results, however, is complicated by the fact that ferred to continuous bright light (10 000 lx) the the entraining agent (light) is also an energy source division phasing is lost within 4-6 d. In continuous for the cells. In this regard, it is significant that dim light (1000 lx), however, division in the popula- various mutant strains of Euglena that are either tion remains phased for at least 14 d. Under these photosyntheticly impaired (Jarrett and Edmunds conditions, the interval between division bursts (the 1970) or permanently bleached (Mitchell 1971) be- free-running period) is close to 24 h, and does not have very similarly to the autotrophic strains. Further- change significantly with temperature (i.e. it is more, these mutants provide a dramatic demonstration temperature compensated). of the Circadian-Infradian Rule (Ehret and Wille The only dinoflagellate species that has been 1970). When the P Zul mutant (blocked in the elec- studied systematically with regard to division timing tron transport chain between photosystem I and II) on different photoperiods and temperature regimes is grown in organic medium on L:D 10:14 at 25°C is Ceratiwn furca (Weiler and Eppley 1979). The it grows exponentially with a generation time of timing of the onset of division seems to be independent 10 h, clearly in ultradian growth mode (Jarrett arid of temperature (on L:D 12:12) and photoperiod (at Edmunds 1970). If one then lowers the growth tempe- 20°C), and always occurs between 9 and 11 h after the rature to 19°C, thus reducing the mean generation beginning of the dark period (Table 3). This is in direct time to 24 h, the culture becomes synchronized to contrast with the division pattern in Euglena gracilis , the light/dark cycle with a period length of 24 h and where the onset of division seems to maintain a fixed a step size of 1.96. The rhythm in division will phase relationship relative to the onset of the light persist for several weeks in continuous darkness or period (see Table 2). Although there is good evidence continuous bright light (9000 lx) (Edmunds et al. that in Ceratium, as in Euglena, division is under 1971) and a single L:D transition is sufficient to clock control (Meeson personal communication), it is entrain an exponential culture and elicit a free-running clear that the phase angles between the clock, the cell rhythm (Edmunds 1974). This is a clear demonstra- cycle, and the light/dark cycle are different in these tion of clock control of the timing of cell division, two species. and an uncoupling of the clock in ultradian growth. It is ot extreme importance to recognize that the differences in the division timing between the Dinoflagellates various species could be influenced by experimental conditions and what stage of division is being moni- for example, has shown that in Cera- Because of the inherently slow growth rates of tored. Meeson, although the duration of the "early divi- dinoflagellates and the ease with which division than furca , stages are independent of temperature and light, stages can be recognized, division rhythms in this sion" division" stages are not. group are usually examined by monitoring the fre- the duration of the "late stage, which consists of the interval quency of paired cells or paired nuclei in light/dark The early division the oblique cleavage phased cultures. Strict phasing has been observed between the time at which the middle of cytokinesis, in all species examined. That is, there is always furrow becomes visible to of conditions, whereas an interval during which dividing cells cannot be lasts for about 2 h regardless division stage, from the last stages of cyto- detected or there is no increase in cell numbers the late h on a light/ (Table 3). kinesis to cell separation, can last 12 h in continuous light. The time at which division occurs relative to dark cycle and up to 48 the light/dark cycle varies between species. The most common trend, documented in cultures of Gonyaulax polyedra , G. sphaeroidea , Ceratium Green Algae furca Cachonina , and Prorocentrwn sp., is maximum division frequencies appearing near the CHLORELLA dark to light transition (Table 3). In several species, however, the division process spans the dark period Studies of cell synchrony and the cell cycle of (Amphidinium carteri , Gymnodinium splendens , and Chlore/la have been dominated over the years by Scrippsiella trochoidea), or begins in the middle of Japanese and German researchers (e.g. Tamiya et the light period and is completed by the onset of al. 1953; Lorenzen 1957; Pirson and Lorenzen 1958). darkness (Scrippsiella sweeneyi and Prorocentrwn Extensive reviews have been written on the subject micans). (Tamiya 1964; John et al. 1973; Schmidt 1966) and

160 TABLE 3. Division patterns of Dinoflagellate species cultured on light/dark cycles. The intervals during which division was occurring (subjectively determined) are indicated with an x. Each x is 1 h and the blackened interval represents the dark period of a 24-h light/dark cycle.

Light Variable Species Reference Temp., °C intensity measured Division pattern

Amphidinium carteri Chisholm and Brand 1980 16 40 ,u,E•m-e•s-1 cell no. XXXXXX X XXXXX (Amphi)

Gynmodinium splendens Hastings and Sweeney 1964 % paired X XXXXXXXXX XX

Scrippsiella trochoida Nelson and Brand 1979 20 5 x 10-2 ly•min-1 cell no. XXXXXXXXXX (Peri)

Scrippsiella sweeneyi Sweeney and Hastings 1964 — cell no. XXXX

Prorocentrutn tnicans Hastings and Sweeney 1964 — % paired XXXXXXXXXX

Gonyaulax polyedra Sweeney and Hastings 1958 20 1400 ft can. cell no. XXXXXX

% paired XX XXX%

Sweeney and Hastings 1964 cell no. XX XXXXXX

% paired = XXXXXX

Gonyaulax sphaeroidia Hastings and Sweeney 1964 — % paired

Cachonina niei Loeblich 1977 24 1000 ft can. cell no. XXX%

Prorocentrum sp. Chisholm and Brand 1980 16 40 pE•nr2 • s-2 cell no. XXXXXX (Exuv)

Ceratium furca Weiler and Eppley 1979 15 3 x 10" quanta•cm -2 •s-1 % paired" "`" XX

10 3 x 10" quanta•cm-2 •s-1 XX XX

25 3 x 10" quanta • cm-2 • s-1

20 3 x 10 1" quanta•cm-2 •s- I XXX

20 3 x 10" quanta•cm-2 •s-1 XXX

20 3 x 10" quanta•cm-2 •s-1 XX XX

20 3 x 10 1" quanta.cm-2 •s-1 XXXX

20 3 x 10" quanta•cm -2 •s-1 XXX

"Reflects interval when division frequency was 1/2 maximal. many reviews of synchrony in algae in general are illumination or darkness after certain preconditioning. dominated by examples from the Chlorella literature They have relied extensively on selection synchrony (e.g. Tamiya 1966; Lorenzen 1970; Pirson and for this dissection. The detailed description of the Lorenzen 1966). Given these extensive reviews, my cell cycle that has resulted from their work is struc- purpose here is simply to examine certain aspects of tured around a distinction between two types of cells: the synchrony in Chlorella as they relate to and dark cells (D) and light cells (L) (Fig. 6). The former compare with those of other key species analyzed are small, highly pigmented cells with relatively high thus far. quantum efficiency of photosynthesis and low res- The approach of the Tamiya school (see e.g. piratory activity, which turn into light cells when Tamiya 1964; Morimura 1959; Hase et al. 1957) illuminated. These light cells in turn ripen and ulti- has been to dissect the cell cycle by classifyirig the mately divide into a number of autospores, which various stages according to their responsiveness to become dark cells.

161 the same number of daughter cells at the various temperatures. The latter response is not what we would expect for a clock-controlled species but it must be stressed that these experiments were done by monitoring the synchronous development of D„ cells isolated by fractional centrifugation. The popu- lation was not entrained by a light/dark cycle. The results of the German workers, who tended to use programmed light/dark cycles to induce syn- chronous division, show characteristics that are in keeping with a clock-coupled mechanism for phasing. As early as 1957, Lorenzen noted thai when Chlorella pyrenoidosa was grown on various light/dark cycles, the onset of division appeared to occur a fixed number of hours (16-20) after the onset of the light period, regardless of the L:D regime (Table 4). Pirson and Lorenzen (1958) extended this observation to include changes in temperature and light intensity as well as photoperiod, and noted again that although growth rate (Schub number) was dependent on these envi- FIG. 6. The Ch/are/la cell cycle. The cycle begins with ronmental parameters the onset of division was always "nascent dark cells" (D„) which become "active dark gated to occur 20 h after the onset of the light period. cells" (D„) when illuminated, and the growth stage (D„ the time that the onset of to L2 ) is initiated. D — L marks the transition to the "ri- These authors suggested at pening" stage. "Unripened light cells" (L,) are large but the light period acted as a Zeitgeber to some "endo- cannot divide when placed in the dark; ripened light cells genous machinery" (Tamiya 1964). Subsequently, (L„) can divide completely when placed in the dark; and Hesse (1971) has shown conclusively that the cell L.1 cells are just about to divide. The dark arrows indicate cycle in Chlorella can be clock controlled. those transitions that do not require light. (After Tamiya et al. 1961.) CHLAMYDOMONAS There are many similarities between Chlatny- The division process is not only independent domonas and Chlore/la in terms of their responses of light in Chlore/la but is actually retarded by it. If to light/dark regimes. Chlatnydomonas exhibits a L: D synchronized culture is released into continuous some of the tightest synchrony patterns observed for light, the division burst that would have occurred unicellular algae. Zoospore production is restricted takes 8-9 h in the light rather than the usual 3- to 4-h to a 2-h gate at the end of the dark period (Bernstein span observed in the dark (Sorokin and Krauss 1959). 1960, 1964) on both L:D 12:12 and 10:14 (Mihara Thus, in cultures grown in continuous light, D„ and Hase 1971). Thus, this genus is ideal for cell cells (see Fig. 6) do not emerge from the mother cells cycle studies, and the conditions for optimizing as such, but remain inside the mother cell and emerge synchrony of all cell cycle parameters have been as D„ or D — L cells at some later time (Morimura extensively examined (Lien and Knutsen 1979). 1959). Although the growth process continues while Mihara and Hase (1971) have shown that when the D cells are inside the mother cell, it seems likely Chlatnydomonas reinhardtii (IAM C-9) is grown on that the growth efficiency is not the same as that of L:D 12:12 and released into continuous bright light illuminated D„ cells. This explains the higher growth (5000 lx), the liberation of zoospores occurs at the rates obtained in light/dark synchronized cultures same time as in the light/dark cycle, although other (Sorokin and Krauss 1959). cellular events such as nuclear and chloroplast divi- According to Morimura (1959) the time required sion, and cytokinesis tend to lose synchrony (see for the growth and ripening stages of the cell cycle also Bruce 1970). During the second cycle of con- (Fig. 6) are independent of light intensity, but de- tinuous light zoospore liberation begins earlier pendent on temperature at saturating light intensities. than usual and synchrony begins to deteriorate. These In contrast, the number of autospores produced is authors have shown that zoospore liberation is timed independent of temperature but increases with in- relative to the beginning of the light period, and that creasing light intensity. In other words, the cells dim light (700 lx) is not sufficient to set the timing. respond to reduced light intensities by producing Furthermore, as in Chlore/la (Morimura 1959), the fewer daughter cells at the same rate, whereas they timing of zoospore liberation is not affected by light respond to lowered temperatures by extending the intensity on L:D 12:12 (although the number of growth and ripening periods appropriately to produce zoospores is) but is quite delayed at reduced tempe-

162 TABLE 4. The timing of cell division in Chlorella pyrenoidosa grown on various light/dark cycles, at 22°C and 6000 lx. Intervals when division was occurring are marked with an x; darkness is indicated by a solid bar. (From Lorenzen 1957.)

L:D 8:8 XXXXXXXX XXXXXXXX XXXXXXXX

10.5:7 XXXXXXX XXXXXXX XXXXXXX , XXXXXXX XXX

12:12 XXXX XXXX XXXX XXXXXXXX XXXXXXXX

16:8 XXXX XXXX XXXX XXXXXXXX XXXXXXXX

16:16 XXXXXXXX XXXXXXXX XXXX

24:12 XXXXXXXXXXXX XXXXXXXXXXXX XX

24:24 XXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXX 1 1 1 1 1 1 1 1 1 1 1 1 24 18 12 6 0 12 18 24 30 36 42 48 54 (h)

ratures in LL. The temperature effect on the timing We are left then with two conflicting views of zoospore liberation in cells grown on L:D 12:12, of the mechanism of synchronization in Chlamy- however, is minor, with a 2-h shift in timing be- domonas , which have been recently recast by Spudich tween 25 and 30°C. It is noteworthy that the develop- and Sager (1980). The first view maintains that the ment and release of zoospores in Chlamydomonas synchrony results from light/dark cycles driving the is not inhibited (or delayed) by light as is reportedly cell cycle, i.e. from dark or light inhibition of certain the case for Chlorella (Sorokin and Krauss 1959). metabolic steps (Bernstein 1964). The second view In direct contrast with the results of Mihara and proposes that the cell cycle is clock coupled and Hase (1971), Bruce (1970) has shown that the rhythm entrained by the light/dark cycle (Bruce 1970). in zoospore release in C. reinhardtii does persist Spudich and Sager (1980) have identified two tran- for at least 14 d in continuous light, strongly sug- sition points in the Chlamydomonas cell cycle (not gesting that the timing of zoospore release is regulated unlike the D and L cell transition in Chlorella), by clock coupling. Bruce used a different strain which they believe explain synchronization and thus (RS strain) of this species and used a continuous support the first hypothesis. Both of these points culture apparatus rather than the daily dilution tech- are in G 1 . At the primary arrest point A, the pro- nique of Mihara and Hase (1971). Temperatures and gression of the cell cycle becomes light dependent light intensities used in the two studies were compa- until the second point, T, is reached and the comple- rable, but the growth rates of the cultures in Bruce's tion of the cycle becomes independent of light. If experiments were much slower (mean population exponentially growing populations of cells are placed doubling times of 30-44 h vs. 12 h for Mihara and in the dark (or subjected to the photosystem II in- Hase). These slow growth rates were attributed to hibitor DCMU) before they reach point T, they light limitation because of self-shading, but the cell remain viable but do not progress through their cycles. densities in Bruce's experiments were of the same Under the condition of the experiments, 6 h of light order of magnitude as those reported by Mihara exposure is required for the cell to progress from and Hase. Whatever the reason for the slower growth A to T. Regardless of how long cells are arrested rates in Bruce' s experiments , the evidence is strong between A and T by a dark block (at least to 48 h), for endogenous control and it is tempting to argue the cells always take the same amount of time in that the infradian growth mode promoted the ex- the light to complete division. Thus, the cells are pression of the circadian rhythm in zoospore release. all blocked at the same point and are not cycling Whether or not the growth rates in Mihara and Hase's in the dark. Although cells in dark arrest are viable experiments can be called ultradian, however, is and metabolically active, they do not mobilize avai- not clear, as distinction may not be appropriate lable carbohydrate reserves to provide energy to for cells that divide into more than two daughter complete the cycle. These reserves are not utilized cells. This whole issue needs clarification. until after the cycle passes the transition point, T.

163 At this time endogenous respiration shows a sharp controlled model. Spudich and Sager (1980) offer an increase, presumably to provide the energy necessary interesting discussion of the alternative hypotheses to progress through the cell division cycle from T regarding mechanisms of synchrony, and conclude to A. that light/dark synchrony of both photosynthetic and nonphotosynthetic algae (e.g. photosynthetic The results of the Spudich and Sager (1980) mutants of Euglena) results from forced periodicity study show that if one grows Chlamydoinonas on in the availability of high-energy metabolites from L:D 6:12 or on L:D 6:48 one gets exactly the same electron transport. Whether this proximate mecha- division pattern. As shown in Table 2, this is not the nism would exclude the possibility of the ultimate case for Euglena. When the minimum light require- mechanism being the clock is not entirely clear. ment for division is 12 h, and Euglena is grown on L:13 12:12, 12:34, 12:36, the step size is smaller SCENEDESMUS AND DUNALIELLA for the extended dark periods (Edmunds 1971; Table 2), suggesting that reserves are being metabolized Fairly tight synchrony can be achieved in cul- for maintenance requirements. tures of Scenedesmus obliquus and S. quadricauda Clearly, the results of Spudich and Sager (1980) grown on light/dark cycles (Table 5), with division are consistent with a synchronization mechanism restricted to a short interval in the dark period. Fur- that operates via a dark arrest point. When exposed thermore, in a pleomorphic strain of S. quadricauda, to darkness, all cells between A and T do not progress photoperiod is a significant determinant of the through their cycle, whereas all those that are past T morphological form of the cells (Steenbergen 1975, progress to A. Thus , within a few light/dark cycles 1978). Under the appropriate conditions of tem- the cell cycles of all cells will be aligned in phase perature and nutrients, unicell-yielding populations with the light/dark cycle. Whether or not the data can be achieved in L:D 14:10, but if the photoperiod are inconsistent with a clock-entrained model, how- is shortened (e.g. L:D 3:21) the coenobial morph ever, is not clear. The clock can be reset on exposure predominates. to bright light after periods of darkness (see Aschoff Unlike other green algae, division in L:D syn- 1965; Palmer 1976); thus, the apparent noncycling chronized Dunaliella tertiolecta does not appear to in the dark followed by simultaneous division in the be totally restricted to the dark period and the division light in Chlatnydomonas is not problematical. The gate can be fairly broad, especially for short and long equivalent actions of darkness and DCMU, however, photoperiods (Table 5). The synchrony can be en- are more difficult to explain in the context of a clock- hanced somewhat and shifted to later in the dark

TABLE 5. Patterns of cell division in various species of green algae grown on light/dark cycles. The interval where division occurs is marked by x; the dark period is indicated by the solid bar. Total interval is 24h and each x represents 1 h.

Species Reference Temp., "C Light Division pattern

XXXX■ Scenedesmus obliquus Lafeber and Steenbergen 1967 30 —15 000 lx MIMIIIM1111 1

XXXX Scenedesnuts quadricauda Komàrek and Simmer 1965

XXXXXXXXXX Dunaliella tertiolecta Eppley and Coatsworth 1966 20 0.05 cal • cm" • min"

Eppley and Coatsworth 1966 20 0.05 cal • cm" • min- ' XXXXXXXXXX

XXXXXXXX Eppley and Coatsworth 1966 20 0.05 cal. cm-2 • min- '

XXXXXXXXXXXX Eppley and Coatsworth 1966 20 0.05 cal • cm-2 • min- '

Eppley and Coatsworth 1966 20 0.05 cal . cm' • min- ' XXXXXX XXXXX XXXXX

Dunaliella tertiolecta Wegmann and Metzner 1971 40/20 8000 lx XXXXXX

X Chlatnydomonas reinhardtii Mihara and Hase 1971 25 5000 lx X

XX Chlainydonionas inoewusii Bernstein 1960 25 8000 lx

Chlorella pyrenoidosa Lorenzen 1957 22 6000 lx XX XXXXX

164 period, if a temperature cycle (L:D 40°C:20°C) is served a similar phenomenon for Hymenomonas superimposed on the L:D cycle (Wegmann and carterae grown in a cyclostat at 20°C on L:D 14:10 Metzner 1971). This is attributed to the fact that the (Chisholm and Costello 1980). At an average popula- temperature optimum for photosynthesis is 35°C in tion growth rate of 1.5 doublings/d there is a major this species whereas the optimum for growth (in- division burst at the beginning of the dark period, crease in population cell number) is 20°C. Cell followed by a second pealc at the dark/light transition. division will not occur at 40°C. It is logical and necessary that such patterns should emerge under conditions of rapid growth, and it is Other Taxonomie Groups interesting that they have never been systematically studied in terms of the distribution of generation An overview of the division patterns in all other times in the population and in the context of cell cycle groups of phytoplankton that have been studied phasing. This is undoubtedly because most research (Table 6) reveals quite clearly that there is a preference on algae grown on light/dark cycles has been directed for nighttime division. The exception in this group towards achieving and understanding the mechanisms is Platymonas striata, but the timing of division of synchrony, rather than understanding how popula- (as determined by the time course of cell numbers) tions respond to a 24-h light/dark cycle in general. in this species is misleading in regard to the timing A mix of goals and approaches would be helpful of cytokinesis (Ricketts 1979). It appears that after here. the formation of daughter cells inside the parental Cyanobacteria theca, which normally occurs in the light period, there must be a maturation period of about 6 h before The cyanobacterium Anacystis nidulans has been they can be released. The escape of the daughter cells synchronized and studied intensively using a variety from the theca is triggered by light; thus tight syn- of induction synchrony methods such as chrony results in which increases in cell number occur L,I ,„,:D 2.5:0.5:5 and alternating temperature cycles early in the light period. Only under conditions of low (26°C:32°C, 8:6 h) (Herdman et al. 1970; Lorenzen temperature, or high growth rates induced by long and Kaushik 1976; Venkataraman and Lorenzen 1969; photoperiods, are any of the daughter cells released in Csatorday and Horvath 1977; Lorenzen and Venkata- the dark (Table 6). raman 1969), but the division patterns of cyanobac- Of the group of species presented in Table 6, teria have rarely been examined on 24-h photocycles. only Emeliania huxleyi has been systematically In search of circadian rhythmicity in prokaryotic studied in terms of its response to photoperiod organisms, Taylor (1979) grew several species of (Paasche 1967). Although it is difficult to separate cyanobacteria on L:D 12:12 at 20°C, which produced the effects of the variable light intensities, it appears infradian growth rates. Unfortunately, however, he that the timing of division is determined primarily did not measure cell number, but rather optical density by the onset of the previous photoperiod, which is of the cultures; thus division timing cannot be deduced similar to the case for Euglena, but opposite to the from these data. Taylor concluded from a variety of pattern for Ceratium (see Weiler and Eppley 1979 measurements, however, that there was no evidence of and discussions above). A more distinctive trend is a circadian clock in the species he examined. the fact that the degree of synchrony is not strongly Recently, S. Lohrenz (unpublished data) moni- influenced by photoperiod and/or light intensity and tored the cell division patterns in the marine cyano- tends to be tightest in the populations grown under bacterium Synecococcus sp., grown in a cyclostat long photoperiods of low light intensity. As Paasche in infradian growth mode. Although the patterns are (1967) recognized, this is different from the case for not clear-cut, peak division rates appear to occur diatoms, where short photoperiods of high light in the light period reaching a maximum towards the intensity significantly enhance the synchrony. end and declining during the dark period (Fig. 7). The It is somewhat difficult to address the question impression one gets (with little justification) is that the of growth rates for this group of species, because observed pattern reflects a cessation of population the data are not easily extracted from some of the growth in the dark rather than a phasing of cell cycles. studies. Paasche's data for E. huxleyi are all for The increase in optical density in the late dark period, populations doubling once per day, , and the species however, is difficult to explain, as are the negative examined by Chisholm and Brand (1980) were all division rates (cell death?) that occur regularly each in infradian growth (generation times 1 d). As day. This latter phenomenon appears to be a common far as can be deduced from the Nelson and Brand occurrence in light/dark grown cultures of non- (1979) data, those species were also doubling close diatoms . Chisholm and Costello (1980) have ieen to once per day. In clear cases of ultradian growth, similar "negative" division rates in cyclostat-grown such as P. striata grown under long photoperiods, cultures of Hymen omonas carterae, and Shifrin multiple division bursts can be seen. We have ob- (personal communication) observed them in the fresh-

165

TABLE 6. Division patterns of various species of phytoplankton cultured on light/dark cycles. The intervals during which division was occurring (subjectively determined) are indicated with an x. Each x is 1 h and the blackened interval represents the dark period of a 24-h light/dark cycle.

Species Clone Reference Temp., °C Light intensity Division pattern

XXXXXXXX Emeliania huxleyi Paasche 1967 21 0.07 cal•cm-'•min- ' (ex-Coccolithus huxleyi) XXXXXXX XX (Haptophyceae) 21 0.028 cal•cm-2 •min- '

XXXXXXX X 21 0.015 cal•cm"'• min-1 3M11.■

21 0.011 cal•cm-2 •min- ' XXXXXX■••■■

XXXXX XXXXXXXXX Emeliania huxleyi BT6 Nelson and Brand 1979 20 5 x 10-2 ly•min-I (ex-Coccolithus havleyi) XXXXXX XXXXXXXXXX (Haptophyceae) G4 Nelson and Brand 1979 20 5 x 10'2 ly•min-I 7■11■I

XXXXXX XXXXXXXXXX 92A Nelson and Brand 1979 20 5 x 10-2 ly•min-I

XXXX XXXXXXXXXX WHA Nelson and Brand 1979 20 5 x 10_2 ly•min-1

XX XX XXXXXXXXXX MCH Nelson and Brand 1979 20 5 x 10_2 ly•min- ' ■•■•■1

XXX% XXXXXXXXXX MCH Chisholm and Brand 1980 16 40 ILE • m-2 . s -1

XX XXXXX XXXXX 451B Nelson and Brand 1979 20 5 x 10-2 ly• min"' ■IMO■I

Xx XXXXX XXXXX 45111 Chisholm and Brand 1980 16 40 bt.E•m-.2 •s-, ■•••••■

XXXXXXXXXX Hymenomonas carterae CoccoII Nelson and Brand 1979 20 5 x 10-2 ly•min- I 1■•••■ (Haptophyceae) XXXXX XXXXX Chisholm and Brand 1980 16 40 1.2.E• • s-1

XXXX XXXXXX Prymesium parvum Prym Chisholm and Brand 1980 16 40 ILE•m-2 •s-I (Haptophyceae)

XXXXXXXXXX Chroomonas salina 3C Nelson and Brand 1979 20 5 x l0-2 ly•min-1 ■••■••• (Cryptophyceae)

XXXXXXXXXX lsochrysis galbana Iso Nelson and Brand 1979 20 5 x 10-2 ly•min"' xxxxx (Chrysophyceae)

XXXXXXXXX X Pavlova hither( Mono Nelson and Brand 1979 20 5 x 10-2 ly•min- ' xxxx (ex-Monoclnysis hither() (Cluysophycene)

40 /LE. xxxxxxx XXXX Olisthocliscus sp. Olistho Chisholm and Brand 1980 16 (Xanthophyceae)

XXXXXXXXXX Pyramimonas sp. Pyr-1 Chisholm and Brand 1980 16 40 p.E•in-'• (Prasinophyceae)

XXXX Platymonas striata Ricketts 1979 20 6500 lx (Prasinophyceae) 25 6500 lx XXX%

XXXX XXXXXX 15 6500 lx ■10111111■1■1

XXXX XXXXXX 20 6500 lx IMI■MII■

20 6500 lx XX XX XXXX

166 0.10

0,05H

000

005-

• (ID)

005

0,00 ••

0.051— 2 3 4 TIME (days)

Flo. 7. Patterns of cell division in Syneco coccus sp. grown on light/dark cycles in a cyclostat. Culture was nutrient replete and maintained at 16°C on L:D 14:10 with a light intensity of 70 iLE•ni-2 •s" and a dilution rate of 0.02 h'. itt(t) was calculated directly from (a) cell density data and (b) measurements of absorption by the cell suspension at 750 nm. It is the rate of change of these parameters. (S. Lohrenz unpublished data.) water green algae Oocystis polymorpha. In continuous mentally different from those of other groups of cultures, these negative rates could reflect error in the phytoplankton. The arguments buttressing this view, dilution rate measurements, but the fact that Nelson as well as a thorough survey of the literature on and Brand (1979) observed the same thing in batch phased division in diatoms, are presented in a recent cultures of several species makes us doubt that this is paper (Chisholm et al. 1980) so only a brief history the explanation. Thus, the decrease in cell number and the essence of the argument will be covered right before the onset of the division burst must either here. reflect cell death (and disintegration), or sticking of some cells to the walls of the container so samples Interest in growth•patterns in diatoms on light/ collected underestimate the population density. We dark cycles began early (Rieth 1939; Braarud 1945; have been unable to find evidence to support either Subrahmayan 1945; von Denffer 1949). A series of of these hypotheses; thus the phenomenon remains papers 20 yr later began revealing that the responses a mystery. of diatoms grown on light/dark cycles were not clear-cut. Palmer et al. (1964) reported that division occurred in the last half of the dark period in Phaeo- Diatoms dactylum tricornutum , while Jorgensen (1966) ob- The patterns of cell division expressed by dia- served the exact opposite pattern in Skeletonema toms grown on light/dark cycles appear to be funda- costatum , and Glooschenko and Curl (1968) found

167 no evidence of phasing at all in the latter species. or light intensities used in the various studies. Di- Eppley et al. (1967), Paasche (1968), and Chisholm tylum , for example, appears to divide primarily during et al. (1978) found similar patterns in Dityhan bright- the night at 15°C and during the day at 20°C (compare wellii with division occurring mainly in the light panels E and F in Fig. 9). period, while Richman and Rogers (1969) found Recently, Nelson and Brand (1979) and Chis- division to be restricted to the dark in this species. holm and Costello (1980) noted that several species More recent evidence (Chisholm et al. 1980) confirms of diatoms, when grown on L:D 14:10 or 12:12, early suspicions that many of these differences were exhibit division patterns characterized by multiple due to differences in the experimental temperatures bursts in population division rate (mg ) that appear

A 2.01- 7 l

1.0

— 780

_J 0 740 _J _.J LiJ

LLJ (9 7001- CC LL.1 •zi

2 3 4 5 6 D AY S 2.0

e•—■

o -o — 1.0

0.0 2 3 4 5 6 7 DAYS FIG. 8. Division patterns of Thalassiosira weissflogii grown in a cyclostat (data from Chisholm and Costello 1980). ,u(t) was calculated from a smoothed function fitted to the cell density data and is the first derivative of that function. (A) Specific division rate IL(t) as a function of time for a culture grown on L:D 14:10, 190 gE•m -2 •s- ', at 20°C at an average population growth rate of 1.31 d- ' (broken line). (B) Average cell volume (as determined with an electronic particle counter) as a function of time for the population in (A). Broken line indicates daily average cell volume. (C) Specific division rate as a function of time for a culture grown on L:D 10:14, 190 ILE•m -2 •s-', at 20°C at an average population growth rate of 1.06 d-'. 168 "typical" synchronypatterns,i.e.containing tills) cultures. high lightintensitywereusedto synchronize points during patterns such situations these interval where (at leastthose to be division patternsof to repeatthemselveswith (see those a out by pattern division rates the = panels A,B,C,G,Q,R,andV, Fig.9). If the is shown patterns cases we normalize 0) light/dark rule ratherthan for It isnoteworthyalsothat in and asingle except V Thalassiosira weis.sflogii as which relatively which reflectwhatmightbeconsidered are FIG. 9. light period.Speciesnames,experimental rate per24-h Nelson andBrand(1979). were not the in the in Fig.8. that displayedby

no changein MAXI MUM DI VIS IO NRA TE that given 50 50 50 5 50 50 50 5C cycle. A interval where diatoms are positive minority) and (for Relative the exception.Theexceptions in the in Table7. burst summarize all 10 which therewere a 24-h period. Solidlinesindicate typical example • form (Fig. 9), .. short photoperiodsof of division cell cell numbersoccurs 20 A in Fig.9are T. weissflogii _ at alltimesthrough- — nr" period length of division patternsof il in l WO, = (ex. 30 our study All we findthat G F H C D B data onthe 0) reflect they weresmoothed T. fluvia- each clay ILO . of no data appear such . ...•...... • patterns are fact and All the an of 10 the HOURS various species •• conditions, and dian growth "synchrony" conditions above, entrain) the light is Lewin 1967;Coombs period light/dark supplied periodically,synchronizes induced by starvation synchrony phased to on L:D8:16, sion the other that used. Based 1980) -- dark period, \f 20 effects normalized to patterns in the There is taxonomie \ and '• the and 30 acting asa cell , .... ••• the converted using used tightest degree of that is pulses of modes. cell and cycle in on of observed references light/dark environmental one T. weissflogii which was in the groups(Chisholm diatoms. When these .•••••• division cyclesofthepopulation. dotted linesindicate ....• . the maximumdivision the questionofultradian cycles flaw Many ... 10 limiting resourcewhich,when ,.....' (Lewin et diatoms isnotentrainedby other nutrients. experiments observations, in et for cycle inthe in theargument the formulaof some al. 1967), (if of 20 30 is analogoustosilicon- each pattern the not most) (Chisholm conditions onthe phasing wasobserved the data cases onshort shortest photoperiod al. 1966;Busby 'y in Fig.9 T X W S U R et the it wasproposed manner and synchrony (but al. 1980). In of the and such presented vs. infra- does not promote Costello of growth photo- cases, most divi- The and 169 or

TABLE 7. Species, experimental conditions, and references for data shown in Fig. 9.

Temper- ature, Panel Species Clone L:D cycle °C Light intensity Reference

A Ditylum brightwelln Dit 8: 16 20 0.032 cal•cm-2 • min" Paasche 1968 4.5:19.5 20 0.120 cal•cm-2 •min- ' 8:16 20 0.011 cal • cm -2 • min- ' D 14:10 20 0.015 cal • cm -2 . min - I 12:12 15 0.030 cal•cm-2 •min -' Chisholm et al. 1980 12:12 21 0.030 cal•cm--2 •min- ' G 8:16 20 0.050 cal•cm -2 •min-I Eppley et al. 1967 H 10:14 18 0.030 cal•cm -2 •min- ' Chisholm et al. 1978

Thalassiosira weissflogii SA 14:10 20 0.050 ly•min- ' Nelson and Brand 1979 Actin 14:10 20 0.050 ly•min- I Actin 10:14 20 190 gE• nr2 • s-1 Chisholm and Costello 1980

• Thalassiosira pseudonana 66A 10:14 20 0.030 cal. cm -2 . min' Chisholm et al. 1978 13 - 1 14:10 20 0.050 ly•min" Nelson and Brand 1979 14:10 20 0.050 ly• min" Nelson and Brand 1979

O Phaeodactylum tricormuum Pet Pd 14:10 20 0.050 ly.min" Nelson and Brand 1979

• Chaetoceros simplex Bbsm 14: 10 20 0.050 ly•min" Nelson and Brand 1979

• Nitzschia turgidula 5:19 20 0.069 cal. cm-2 •min" Paasche 1968 8:16 20 0.022 cal.cm-z • 12:12 20 0.013 cal•cm -2 •min" 16:8 20 0.009 cal. cm -2 •min"

• Cyclotella cryptica W1'- 1 - 8 14:10 20 0.050 ly•min' Nelson and Brand 1979

✓ Skeletonema costatum 12:12 20 0.060 cal • cm -2 . min" Jorgensen 1966

• Navicula ostrearea 16:8 14 3500 lx Neuville and Daste 1977

X Nitzschia paten 12:12 ? 118 p.E•m -2 •s" Hunding 1978

ultradian (generation times shorter than 1 d) growth. investigators have recognized the extremely unna- Thus, we should not expect to see entrainment in tural nature of such systems (e.g. Jannasch 1974), the classical sense. The question then becomes whe- the ease with which steady-state chemostats can be ther we would see entrainment of cell cycles if we sampled and described mathematically usually over- forced the populations into infradian mode by some rides the logic of using light/dark cycles. Only in form of growth limitation. The answer (for one a few investigations (e.g. Eppley et al. 1971; Chis- species at least) is no. When the growth rate of holm et al. 1975; Gotham 1977; Williams 1971; populations of T. weissflogii is reduced to less than one Malone et al. 1975; Chisholm and Costello 1980) doubling per day by lowered temperatures or nutrient have nutrient-limited growth and the phasing of cell limitation, entrainment does not result (Fig. 10b, c). cycles in phytoplankton been studied simultaneously, Moreover, the degree of phasing to the light/dark usually in cyclostat systems (Chisholm et al. 1975) cycle (although difficult to quantify) appears to be that are (physically at least) nothing more than a weaker than in the ultradian case (compare with Fig. chemostat on a light/dark cycle. 10a). This is not characteristic of clock-coupled cell A complete mathematical description of the division, and can only be interpreted as a forced, rather growth of phytoplankton cyclostats has been devel- than entrained, periodicity. oped by Gotham (1977) and Frisch and Gotham (1977, 1978) and recently reviewed by Rhee et al. Nutrient - Limited Cyclostat Growth (1981). Unfortunately, the data base for the theory comes exclusively from Euglena gracilis , because Most of our understanding of nutrient-limited this is the only species for which periodic components growth in phytoplankton is a result of continuously have been exhaustively studied in a cyclostat system. lit, steady-state chemostat cultures. Although many Caution must be exercised in generalizing from this

170 species in particular, because it is so tightly clock controlled. The nutrient-limited growth of algae in a cy- clostat can be described by a set of equations not unlike those developed for chemostats (Frisch and Gotham 1977,1978; Rhee et al. 1981) except that they consider time dependency. When cells are grown on a 24-h light/dark cycle in such a system, only the period average of the derivatives of residual limiting nutrient concentration (S), cell number (X), nutrient uptake rate (V), and instantaneous division rate (i.t) are inde- pendent of time such that:

(1) » = < VX = < VX > 7./ (S„ —

(2) T = 1 T where < . . . . >7, — f .. dt (or period T .• average), T = 24 h, Q is the amount of limiting nutrient per cell, and S„ is the influent concentration of the limiting nutrient. From these equations:

(3) DT = f g(t) dt

where DT is the integrated specific growth over the period, T, and is equivalent numerically to the dilution rate, D, of the cyclostat. The relationship between nutrient cell quotas and growth rates is well established for chemostat- grown cultures (Droop 1973) and has been docu- mented for average values in some cyclostat studies (Chisholm et al. 1975; Eppley et al. 1971). Gotham (1977) has shown that the period averaged growth rate in P-limited Euglena gmcilis grown in a cyclostat is a function of the minimum cell quota per 24-h period (Q,.) for a given dilution rate:

(4) = T (Q e Quo) K„,. + (Qc — Q co )

where Q,.„ = Q,. for 7. = 0, is the theoretical maximum growth rate, and K„,. is a constant. An example of the temporal patterns be-

several days. (A) Ultradian growth mode with nutrient- replete conditions at 20°C and an average growth rate of 1.06 d - ' (data are the average of those in Fig. 8C). (B) Infradian growth mode imposed by lowered temperature (I5°C). Average population growth rate was 0.62 d - ' (0.89 doublings/d). (C) Infradian growth mode imposed FIG. 10. Average division patterns of Thalassiosira by PO.T limitation. Average growth rate was 0.61 cr. weissflogii grown in a cyclostat under various conditions (0.88 doublings/d). Broken function is the g(t) pattern (data from Chisholm and Costello 1980).t) time courses from (B), phase shifted by 8 h for comparison. (See Chis- were calculated as described in Fig. 8 and averaged for holm and Costello 1980.)

171 At the beginning of the dark phase of the photo-

(X) cycle, p(t) jumps dramatically to a peak as cell

ity division begins and Q(t) begins declining due to

ns net loss to daughter cells. Plots of WO as a function De of Q(I) for any given dilution rate (growth rate) in ll

Çe the cyclostat describe closed trajectories (Fig. 12). If the cyclostat were placed in continuous light these trajectories would spiral into a family of ii,(Q) values that describe a Droop hyperbola (Droop 1973). The capacity of the cells to take up PO , also changes «É- I .0- over the 24-h light/dark cycle (Fig. 11). It should be clear from this analysis that for ; species like Euglena, the basic functional relation- Fri _ - ships derived from the period averages in the cy- clostat do not deviate from our understanding of a. • nutrient/growth relationships derived from chemostat a. studies (Rhee et al. 1980). The time courses of :5. `?, -•‹- 0.2 nutrient uptake, Q(t) and p (l), take on significance, cc however, when more than one species or fluctuation in the nutrient environment are considered. Although quantitative analyses have been done (Chisholm and Nobbs 1975; Frisch and Gotham 1978; Rhee et al. 1980), one needs only to ponder the problem in a qualitative sense to realize that consideration of only the daily averaged behavior of organisms cannot predict the results of a dynamic competitive inter- action among species and between species and their environment. One can apply this same reasoning at the cellular level, i.e. the average behavior of a population of cells does not necessarily reflect the behavior of the individual cells that comprise that population (Chisholm et al. 1980).

Patterns of Division Observed in Situ

The first observations of synchronized mitosis in phytoplankton in situ were reported near the turn FIG. 11. Time courses of measined variables from P01 of the century (Gough 1905; Apstein 1911; Jorgensen limited Euglena gracilis grown in a cyclostat on L:D 14:10 1911; Gran 1912) for the members of the genus with a dilution rate of 0.68 d - '. X = cell density showing Ceratium. Since that time, this genus and other washout when there is no division and an increase when dinoflagellates have been the object of numerous division is occurring. /2 = the instantaneous division rate. Q = cell quota or the amount of P per cell. f/ = maximum investigations into the phasing of phytoplankton cell capacity for P uptake at saturating concentrations of P. division in situ because of the easily identified (Modified from Gotham 1977.) division stages. In Ceratium, division stages are usually res- tricted to the late night-early morning hours (El- brachter 1973; Doyle and Poore 1974; Heller 1977; tween Q(t), dQldt(t), X(t), ILO), S(1), and 0( 1) Weiler and Chisholm 1976; Weiler and Eppley for a given average growth rate are shown in Fig. 11. 1979; Weiler 1980) although dividing cells have been One hour after the beginning of the light period observed in midday (Lanskaya 1963; Pavlova and Q(t) = Qe and ,u(t) is zero; WO remains equal Lanskaya 1969). A typical pattern for this genus is to zero throughout the light period because this is shown in Fig. 13. Weiler (1980) found that the outside the division gate, but Q(t) increases from timing of maximum division frequency varied by minimum to a maximum value during this period. only 2 h among 51 different Ceratium species in

172 20 • Paired nuclei A 0 Recently divided 20 15

10 1.5 -ic 5

Y‹ 10 Cerolium faro° 15 S 0.5 L

CEL 10

5

0 TOTAL

1.0 2,0 3,0 % Q x 10 -7 Irnol • cell -I 80 FIG. 12. Phase loop trajectories from experimental data Dinophysis for//i for instantaneous growth rate, L(t), as a function of phos- phate quota, Q(t) in Euglena gracilis grown in a cyclostat. Arrow marks the subsistence quota (below which no growth occurs). DT is the average specific growth rate of the 40 culture (d -'). (Redrawn from Gotham 1977.)

the central Pacific (photoperiods of 10.5-14.0 h), always occurring around 04:00. The timing and duration of division in the field populations agreed 24:00 0800 16:00 well with laboratory cultures grown under similar HOURS conditions (Weiler and Eppley 1979). The width of the division gate in this genus is apparently varia- FIG. 13. The frequency of paired nuclei of various dino- ble. Reported values range from 2 h (Doyle and Poore flagellate species observed as a function of time of day in samples collected in Santa Monica Bay, CA. Open 1974) to about 12 h (Heller 1977), but are commonly circles are 1/2 the recently divided cells. (From Weiler on the order of 5-7 h (Weiler and Eppley 1979; and Chisholm 1976.) Elbrachter 1973). The maximum proportion of cells observed dividing, which reflects the population growth rate (see below), has been found to range divided between 04:00 and 06:00. Pyrocystis from 1 to 40% (Elbrachter 1973), but usually is noctiluca and P . fusiformis show maximal production between 10 and 20%. of reproductive cells during the latter part of the The in situ division patterns of other dinofla- night (Swift and Durbin 1972), but the gate appears gellates are not as well studied as those of Ceratium , to span the entire dark interval. In contrast, division but data exist for several species . Weiler and Eppley in fast-growing field populations of Dinophysis fortii (1979) observed that in six species of Ornithocercus has been noted to span the entire day, with the maxi- division frequency was maximal (14% dividing cells) mum frequency of paired cells occurring soon after at dawn, but dividing cells were observed throughout sunrise (Weiler and Chisholm 1976). The freshwater the 24-h period indicating a broad division gate. dinofiagellate Peridinium cinctum divides primarily This contrasts with the data for O. magnificus reported between 02:00 and 04:00 regardless of the season by Doyle and Poore (1974) in which dividing cells (Pollingher and Semiya 1976). (8% of the total) were observed only at 05:00. Although the patterns of in situ division phasing In the latter study, paired cells of Peridiniuin sp. in dinoflagellates can be characterized generally by were only observed at 03:00 and Ceratocorys horrida a late night or early morning division gate, with a

173 significant interval over the day when no dividing blages by following the time course of silicic acid cells are observed, this is not the case with diatoms incorporation into frustules. Our hypothesis was (Smayda 1975; Lewin and Rao 1975; Chisholm et that if there was a clear preference for day- or night- al. 1978; Williamson 1980). Cells can be observed time division in diatoms, this should be reflected in the division process at all times of the day or night, in the silicon incorporation patterns of the whole but the frequency of dividing cells usually shows water sample. Typical time courses observed (Fig. one or more peaks during each 24-h period. 15) indicated that division (Si incorporation) was In early studies (Allen 1922; Savage and Wim- occurring all the time, with bursts appearing both penny 1936; Wimpenny 1938), the highest frequen- during the day and night. As we were dealing with cies of dividing cells in diatom populations were mixed populations, the patterns could be interpreted detected at night. More recently, , Smayda (1975) in two ways: either different species were dividing at observed that percentage of dividing cells was also different times, or all species were dividing continu- maximal in the night in natural populations of Ditylion ously, , with preferred division times both night and brightwellii and Biddulphia mobiliensis from the Gulf day. At the time, the former interpretation seemed of California (Fig. 14). Lewin and Rao (1975) on most plausible, but in view of the more recent evi- the other hand, found the maximum frequency of dence from culture work discussed above, the latter dividing cells to occur during the day in the surf-zone now seems more likely. The entire rationale behind diatom Chaetoceros armatum , and this timing was this shift in interpretation is discussed in detail in independent of the time of year. The various species Chisholm and Costello (1980) and Chisholm et al. examined by Williamson (1980) exhibited all possible (1980). patterns: no rhythmicity (Chaetoceros vanheurckii), Two reports of in situ synchrony have been dominant daytime peak (Thalassiothrix nitzschioides made for freshwater green algae. Simmer and Sodom- and Tropidoneis antarctica), or dominant nighttime kova (1968) observed division in Scenedesmus peak (Thalassiosira rotula). acumenatus (cultivated in large outdoor cultures) In an effort to avoid the tedium of scoring to be concentrated in the dark period. Stayley (1971), percent dividing cells for monitoring division in dia- who has made the only direct observation of cell toms, Chisholm et al. (1978) attempted to take advan- division timing in nature, actually imersed a micro- tage of the coupling between frustule deposition scope in a small pond and photographed the growth and cell division (Lewin et al. 1966) and tried to and cell division in attached chlorelloid algae over assay the timing of division in natural diatom assem- the day. Most of the size increase in the cells occurred during the day and division at night. Growth rates of the cells ranged from 1.3 to 1.8 doublings/d. Ditylum brightwelli Measurements of the frequency of division stages in natural populations of phytoplankton can 70- be used to calculate in situ growth rates of the popu- • lations. Various formulae have been used for this 60- calculation, which depend on either the maximum

CELLS frequency of paired division stages (Swift and Durbin

) 1972; Elbrachter 1973; Lewin and Rao 1975; Smayda 50- ED 1975; Swift et al. 1976; Weiler and Chisholm 1976)

AIR • or the sum of the frequency of paired stages observed,

(P • 40 - • over a 24-h period (Weiler and Chisholm 1976; Heller

ET • 1979; Weiler and Eppley 1979), for the growth rate BL • calculation. The expressions that have been used 30H DOU are diverse, often of unclear origin, and vary in • • degree of correctness and in underlying assumptions. 2 0 -1 • McDuff and Chisholm (1981) have reviewed CENT • • •

ER the various formulations used in the past, and P • • 11115 present a derivation of the appropriate expressions 10H • from first principles. As noted by others (Swift • et al. 1976; Weiler and Chisholm 1976; Weiler and 0 111.11.11.1 1 1 111111.111111111. Eppley 1979), in situations where the maximum 0CIO0 0be0 013:00 12D0 Iwo zono frequency of paired division stages (f) observed in a 24-h period is a measure of all the cells that are FIG. 14. The percentage of doublet (paired) cells of dividing in that interval, the expression Ditylum brightwellii in net tows as a function of time of 1 day. Samples were collected in the Gulf of California. ln (1 + f„,„,) (From Smayda 1975.) (5) = —

174 1 " 0.4 —A (6) = — ln (f; + 1) o Natural light n td = • Growth chamber where n is the number of samples taken and td is the duration of the division stage. When division is 0.3 phased or synchronous, the sampling period (n t,.) must be an even multiple of 24 h to obtain an accurate o estimate of the average o daily growth rate. o The critical parameter in equation (6) is the

— o duration of the paired division stage, td . This can 0.2 c • be estimated from the interval between the peaks in division frequency and frequency of recently divided 0 cells (if identifiable). Alternatively, it can be calcu- lated from data from laboratory cultures in which CI) the increase in cell number (N) is followed along 0.1 with the frequency of paired division stages. Since

(7) p,= In (NI IN0 )11t o substituting into equation (6) we find that a) o At CC In (f; + 1) n In (N,/N0) ;=1 40 o `c3 • o Measuring the frequency of division stages in c • natural populations of phytoplankton is perhaps one 20 of the best (surely the most direct) tools to estimate o the growth rates of phytoplankton in the sea. The accuracy of this method, however, is critically de- 65 pendent on the following factors: (i) that there is no differential mortality between the various division o stages (and there may be for some species (Richman 1400 2 200 0600 1400 and Rogers 1969)) and (ii) that either the duration of HOURS the division stage monitored is precisely known and FIG. 15. (A) Time course of Si(OH) 1 incorporation into an appropriate sampling interval is chosen or (iii) all diatom frustules in a surface seawater sample collected 3 km the cells undergoing division in a given day can be off La Jolla, CA, incubated under natural conditions, and observed doing so at one time. in a growth chamber at 17°C on L:D 12:12. Initial Si(OH), concentration was 3.5 /Le (B) Data from (A) converted to rates of Si(OH) 1 incorporation for each inverval between Adaptive Significance and measurements. Negative rates are plotted as zero. (From Chisholm et al. 1978.) Ecological Implications

It is clear from this review that phytoplankton where t = 1 d, correctly describes the specific species differ in their growth patterns on light/dark growth rate of the population per day. The use of cycles. The discussion would not be complete without this expression is only correct when the duration of some attempt at considering the selective pressures the paired division stage is long relative to the width that may have resulted in the observed patterns of of the division gate, or when recently divided cells division. can be recognized (Weiler and Chisholm 1976; In clock-controlled species, the timing of divi- McDuff and Chisholm 1981). sion is phase restricted, i.e. cells can only divide when For the more general case, which includes expo- the division gate is open. The possible selective nential growth, one must calculate the average of advantage of such rigid control over the timing p, from sequential observations of the frequency of of the cell cycle has been discussed repeatedly in paired division stages (f) over a 24-h period to obtain recent years (e.g. Smayda 1975; Starkweather 1975; a daily averaged growth rate Chisholm et al. 1978; Weiler 1978). Interesting

175 hypotheses suggesting the avoidance of grazing pres- distribution of properties among cells comprising sure (McAllister 1970; Starkweather 1975; Chisholm populations, it is difficult to do more than speculate et al. 1978) or the reduction of competition (Williams on the selective pressures that have resulted in 1971; Eppley et al. 1971; Stross et al. 1973; Doyle the patterns of division expressed by phytoplankton and Poore 1974; Chisholm and Nobbs 1976; Chis- populations. holm et al. 1978) have been advanced, but they are difficult, if not impossible, to test. It seems likely that part of the selective advantage at least must involve the optimization and maintenance of the temporal Acknowledgments order of cellular events (Gotham 1977; Chisholm et al. 1980). In other words, clock control of division This work was supported in part by NSF-OCE- timing could simply serve to insure that the cell 7708999, by the MIT Sea Grant College Program division cycle is optimally phased to the periodic under grant No. NA-79AA-D-00101 from the Office supply of energy to the cell. of Sea Grants, National Oceanic and Atmospheric In some species, in particular certain diatoms, Administration, U.S. Department of Commerce, and the timing of division is not phase restricted (Chis- by funds awarded to S.W.C. as Doherty Professor holm et al. 1980); thus the advantage of strict temporal of Ocean Utilization. ordering does not appear to be universal. Division rhythms expressed by diatom populations appear to be forced oscillations rather than entrained. Strict cell division phasing occurs only when the cells are References subjected to a strong forcing oscillation such as short photoperiods or nutrient pulses (Chisholm et al. ABBO, F. E., AND A. B. PARDEE. 1960. Synthesis of 1980). Under most growth conditions and photo- macromolecules in synchronously dividing bacteria. periods, division occurs at all times during the photo- Biochim. Biophys. Acta 39: 478-485. cycle; thus individual cells in the population are not AGRELL, I. 1964. Natural division synchrony and mitotic gradients in metazoan tissues, p. 39-70. In E. Zeuthen responding to the light/dark cycle in the same manner [ed.] Synchrony in cell division and growth. Interscien- (Chisholm and Costello 1980). Each cell "processes" ce, New York. environmental inputs as it experiences them rather ALLEN, W. E. 1922. Quantitative studies on marine phyto- than aligning the cell cycle to the environmental plankton at LaJolla in 1919. Univ. Calif. Berkeley, periodicities by "waiting" for the division gate. Pub!. Zool. 22: 329-347. It has been suggested that the flexibility in the APSTEIN, C. 1911. Biologische studie uber Ceratium division patterns of diatom cells grown on light/ tripos var. subsalsn Ostf. Wiss. Meeresunters. N.F. (phenotypic) Bd. 12 (S): 137-162. dark cycles reflects extensive nongenetic ASCHOFF, J. 1965. Response curves in circadian perio- variability in the populations (Chisholm et al. 1980), dicity, p. 95-111. In J. Aschoff [ed.] Circadian clocks. particularly in terrns of the distribution of generation North Holland Pub!. Co., Amsterdam. times. This necessarily results in asynchronous BERNSTEIN, E. 1960. Synchronous division in Chlamy- division, unless a forcing oscillation in a limiting domonas moewusii. Science (Washington, D.C.) 131: substance is strong enough to align the cell cycles. 1528-1529. The relative selective advantages of these two 1964. Physiology of an obligate photoautotroph growth "strategies" (i.e. strict phasing of cell cycles (Chlamydomonas moewusii). I. Characteristics of versus the absence of phase-restricted division) synchronously and randomly reproducing cells and population curves. J. variety of contexts in a hypothesis to explain their have been discussed in a Protozool. 11(1): 56-74. and Koshland 1976; Gotham recent years (Spudich BRAARUD, T. 1945. Experimental studies on marine 1977; Chisholm et al. 1980). Although difficult to plankton diatoms. Nor. Vidensk. Akad. Oslo I. Mat. prove, it seems logical that species with strict, clock- Naturvidensk. KI. No. 10. coupled temporal ordering might be selected for BRUCE, V. G. 1970. The biological clock in Chlamydo- in stable, predictable environments, whereas those monas reinhardtii. J. Protozool. 17: 328-334. species with less rigid cell cycle timing would thrive 1972. Mutants of the biological clock in Chlamy- in environments where the population is subjected to domonas reinhardtii. Genetics 70: 537-548. widely varying conditions. The difficulty in examin- BRUCE, V. G., AND C. S. PITTENDRIGH. 1956. Tempe- rature independence in a unicellular "clock." Proc. ing this hypothesis in more detail is in defining Natl. Acad. Sci. U.S.A. 42: 676-682. predictable and unpredictable environments , BÜHNEMANN, F. 1955. Die rhythmische sporenbildung particularly on the space and time scales experienced von Oedogonium cardinchan. Zentralbl. Wittr. Biol. by single cells. Until we know more about the 74: 1-54. temporal and spatial distribution of the key environ- BÜNNING, E. 1967. The physiological clock. Springer- mental parameters a cell experiences, as well as the Verlag, New York. 167 p.

176 BUSBY, W. F., AND J. LEWIN. 1967. Silicate uptake ration in silicon-starvation synchrony of Navicula and silica shell formation by synchronously dividing pelliculosa. Plant Physiol. 42: 1607-1611. cells of the diatom Navicula pelliculosa (Bréb) HiIse. CSATORDAY, K., AND G. HORVATH. 1977. Synchro- J. Phycol. 3: 127-131. nization of Anacystis nidulans: oxygen evolution BUSH, K. J., AND B. M. SWEENEY. 1972. The activity during the cell cycle. Arch. Microbiol. 111: 245-246. of ribulose diphosphate carboxylase in extracts of DENFFER, D. VON. 1949. Die planktische massenkultur Gonyaulax polyedra in thé day and the night phases pennater Grunddiatomeen. Arch. Milcrobiol. 14: of the circadian rhythm of photosynthesis. Plant 159-202. Physiol. 50: 446-451. DOYLE, R., AND R. V. POORE. 1974. Nutrient competition CHANCE, B., E. K. PYE, A. K. GHOSH, AND B. HESS. and division synchrony in phytoplankton. J. Exp. 1973. Biological and biochemical oscillators. Aca- Mar. Biol. Ecol. 14: 201-210. demic Press, New York. 374 p. CHIstiouo, S. W., F. AZAM, AND R. W. EPPLEY. 1978. DROOP, M. R. 1973. Some thoughts on nutrient limitation Silicic acid incorporation in marine diatoms on light/ in algae. J. Phycol. 9: 264-272. dark cycles: use as an assay for phased cell division. EDMUNDS, L. N. JR. 1965a. Studies on synchronously Limnol. Oceanogr. 23: 518-529. dividing cultures of Euglena gracilis Klebs (Strain Z). CHISHOLM, S. W., AND L. E. BRAND. 1980. Persistence I. Attainment and characterization of cell division. J. of cell division phasing in marine phytoplankton in Cell. Comp. Physiol. 66: 147-158. continuous light after entrainment to light/dark cycles. 1965b. Studies on synchronized dividing cultures J. Exp. Mar. Biol. Ecol. 51:107-118. of Euglena gracilis Klebs (Strain Z). II. Patterns of CHISHOLM, S. W., AND J. C. COSTELLO. 1980. Influence biosynthesis during the cell cycle. J. Cell. Comp. of environmental factors and population composition Physiol. 66: 159-182. on the timing of cell division in Thalassiosira 1966. Studies on synchronously dividing cultures fluviatdis (Bacillariophyceae) grown on light/dark of Euglena gracilis Klebs (Strain Z). III. Circadian cycles. J. Phycol. 16: 375-383. components of cell division. J. Cell. Physiol. 67: CHISHOLM, S. W., F. M. M. MOREL, AND W. S. 35-44. SLOCUM. 1980. The phasing and distribution of cell 1971. Persisting circadian rhythm of cell division division cycles in marine diatoms, p. 281-300. In in Euglena: Some theoretical considerations and the P. Falkowski [ed.] Primary productivity in the sea. problem of intercellular communication, p. 594-611. Brookhaven Symp. Biol. 31. In M. Menaker [cd.] Biochronometry. Natl. Acad. CHISHOLM, S. W., AND P. A. NOBBS. 1976. Simulation Sci. USA. of algal growth and competition in a phosphate-limited 1974. Circadian clock control of the cell develop- cyclostat, p. 337. In R. P. Canale [cd.] Modeling mental cycle in synchronized cultures of Euglena, biochemical processes in aquatic ecosystems. U. Mich. p. 287-297. In R. L. Bieleski, A. R. Ferguson, and Ann Arbor Science Publ. M. M. Cresswell [cd.] Mechanisms of regulation of CHISHOLM, S. W., AND R. G. STROSS. 1976. Phosphate plant growth. R. Soc. N.Z. Bull. 12. uptake kinetics in Euglena gracilis (Z) (Eugleno- 1976. Models and mechanisms for endogenous phyceae) growth on light/dark cycles. II. Phased, timekeeping, p. 280-361. In J. D. Palmer [cd,] PO I-limited cultures. J. Phycol. 12: 217-222. An introduction to biological rhythms. Academic Press, CHISHOLM, S. W., R. G. STROSS, AND P. A. NOBBS. New York. 1975. Light/dark phased cell division in Euglena 1978. Clocked cell cycle clocks: implications gracilis (Z) (Euglenophyceae), in PO I limited cultures. toward chronobiology and aging, p. 125-184. In J. Phycol. 77: 367-373. H. V. Samis and S. Capobianco [cd.] Aging and Coolc, J. R. 1961a. Euglena gracilis in synchronous divi- biological rhythms. Adv. Exp. Med. Biol. Vol. 108. sion. I. Dry mass and volume characteristics. Plant EDMUNDS, L. N. JR., L. CHUANG, R. M. JARRETT, Cell Physiol. 2: 199-202. AND W. O. TERRY. 1971. Long-term persistence of 1961b. Euglena gracilis in synchronous division. free-running circadian rhythms of cell division of II. Biosynthetic rates over the life cycle. Biol. Bull. Euglena and the implication of autosynchrony. J. Woods Hole Mass. 121: 277-289. Interdiscip. Cycle Res. 2(2): 121-123. 1963. Adaptations in growth and division in EDMUNDS, L. N., AND R. R. FUNCH. 1969a. Circadian Euglena effected by energy supply. J. Protozool. 10: rhythm of cell division in Euglena: effects of a random 436-444. illumination regimen. Science (Washington, D.C.) 1971. Influence of culture pH and phosphate on 165: 500-503. synchrony of Euglena gracilis. Exp. Cell Res. 69: 1969b. Effects of "skeleton" photoperiods and 207-211. high frequency light/dark cycles on the rhythm of COOK, J. R., AND B. COOK. 1962. Effects of nutrients cell division in synchronized cultures of Euglena. on the variation of individual generation times. Exp. Planta 87: 134-163. Cell Res. 28: 524-530. EHRET, C. F. 1974. The sense of time: evidence for its CooK, J. R., AND T. W. JAMES. 1960. Light-induced molecular basis in the eukaryotic gene-action system. division synchrony in Euglena gracilis var. bacillaris. Adv. Biol. Med. Phys. 15: 47-77. Exp. Cell Res. 21: 583-589. EHRET, C. F., AND K. W. DOBRA. 1977. The infradian COOMBS, J., C. SPANIS, AND B. E. VoLcANI. 1967. eukaryotic cell: a circadian energy-reserve escapement, Studies on the biochemistry and fine structure of silica p. 563-570. Proceedings of XII International Con- shell formation in diatoms. Photosynthesis and respi- ference, International Society for Chronobiology

177 (Section on Cellular and Metabolic Mechanisms). GRAN, H. H. 1912. Pelagic plant life, p. 307-386. In Casa Editrice "II Ponte," Milano. J. Murray and J. Hjort [ed.] The depths of the ocean. EHRET, C. F., J. C. MEINERT, K. R. GROH, K. W. Macmillan and Co., London. DOBRA, AND G. A. ANTIPA. 1977. Circadian HARTWELL, L. H. 1971. Genetic control of the cell regulation: growth kinetics of the infradian cell, p. division cycle in yeast. II. Genes controlling DNA 62-64. In B. Drewinko and R. M. Humphrey [ed.] replication and its initiation. J. Mol. Biol. 59: 183-194 , Growth kinetics and biochemical regulation of normal 1974. Saccharomyces cerevisine cell cycle. and malignant cells. Williams and Wilkins Co., Bal- Bacteriol. Rev. 38: 164-198. timore. HASE, E., Y. MORIMURA, AND H. TAMIYA. 1957. Some EHRET, C. F., AND E. TRUCCO. 1967. Molecular models data on the growth physiology of Chlorella studied for the circadian clock. 1. The chronon concept. J. by the technique of synchronous culture. Arch. Bio- Theor. Biol. 15: 240-262. chem. Biophys. 69: 149. HASTINGS, J. W. 1960. Biochemical aspects of rhythms: EHRET, C. F., AND J. J. WILLE. 1970. The photobiology shifting by chemicals. Cold Spring Harbor of circadian rhythms in protozoa and other eukaryotic Phase microorganisms, p. 369-416. In P. Halladal [ed.] Symp. Quant. Biol. 25: 131-143. HASTINGS, features: group report 1, Photobiology of microorganisms. Wiley-Interscience, J. W. 1975. Basic H.-G. Schweiger [ed.] London, New York, Sydney, and Toronto. p. 49-62. in J. W. Hastings and The molecular basis of circadian rhythms. Report of ELBRACHTER, M. 1973. Population dynamics of Ceratium the Dahlem Konferenzen, Berlin, Life Sci. Res. Rep. I. in coastal waters of the Kiel Bay. Oikos Suppl. 15: HASTINGS, J. W., AND B. M. SWEENEY. 1958. A per- 43-48. sistant diurnal rhythm of luminescence in Gonynulax ENRIGHT, J. T. 1965. Synchronization and ranges of polyedra. Biol. Bull. 115: 440-458. entrainment, p. 113-124. In J. Aschoff [ed.] Circadian HASTINGS, J. W. , L. ASTRACHAN, AND B. M. SWEENEY. clocks, North Holland Publ. Co., Amsterdam. 1960. A persistent daily rhythm in photosynthesis. EPPLEY, R. W., AND J. L. COATSWORTH. 1966. Culture J. Gen. Physiol. 45: 69-76. of the marine phytoplankter, Dunaliella tertiolecta, HASTINGS, J. W., AND H. -G. SCHWEIGER. 1976. The with light/dark cycles. Arch. Mikrobiol. 55: 66-80. molecular basis of circadian rhythms. Rep. Dahlem EPPLEY, R. W., R. W. HOLMES, AND E. PAASCHE. Workshop, Abakon Ver. Berlin. 461 p. 1967. Periodicity in cell division and physiological HASTINGS, J. W., AND B. M. SWEENEY. 1964. Phased behavior of Dityhan brightwellii, a marine plankton cell division in the marine dinoflagellates, p. 307-321. diatom, during growth in light/dark cycles. Arch. In E. Zeuthen [ed.] Synchrony in cell division and Mikrobiol. 56: 305-323. growth. Interscience Publ., New York. EPPLEY, R. •W. , J. N. ROGERS, J. J. MCCARTHY, AND HELLER, M. D. 1977. The phased division of the fresh- A. SoURNIA. 1971. Light/dark periodicity in nitrogen water dinoflagellate Ceratium hirundinella and its assimilation of the marine phytoplankters Skeletonema use as a method of assessing growth in natural popu- costatum and Coccolithus lualeyi in N-limited lations. Freshwater Biol. 7: 527-533. chemostat culture. J. Phycol. 7: 150-154. HELLMAN, W. S. 1976. Biological rhythms and physio- ERIKSON, R. 0. 1964. Synchronous cell and nuclear logical timing. Annu. Rev. Plant Physiol. 27: 159- division in tissues of the higher plants, p. 11-37. In 179. E. Zeuthen [ed.] Synchrony in cell division and HERDMAN ET AL. 1970. Synchronous growth and genome growth. Interscience, New York. replication in the blue-green alga, Anacystis nidulans. FELDMAN, J. F. 1968. Circadian rhythmicity in amino acid Arch. Mikrobiol. 73: 238-249. incorporation in Euglena gracilis. Science (Washing- HESSE, M. 1971. Endogene schwankungen der produk- ton, D.C.) 160: 1454-1456. tionsfahigkeit bei Chlorella und ihre beeinflussung FELDMAN, J. F., AND M. N. HOYLE. 1973. Isolation durch licht sehr niedriger intensitat und durch tiefe of circadian clock mutants of Neurospora crassa. temperatur. Dissertation, Gottingen. p. 1-84. Genetics 75: 605-613. HOTTA, Y., M. ITO, AND H. STERN. 1966. Synthesis FRISCH, H. L., AND I. J. GOTHAM. 1977. On periodic of DNA during meiosis. Proc. Natl. Acad. Sci. U.S.A. algal cyclostat populations. J. Theor. Biol. 66: 665. 56: 1184-1191. 1978. A simple model for periodic cyclostat HUNDING, C. 1978. Growth cycle of a freshwater diatom. growth of algae. J. Math. Biol. 7: 149-169. Mitt. Int. Ver , Limnol. 21: 136-146, GALLERON, C., AND A. M. DURRAND. 1979. Cell cycle JAMES, T. W. 1966. Cell synchrony, a prologue to dis- and DNA synthesis in a marine dinoflagellate, Am- covery, p. 1-13. In I. L. Cameron and G. M. Padella phidinhun cartel. Protoplasma 100: 155-165. [ed.] Cell synchrony. Academic Press, New York, GLOOSCHENKO, W. A., AND H. CURL. 1968, Obtaining NY. synchronous cultures of algae. Nature (London) 218: JANNASCH, H. W. 1974. Steady-state and the chemostat 573-574. in ecology. Limnol. Oceanogr. 19: 716. GOTHAM, I. J. 1977. Nutrient limited cyclostat growth: JARRETT, R. M., AND L. N. EDMUNDS JR. 1970. Per- a theoretical and physiological analysis. Ph. D. thesis, sisting circadian rhythm of cell division in a photo- New York State Univ., Albany, NY. 222 p. synthetic mutant of Euglena. Science (Washington, GOUGH, L. H. 1905. Report on the plankton of the English D.C.) 167: 1730-1733. Channel in 1903. Rep. (S. Area) Fish, Hydrogr. Invest. JoHN, P. C. L., W. McCuu.ouGH, A. Q. ATKINSON, N. Sea 1: 325-377. B. G. FORD, AND B. E. S. GUNNING. 1973. The

178 cell cycle in Chlorella , p. 61-76. In M. Balls and LORENZEN, H. 1957. Synchrone zellteilungen von Chlo- F. S. Bellett [cd.] The cell cycle in development rella bei verschiedenen Licht-Dunkel-Wechseln. Flora and differentiation. Cambridge Univ. Press, Cam- (Jena) 144: 473. bridge, MA. 1970. Synchronous cultures, p. 187-212. In P. JORGENSEN, E. G. 1911. Die Ceratien , eine kurze mono- Halldal [ed.] Photobiology of microorganisms. Wiley- graphie der gattung Ceratium Schrank. Int. Rev. Interscience, London. Gesamten Hydrobiol. Hydrogr. Biol. Suppl. Ser. 2: LORENZEN, H., AND B. D. KAUSHIK. 1976. Experiments 1-124. with synchronous Atzacystis nidulans. Ber. Dtsch. 1966. Photosynthetic activity during the life cycle Bot. Ges. 89: 491-498. of synchronous Skeletonema cells. Physiol. Plant. LORENZEN, H., AND G. S. VENKATARAMAN. 1969. 19: 789-799. Synchronous cell divisions in Anacystis nidulans KLEVECZ, R. R. 1969. Temporal order in mammalian Richter. Arch. Mikrobiol. 67: 251-255. cells. I. The periodic synthesis of lactate dehydro- MCALLISTER, C. D. 1970. Zooplankton rations, phyto- genase in the cell cycle. J. Cell Biol. 43: 207-219. plankton mortality and the estimation of marine pro- 1975. Molecular manifestations of the cellular duction, p. 419-457. In J. Steele [ed.] Marine food clock, p. 1-19. In J. C. Hampton [cd.] The cell cycle chains. Oliver and Boyd, Edinburgh. in malignancy and immunity. 13th Annu. Hanford McDuFF, R., AND S. W. CHISHOLM. 1981. The Biol. Symp. cal- 1976. Quantized generation time in mammalian culation of in situ growth rates of phytoplankton populations from fractions of cells undergoing mitosis: cells as an expression of the cellular clock. Proc. a clarification. Limnol. Oceanogr. (In press) Natl. Acad. Sci. U.S.A. 73: 4012-4016. KOMAREK, J., AND J. SIMMER. 1965. Synchronization MCMURRY, L., AND J. W. HASTINGS. 1972. No desyn- of the cultures of Scenedesmus quadricauda (TURP.) chronization among four circadian rhythms in the Breb. Biol. Plant. 7: 409. unicellular alga, Gonyaulax polyedra. Science (Wash- 1(oNoFKA, R. J., AND S. BENZER. 1971. Clock mutants ington, D.C.) 175: 1137-1139. of Drosophila melanogaster. Proc. Nat. Acad. Sci. MALONE, T. C., C. GARSIDE, K. C. HAINES, AND O. A. U.S.A. 68: 2112-2116. ROELS. 1975. Nitrate uptake and growth of Chaeto- LAFEBER, A., AND C. L. M. STEENBERGEN. 1967. ceros sp. in large outdoor continuous cultures. Limnol. Simple device for obtaining synchronous cultures of Oceanogr. 20: 9-19. algae. Nature (London) 213: 527-528. MAZIA, D., AND K. DAN. 1952. The isolation and bio- LANSKAYA, L. A. 1963. Fission rate of plankton algae chemical characterization of the mitotic apparatus of of the Black Sea in cultures, p. 127-132. In C. A. dividing cells. Proc. Natl. Acad. Sci. U.S.A. 38: Oppenheimer [ed.] Symposium on marine micro- 826-838. biology. Charles C. Thomas, Publisher, Springfield, MENAKER, M. 1971. Biochronometry. Natl. Acad. Sci. IL. USA. 662 p. LAVAL - MARTIN, D. L., D. J. SHUCH, AND L. N. MIHARA, S., AND E. HASE. 1971. Studies on the vegetative EDMUNDS JR. 1979. Cell cycle related and endoge- life cycle of Chlamydotnonas reinhardtii Dangeard nously controlled circadian photosynthetic rhythms in synchronous culture. I. Some characteristics of the in Euglena. Plant Physiol. 63: 495-502. cell cycle. Plant Cell Physiol. 12: 225-236. LEEDALE, G. F. 1959. Periodicity of mitosis and cell MITCHELL, J. L. A. 1971. Photoinduced division syn- division in the Euglenineae . Biol. Bull. 116: 162- chrony in permanently bleached Euglena gracilis . 174. Planta 100: 244-257. LEWIN, J. C., AND V. N. R. RAà. 1975. Blooms of MITCHISON, J. M. 1971. The biology of the cell cycle. surf-zone diatoms along the coast of the Olympic Cambridge Univ. Press, Cambridge, MA. 313 p. Peninsula, Washington. VI. Daily periodicity pheno- 1973. Differentiation in the cell cycle, p. 1-11. mena associated with Chaetoceros armatum in its In M. Balls and F. S. Billet [cd.] The cell cycle in natural habitat. J. Phycol. 11(3): 330-338. development and differentiation. Cambridge Univ. LEWIN, J. C., B. E. REIMANN, W. F. BUSBY, AND Press, Cambridge. B. E. VOLCANI. 1966. Silica shell formation in syn- 1974. Sequences, pathways and timers in the cell chronously dividing diatoms, p. 169-188. In I. L. cycle, p. 125-142. In G. M. Padilla, I. L. Cameron, Cameron and G. M. Padilla [ed.] Cell synchrony. and A. Zimmerman [ed.] Cell cycle controls. Aca- Academic Press, New York, NY. demic Press, New York, NY. LIEN, T., AND G. KNUTSEN. 1979. Synchronous growth MORI MURA , Y. 1959. Synchronous culture of Chlorella. of Chlamydomonas reinhardtii (Chlorophyceae): a I. Kinetic analysis of the life cycle of Chlorella chip- review of optimal conditions. J. Phycol. 15: 191-200. soidea as affected by changes of temperature and light LOEBLICH, A. R. 1977. Studies on synchronously dividing intensity. Plant Cell Physiol. 1: 49. populations of Cachonina niei , a marine dinoflagellate. NELSON, D. M., AND L. E. BRAND. 1979. Cell division Bull. Jpn. Soc. Physiol. 25: 119-126. periodicity in 13 species of marine phytoplankton on LONERGAN, T. A., AND M. L. SARGENT. 1978. Regula- a light/dark cycle. J. Phycol. 15: 67-75. tion of the photosynthesis rhythm in Euglena gracilis. NEUVILLE, D., AND P. DASTE. 1977. Premiers essais I. Carbonic anhydrase and glyceraldehyde-3-phosphate de synchronisation des divisions cellulaires chez la dehydrogenase do not regulate the photosynthesis diatomée Navicula ostrearia (Gaillon) bory en culture rhythm. Plant Physiol. 61: 150-153. axenique. C.R. Acad. Sci. Paris Ser. D 284: 761-764.

179 NJUS, D., F. SULZMAN, AND J. W. HASTINGS. 1974. in vivo Chlorophyll a fluorescence. Plant Physiol. Membrane model for the circadian clock. Nature 60: 388-392. (London) 248: 116-120. RHEE, G. Y., I. J. GOTHAM, AND S. W. CHISHOLM. PAASCHE, E. 1967. Marine plankton algae grown with 1981. Use of cyclostat cultures to study phytoplankton light/dark cycles. I. Coccolithus huxleyi Physiol. ecology, p. 159-186. In P. Calcott [ed.] Continuous Plant. 20: 946-956. cultures of cells. Vol. II. CRC Press. (In press) 1968. Marine plankton algae grown with light/ RICHMAN, S., AND J. N. ROGERS. 1969. The feeding dark cycles. II. Ditylunt brightwellii and Nitzschia of Calanus helgolandicus on synchronously growing turgidula Physiol. Plant. 21: 66-77. populations of the marine diatom Ditylum brightwellii. PADILLA, G. M., AND J. R. COOK. 1964. The develop- Limnol. Oceanogr. 14: 701-709. ment of techniques for synchronizing flagellates, p. RICKETTS, T. R. 1979. The induction of synchronous cell 521-535. In E. Zeuthen [ed.] Synchrony in cell divi- division in Platymonas striata Butcher (Prasinophy- sion and growth. Interscience, New York, NY. ceae). Br. Phycol. J. 14: 219-223. PALMER, J. D. 1976. An introduction to biological rhythms. RIETH, A. 1939. Photoperiodizitât bei zentrischen dia- Academic Press, New York, NY. 375 p. tomeen. Planta 30: 294-296. PALMER, J. D., L. LIVINGSTON, AND D. ZUSY. 1964. RUSCH, H. P. 1969. Some biochemical events in the growth A persistent diurnal rhythm in photosynthetic capacity. cycles of Physarunz polycephalum. Fed. Proc. Fed. Nature (London) 203: 1087-1088. Am. Soc. Exp. Biol. 28: 1761-1770. PALMER, J. D., AND F. E. ROUND. 1965. Persistent SAVAGE, R. E., AND R. S. WIMPENNY. 1936. Phyto- vertical migration rhythms in the benthic microflora. plankton and the herring. Part II. 1933 and 1934. I. The effect of light and temperature on the rhythmic Fish. Invest. G.B. Ser. II, 15(1). 88 p. behavior of Euglena obtusa . J. Mar. Biol. Assoc. SCHEVING, L. E., AND F. HALBERG. 1980. Chronobio- U.K. 45: 567-582. logy. Sijthoff Intl. Publ., Netherlands. 312 p. 1967. Persistent, vertical migration rhythms in SCHMIDT, R. 1966. Intracellular control of enzyme syn- benthic microflora. VI. The tidal and diurnal nature thesis and activity during synchronous growth of of the rhythm in the diatoms Hantzschia virgata . Biol. Chlorella , p. 189-235. In I. L. Cameron and G. M. Bull. Woods Hole Mass. 132: 44-55. Padilla [ed.] Cell synchrony. Academic Press, New PAVLOVA, E. V., AND L. A. LANSKAYA. 1969. Energy York, NY. expenditures for movement in some Black Sea dino- SI MMER, J., AND M.. SODOMKOVA. 1968. Synchronization flagellates. Prog. Protozool. III: 174-175. of natural populations of Scenede.smus acuminatus PETERSON, D. F., AND E. C. ANDERSON. 1964. Quan- (Lagerh) Chod. induced by rhythmical changes in the tity production of synchronized mammalian cells in environment. Acta Univ. Carol. Biol. 1967: 251-254. suspension culture. Nature (London) 203: 642-643. SMAYDA, T. J. 1975. Phased cell division in natural PIRSON, A., AND H. LORENZEN. 1958. Ein endogener populations of the marine diatom Ditylum brightwellii zeitfaktor bei der teilung von Chlorella . Z. Bot. 46: and the potential significance of die phytoplankton 53. behavior in the sea. Deep-Sea Res. 22: 151-165. 1966. Synchronized dividing algae. Annu. Rev. SMITH, J. A., AND L. MARTIN. 1973. Do cells cycle? Plant Physiol. 17: 439-457. Proc. Natl. Acad. Sci. U.S.A. 70: 1263-1267. are PITTENDRIGH, C. S. 1975. Circadian clocks: what SOROKIN, C., AND R. W. KRAUSS. 1959. Maximum they?, p. 11-48. In J. W. Hastings and H.-G. growth rates in synchronized cultures of green algae. molecular basis of circadian Schweiger [ed.] The Science (Washington, D.C.) 129: 1289. rhythm. Life Sci. Res. Rep. 1. SOURNIA, A. 1974. Circadian periodicities in natural 1979. Some functional aspects of circadian pace- populations of marine phytoplankton. Adv. Mar. Biol. makers, p. 3-12. In M. Suda, O. Haysaishi, and H. 12: 325-389. Nakagawa [ed.] Biological rhythms and their central SPUDICH, J. L., AND D. E. KOSHLAND, JR. 1976. Non- mechanism. Elsevier/North Holland Biomedical genetic individuality: chance in the single cell. Nature Press, Amsterdam. 262: 467-471. POLLINGHER, U., AND C. SERRUYA. 1976. Phased divi- (London) sion of Peridinittm cinctum F. Westii (Dinophyceae) SPUDICH, J., AND R. SAGER. 1980. Regulation of the and development of the Lake Kinneret (Israel) bloom. Chlamydomonas cell cycle by light and dark. J. Cell J. Phycol. 12: 162-170. Biol. 85: 136-145. PRESCOTT, D. M. 1959. Variations in the individual STARKWEATHER, P. L. 1975. Diel changes in the feeding generation times of Tetrahymena geleii HS. Exp. behavior of Daphnia pulex Leydig (Crustacea: Bran- Cell Res. 16: 279. ciopoda). Ph.D. thesis, Dartmouth College, Hanover, 1976. Reproduction of eukaryotic cells. Academic NH. 190 p. Press, New York, NY. 177 p. STAYLEY, J. T. 1971. Growth rates of algae determined PREZELIN, B. B., B. W. MEESON, AND B. M. SWEENEY. in situ using an immersed microscope. J. Phycol. 1977. Characterization of photosynthetic rhythms in 7: 13-17. marine dinoflagellates. I. Pigmentation photosyn- STEEN BERGEN, C. L. M. 1975. Light-dependent morpho- thetic capacity and respiration. Plant Physiol. 60: 384- genesis of unicellular stages in synchronized cultures 387. of Scenedesmus quadricatula . Acta Bot. Neerl. 24: PREZELIN, B. B., AND B. M. SWEENEY. 1977. Charac- 391-396. terization of photosynthetic rhythms in marine dino- 1978. Pleomorphism of Scenedesmus quadri- flagellates. II. Photosynthesis-irradiance curves and cauda (Turp.) (Chlorophyceae) in synchronized

180 cultures. Mitt. Int. Ver. Theor. Angew. Limnol. 21: teristics of Chlorella. Carnegie Inst. Washington Publ. 216-223. 600: 76. STROSS, R. G., S. W. CHISHOLM, AND T. A. DOWNING. TAYLOR, W. R. 1979. Studies on the bioluminescent glow 1973. Causes of daily rhythmicity in photosynthetic rhythm of Gonyaulax polyedra. Ph.D thesis, Univ, rates of phytoplankton. Biol. Bull. 145: 200-209. Michigan, Ann Arbor, MI. 260 p. SUBRAHMANYAN, R. 1945. On the cell division and mitosis TERRY, O. W., AND L. N. EDMUNDS. 1970. Rhythmic in some South Indian diatoms. Proc. Indian Acad. settling induced by temperature cycles in continuously Sci. Ser. B 22: 331-354. stirred autotrophic cultures of E. gracilis (Z strain). SUDA, M., O. HAYAISHI, AND H. NAKAGAWA. 1979. Planta 93: 128-142. Biological rhythms and their central mechanism. VENKATARAMAN, G. S., AND H. LORENZEN. 1969. Elsevier/North Holland Biomedical Press, Am- Biochemical studies on Anacystis nidulans during sterdam. 265 p. its synchronous growth. Arch. Mikrobiol. 69: 34-39. SULZMAN, F. M., AND L. N. EDMUNDS. 1972. Persisting WEGMANN, K., AND H. METZNER. 1971. Synchroni- circadian oscillations in enzyme activity in non- zation of Dwialiella cultures. Arch. Mikrobiol. 78: dividing cultures of Euglena. Biochem. Biophys. 360-367. Res. Commun. 47: 1338-1344. WEILER, C. S. 1978. Phased cell division in the dino- SWEENEY, B. M. 1969. Rhythmic phenomena in plants. flagellate genus Ceratiwn: temporal pattern, use in Academic Press, New York and London. 147 p. determining growth rates, and ecological implications. SWEENEY, B. M., AND J. W. HASTINGS. 1958. Rhythmic Ph. D. thesis, Univ. California San Diego, La Jolla, cell division in populations of Gonyaulax polyedra. CA. 126 p. J. Protozool. 5: 217-224. 1980. Population structure and in situ division 1964. The use of the electronic cell counter in rates of Ceratiwn in oligotrophic waters of the North studies of synchronous cell division in dinoflagellates, Pacific central gyre. Limnol. Oceanogr. 25: 610-619. p. 579-587. In E. Zeuthen [cd.] Synchrony in cell WEILER, C. S., AND S. W. CHISHOLM. 1976. Phased division and growth. Interscience, New York, NY. cell division in natural populations of marine dino- SWEENEY, B. M., AND F. T. HAXO. 1961. Persistence flagellates from shipboard cultures. J. Exp. Mar. Biol. of a photosynthetic rhythm in enucleated Acetabularia. Ecol. 25: 239-247. Science (Washington, D.C.) 134: 1361-1363. WEILER, C. S., AND R. W. EPPLEY. 1979. Temporal pattern of division in the dinoflagellate genus Ceratium Sw ï FT, E., AND E. G. DuRBIN. 1972. The phased division- and its application to the determination of growth rate. and cytological characteristics of Pyrocystis spp. can J. Exp. Mar. Biol. Ecol. 39: 1-24. be used to estimate doubling times of their populations WILLE, J. J., AND C. F. EHRET. 1968. in the sea. Deep-Sea Res. 19: 189-198. Light synchro- nization of an endogenous circadian rhythm of cell SWIFT, E., M. STUART, AND V. MEUNIER. 1976. The division in Tetrahymena. J. Protozool. 5: 785-789. in situ growth rates of some deep-living oceanic WILLIAMS, F. M. 1971. Dynamics of microbial popula- dinoflagellates: Pyr6cystis fusiformis and Pyrocystis tions, p. 197-267. In B. C. Patten [ed.] Systems noctiluca. Limnol. Oceanogr. 21: 418-426. analysis and simulation in ecology. Vol. I. Academic TAMIYA, H. 1964. Growth and cell division in Chlorella, Press, New York, NY. p. 247-305. In E. Zeuthen [ed.] Synchrony in cell WILLIAMSON, C. E. 1980. Phased cell division in natural division and growth. Interscience, New York, NY. and laboratory populations of marine phytoplankton 1966. Synchronous cultures of algae. Annu. Rev. diatoms. J. Exp. Mar. Biol. Ecol. 43: 271-279 , Plant Physiol. 17: 1-26. WIMPENNY, R. S. 1938. Diurnal variation in the feeding TAMIYA, H. Y., Y. MORIMURA, M. YOKOTA, AND and breeding of zooplankton related to the numerical R. KUNIEDA. 1961. Mode of nuclear division in balance of the zoo-phytoplankton community. J. Cons. synchronous cultures of Chlorella: comparison of Perm. Int. Explor. Mer. 13: 323-337. various methods of synchronization. Plant Cell ZEUTHEN, E. 1974. A cellular model for repetitive and Physiol. 2: 383. free-running synchrony in Tetrahymena and Schizo- TAMIYA, H., K. SHIBATA, T. SASA, T. IWAMURA, AND saccharomyces, p. 1-30. In G. M. Padella, I. L. Y. MORIMURA. 1953. Effect of diurnally intermittant Cameron, and A. Zimmerman [ed.] Cell cycle controls . illumination on the growth and some cellular charac- Academic Press, New York, NY.

181 Nitrogen Metabolism of Microalgae

P. J. SYRETT Department of Botany and Microbiology, University College of Swansea, Swansea, Wales, UK

Introduction by oceanic phytoplankton must be of the order of 6 x 10" t per annum (Fogg 1978). Study of the nitrogen metabolism of algae may be said to have started some 60 years ago when Nitrogen Sources for Algal Growth Warburg and Negelein (1920), following Warburg's introduction of the use of the alga Chlorella for GASEOUS NITROGEN studying photosynthesis, showed that light stimulated nitrate reduction by this organism. Since then studies Microscopic algae include both procaryotes and of algal nitrogen metabolism have been published eucaryotes although many would now regard the at an ever-increasing rate. The topics studied have procaryotes, the blue-green organisms (Cyano- diversified and just as Warburg's initial work with phyceae or Cyanobacteria) as bacteria. The ability Chlorella led to an understanding of the fundamentals to fix gaseous nitrogen is confined to procaryotes of photosynthesis so have later studies on algal and, among algae, only some blue-greens are able nitrogen metabolism thrown light on fundamental to do so. Fogg (1978) has recently reviewed nitrogen cellular processes. fixation in the oceans. It is not known with certainty Along with the diversification of topics has been how much occurs but it cannot be very great because a great increase in the number of different species nitrogen-fixing microorganisms are not abundant that have been studied. It remains true, however, in oceans. The best-known planktonic blue-green that there are few detailed laboratory studies of truly alga which almost certainly fixes nitrogen is Tri- planktonic algae and this is largely because bio- chodesmium (Oscillatoria) which can form extensive chemical physiologists prefer to work with organisms blooms in tropical waters. Nevertheless Fogg's very that can be grown readily and axenically in labo- tentative estimate of the quantity of gaseous nitrogen ratory culture. These conditions still impose a con- fixed in the oceans is 0.1 x 10" t per annum or siderable limitation on the variety of algae that can about 1% of the total annual N incorporation by be used. For these reasons therefore this article is phytoplankton. Even this estimate may be too high about the nitrogen metabolism of microalgae, rather and it is not clear what has limited the development than of phytoplankton only, and one presumes that of autotrophic marine nitrogen-fixing organisms. some of our knowledge of the nitrogen metabolism Biochemically, nitrogen fixation takes place by of algae such as Morelia , Chlamydomonas , Phaeo- reduction of N2 to Ne, the reaction being catalyzed dactylum , and Anacystis will apply to their much by nitrogenase. Six electrons are required for each less studied oceanic cousins. N2 molecule reduced; these are derived from a low After carbon and excluding hydrogen and potential reductant such as reduced ferredoxin. ATP oxygen that can come from water, nitrogen is quan- is also required, most probably 12 molecules for titatively the most important single element contri- each N., molecule (Mortenson and Thorneley 1979). buting to the dry matter in algal cells. The proportion In blue-green algae both reduced ferredoxin and of nitrogen as a percentage of dry weight can vary ATP can probably be generated either by photo- from 1 to 10%. It is low in diatoms where the silica chemical reactions or by dark respiratory reactions in the cell wall makes a substantial contribution to (Gallon 1980). dry matter and in nitrogen-deficient organisms that have accumulated large amounts of carbon com- COMBINED INORGANIC NITROGEN pounds such as oils or polysaccharides (Fogg and Collyer 1953). But in an exponentially growing non- Almost all chlorophyll-containing algae studied diatomaceous microalga, nitrogen is about 7-10% in culture will grow with either nitrate (NO), nitrite of the dry matter and carbon about 50% (Vaccaro (NO.',), or ammonium (NH; ) as a nitrogen source. 1965). The C:N ratio is therefore around 5. As es- Care has to be taken to avoid large changes in pH timates put the primary productivity of the oceans at and the toxic effects of high concentrations of NO.', 3.1 x 10" t per annum due to phytoplankton and sometimes of NH1- . Maximum growth rates are photosynthesis (Platt and Subba Rao 1975), this generally much the same with either NO :', or NFIT suggests that the total amount of nitrogen assimilated as N source (see Table 1) but a careful study by

182

TABLE 1. Comparison of the growth rates of some algae cultured on various N sources (from Leftley 1980).

Specific growth rate constant Concentration of N (k) (log It, units • d') (mmol L- ')

Organism NHt NO3 Urea NI-it NO;;" Urea Reference

Cyanophyta Agmenellum quadruplicatum 3.3 3.0 1.7 0.6 variable 4.3 Kapp et al. (1975) Anabaena variabilis 0.69 0.69 0.81 3.0 10.0 1.0 Kratz and Myers (1955) Nostoc muscorum 0.48 0.50 0.54

Chlorophyta Chlorella ellipsoidea 0.50 0.40 - 12.0 66.0 Samejima and Myers (1958) Chlorella pyrenoidosa 0.5 0.45 0.47 13.0 "

Chrysophyta Amphiphora alata 1.14 1.30 1.30 0.„ 1 0,., 1 Carpenter et al. (1972) Chaetoceros simplex 1.54 1.58 1.35 Chaetoceros sp. 0.96 1.30 1.03 Chrysochromulina sp. 1.12 1.14 1.35 Cyclotella cryptica 0.57 0.57 0.49 0.1 0.9 0.2 Liu and Hellebust (1974) Skeletonema sp. 1.41 1.66 0.99 0.1 0.1 0.1 Carpenter et al. (1972) Stephanopyxis costata 1.23 1.35 1.58

Paasche (1971) established that Dunaliella tertiolecta diatoms and 3 coccolithophorids, 16 grew on gluta- grows 10-30% faster on NH; than on NO. The mine and 12 on urea. Growth rates in urea are gener- difference may partly reflect the greater need for ally similar to those in N0f3 or NH; (Table 1). Gluta- reductant (ultimately photogenerated) for assimilation mine contains two N atoms and Chlorella can utilize of NO (3 but the NH;-grown cells also have higher both for growth although unable to grow on glutamic ribulose biphosphate carboxylase activities. Although acid (Lynch and Gillmor 1966). The explanation ap- laboratory algae grow well on NO: , Antia et al. pears to be that glutamine can enter the cells of this (1975) point out that this may be partly a consequence Chlorella strain and hence be metabolized whereas of the fact that they have almost always been isolated glutamic acid cannot. The alga Monodus may be from natural populations by selective growth on somewhat exceptional in producing an extracellular media containing Indeed Antia et al. (1975) glutaminase which hydrolyzes glutamine to NH: and showed that one marine species, Hemiselmis vi- glutamic acid; the organism uses the NH; released for rescens , that had been isolated by Droop in a medium growth but not the glutamic acid (Belmont and Miller containing glycine and purines, was exceptional 1965). among the 26 planktonic species they studied in being unable to utilize or N0',. Amino acids, particularly glycine, serine, ala- nine, glutamic acid, and aspartic acids, also serve as AMIDES AND AMINO ACIDS N sources for the growth of some algae, although of the diatoms studied by Wheeler et al. (1974) it Thomas (1968) has compiled useful tables was nearshore species of Nitschia and Phaeodac- summarizing the information that was available up tylum that most readily grew on these compounds. to then about the utilization of organic N by fresh- Uptake mechanisms for some amino acids, of high water and marine algal species. Additional studies affinity (i.e. half saturation at less that 10 p,M), with marine forms have been carried out by Wheeler have been demonstrated in pennate diatoms and in et al. (1974), Antia et al. (1975), and Neilson and Platymonas (Hellebust and Lewin 1977; Wheeler Larsson (1980). In general, the amides, urea, gluta- et al. 1974). This suggests an ecological significance mine, and asparagine are good sources of nitrogen for for amino acid utilization but sometimes an organism algal growth. For example, Antia et al. found that 23 will take up an amino acid rapidly while not being of the 26 marine species they studied could grow with able to grow on it, e.g. Thalassiosira with lysine urea. Guillard (1963) found that of 16 marine centric (Wheeler et al. 1974).

183 PURINES AND PYRIMIDINES TABLE 2. Nitogcn sources for algal growth with oxidation states (± one electron is required to change the oxidation Study of the use of purines as nitrogen sources state of one N atom by 1 unit). is hindered by their low solubility in water. Never- theless uric acid, xanthine, guanine, or adenine can be used as sole N sources by some blue-greens, Oxidation chlorophytes, and diatoms (Van Baalen 1962; Birdsey state and Lynch 1962; Cain 1965). Antia and Landymore (1974) point to the instability of uric acid and xanthine ,i +5 in seawater which means that in nature it may be Nitrate ions NO these compounds that are degradation products of Nitrite ions NO+3 used. Antia et al. (1975) tested the more stable purine, hypoxanthine, and found that 17 out of 26 marine Nitrogen gas N2 strains tested grew well on it, although some did so only after a very long lag period (e.g. about 30 d Ammonium ions NFIt } —3 with Isochrysis). Urea Pyrimidines appear to be much less utilizable as N sources for growth (Cain 1965). Nevertheless Glutamine — 4 (see text) the soil alga Chlorella has an active mechanism for Amino acids — 5 (see text) the uptake of uracil (Knutsen 1972). Purines GLUCOSAMINE emphasize this in Table 2, amino acids (and purines Mention should be made of this compound in which the N is derived from amino N) are shown Antia et al. (1975) found to be used as a N source at an oxidation state of —5 although, overall, amino for growth by several marine phytoplankters although acids, like alanine, are at the same oxidation level none grew well on it. as carbohydrate plus NH;. TAXONOMIC DIFFERENCES There appears to be no clear relationship between Sources of Combined Nitrogen the ability to use a particular group of nitrogen com- in Seawater pounds and taxonomic class. The work of Antia et al. (1977) summarize naturally al. (1975) shows that within both the chlorophytes Parsons et combined nitrogen and diatoms some species (e.g. Tetraselmis maculata occurring concentrations of as NO < 0.01-50 i.LM, and Nitzschia acicularis) show ability to utilize a compounds in seawater urea < wide range of organic N sources while other species NO <0.01-5 tM, NH ; <0.1-5 bt/d, M. In the (e.g. Nannochloris oculata and Cyclotella cryptica) 0.1-5 4/, and amino acids < 0.2-2 amino acid present cannot do so. Possibly, as Naylor (1970) suggests, North Atlantic, alanine was the most widely dis- more extensive studies will reveal evolutionary in highest concentration and the relationships but it may well be that ability to utilize, tributed (Pocklington 1971). It is obvious that there or lack of it, depends on rather small genetic dif- must be considerable variation in the concentrations ferences. of nitrogenous compounds both in space and in time. Higher concentrations, sometimes much higher, are likely around coasts and when rates of phytoplank- STATES OXIDATION ton metabolism are low. Information about concen- The nitrogen sources for algal growth occur in trations of purines in seawater appears to be lacking widely different states of oxidation (Table 2). Thus, although their ready utilization by several algal spe- if nitrogen gas is regarded as an oxidation state of cies makes them of interest. zero, N(:)! is the most oxidized inorganic form (oxi- dation state, +5) and NH; the most reduced (oxi- dation state, —3); this means that eight electrons Assimilation of Nitrogen Compounds are necessary to reduce a nitrogen atom in the form of NO,', to NH;. Amide N can also be regarded as Figure 1 depicts a much idealized algal cell at an oxidation state of —3 since no oxido/reduction setting out some of the major features of nitrogen occurs in the formation of an amide from NHL assimilation that will be discussed in more detail However, the formation of an amino acid from an later. The diagram shows all nitrogen sources crossing oxo-acid and NH; is a reductive reaction which the cell wall and outer cell membrane (the plasma- requires two electrons per N atom incorporated. To lemma) to enter the cytoplasm. For all the com-

184 NO ' ----> NO ' 3 1 3 • sk Vacuol e NO .E mitochondrion 2

th /

Urea —> Urea % r [WI:I ',P

CO ' V 2 se amino acids , 401 (C) n'------... amino NH + --> NH 1. acids .•.....,,.... 4 4 amino acids

chloroplast

FIG. 1. Main features of nitrogen assimilation in a eucaryotic algal cell. See text for discussion. pounds depicted, there is evidence for active uptake metabolized (Czygan 1965); firm evidence for such systems dependent on metabolism although there reactions is lacking. Similarly there are suggestions may be passive entry by free diffusion as well. The for a direct incorporation of urea N into organic diagram shows exchange between the compounds compounds (Kitoh and Hori 1977); again the bulk in the cytoplasm and compounds in the vacuole. of the evidence is against this. Third, the vacuole It also shows the involvement of the chloroplast is depicted as containing a metabolically inactive and, to a lesser extent, of the mitochondrion. One solution into and from which solutes pass. But in of our main areas of ignorance about the nitrogen higher plant cells the vacuole contains hydrolytic metabolism of microalgae is that we do not know, enzymes (Boller and Kende 1979). Whether such with any certainty, exactly where in the cells the enzymes are present in algal vacuoles and, if so, various metabolic reactions are located nor, indeed, whether they affect nitrogen metabolism is unknown. where the metabolic "pools" are. We do not know, therefore, what part the internal membranes sur- Assimilation of Ammonium (NI-n) rounding the vacuole, chloroplast, and mitochon- drion play in the regulation of metabolism but we FORMATION OF GLUTAMIC ACID suspect it may be an important one. The figure contains a number of assumptions Studies of NH; assimilation have been much which will not be discussed again. The first is that aided by the use of nitrogen-deficient Chlorella nitrate transport across the membrane(s) and nitrate cultures. Cultures that are deprived of nitrogen go reduction in the cytoplasm are separate events; in on accumulating polysaccharides by photosynthesis. contrast, Butz and Jackson (1977) have argued that When such cultures are supplied with NH; ions, the enzyme nitrate reductase is located in cellular they are taken up rapidly and converted to organic N membranes and has the dual functions of both trans- compounds. At the same time the accumulated port and reduction of nitrate. Second, the nitrogen polysaccharide in the cells disappears, partly because in nitrate, nitrite, and urea is shown as being con- of a greatly increased role of respiration during the verted into NH; before being assimilated into or- rapid assimilation of NI-1.; , but largely because it is ganic N compounds. There have been suggestions converted into organic N compounds (Syrett 1953a, of a partial reduction of nitrate and nitrite to the b; Hattori 1958; Reisner et al. 1960). During the first level of hydroxylamine followed by combination with 15 min of NH; assimilation the greatest increases an organic acid to form an oxime that is then further of organic N are in glutamine and alanine with other

185 amino acids showing smaller increases (Reisner et At the time these experiments were carried out al. 1960) and, in experiments in which 'NH; was they were interpreted as being consistent with the fed, after 1 min the greatest incorporation was into then accepted mechanism of ammonium assimilation the amide group of glutamine (36.8 atom% excess), by plants. The key compound was glutamic acid followed by the amino groups of glutamine and which was formed from NH; and a-oxoglutaric glutamic acid (16% atom% excess), with alanine acid by the reductive reaction catalyzed by glutamic showing lower labeling (6.5% atom% excess) (Baker dehydrogenase (GDH) namely: and Thompson 1961).

COOH COOH glutamic CH 2 dehydrogenase (GDH) CH2 (1) I + NH3 + NAD(P)H + + NAD(P) + + H2 0 CI-I2 2 CH

CO CHNH,

COOH COOH

a-oxoglutaric acid glutamic acid

Other amino acids, including alanine, were native pathway of NH; assimilation (Tempest et al. regarded as being formed from glutamic acid by 1970; Brown et al. 1974) and Miflin and Lea (1976) transamination and glutamine by the further addition showed that the pathway could also operate in leaves of NH; to the y-carboxyl group of glutamic acid, of higher plants and in blue-green and green algae. a reaction requiring ATP and catalyzed by the enzyme In this pathway, glutamine rather than glutamic acid glutamine synthetase (GS) (reaction 2 below). All is the first product of NH; assimilation and the the necessary enzymes were known to be present NH; incorporated into the amide group of glutamine in Chlore/la (Baker and Thompson 1961). is then transferred to a-oxoglutaric acid in a re- However, in the early 1970s work by Brown ductive reaction from which two molecules of glu- and his colleagues with bacteria revealed an alter- tamic acid result; thus COOH CONH2

CH2 2 CH glutamine synthetase (GS) (2) CH2 + NH3 + ATP CH, + ADP + Pi

CHNH2 CHNH2

COOH COOH

glutamic acid glutatnine GOGAT

(3) COOH COOH [2F11 CH2 CH2

2 CH CH2

CHNH2 CO

COOH COOH

glutamic acid a-oxoglutaric acid

186 The enzyme catalyzing reaction (3) is glutamine- 2000 NH4' produced oxoglutarate aminotransferase or glutamate synthase; with NO,' added/ 1, it is often called GOGAT. GOGAT catalyzes a 1600-1 reductive reaction and, in higher plants and algae, the reductant can be either NADH or reduced ferre-

doxin. By reactions (2) and (3) operating in sequence, 1200 — NH ; is assimilated into glutamic acid by combination with a-oxoglutaric acid but GDH is not involved. The produced with no addition overall reaction is still reductive, as is reaction (1), but 800 _AM it also requires ATP which reaction (1) does not. It is NH (B - A) • 4 particularly attractive to think of reaction (2) as being o -11 the major reaction responsible for the incorporation of 400-1 • NH ; into organic combination because US has a much /0/ higher affinity for ammonium than does GDH (Miflin • and Lea 1976). In a number of marine phytoplankters, O 24 48 for example, the apparent K,, (concentration sup- HOURS porting half maximum velocity of reaction) of GDH for NM- was 4 500-10 000 ,tx./14 (Ahmed et al. 1977). FIG. 2. Ammonium production by illuminated MS0- For Skeletonema , Falkowski and Rivkin (1976) found (1 mM)-treated Chlamydomonas Cell density, 1.3 x 10' a K„, for GDH of 28 000 gM; for US it was only cells/mL. The horizontal broken line shows the quantity 29 ,ti,M which is similar to the value for higher plant of NH1- expected from the complete reduction of the NICI enzymes (Miflin and Lea 1976). added to B (unpublished data of C. R. Hipkin, T. A. V. Rees, and S. A. Everest). GDH OR GS /GOGAT? Much work has now been done with higher caldarium, where NH; assimilation is inhibited by plants and some with algae to discover whether NH; either MS0 or azaserine and again, in the presence assimilation mainly takes place through GDH or of MSO, N0f, is reduced quantitatively to NH./ through the GS /GOGAT reactions. One approach which accumulates. Inhibition of US activity and has been to make use of inhibitors. L-Methionine- of NH; assimilation by MSO also occurs in Chia- DL-sulphoximine (MSO or MSX) forms an analogue mydomonas (Cullimore and Sims 1981; C.R. Hipkin, of the y-glutamyl phosphate enzyme complex of T. A. V. Rees, and S. A. Everest unpublished data). US and is a powerful inhibitor of US. Azaserine is But here a new phenomenon is seen, namely a sub- an analogue of glutamine and inhibits GOGAT. stantial production of NH ; in the presence of MS0 in These substances do not inhibit GDH (Miflin and the absence of any added nitrogen source; NH t pro- Lea 1976). If ammonium assimilation is primarily duction is increased on addition of NO .'2 (Fig. 2). by the GS /GOGAT pathway it should be possible The other approach that has been used is to to inhibit it substantially with these inhibitors. That survey algae for the presence of the key enzymes, this is so has been shown for the blue-green alga GDH, US, and GOGAT, and to estimate their ac- Anacystis , where azaserine and MSO at 1 triM con- tivities. These studies can sometimes be misleading centration caused 90% inhibition of NH; assimilation because an enzyme activity may appear to be absent by illuminated cells (Ramos et al. 1980a). MS0- or low in cell-free extracts because of failure to treated Anacystis reduced nitrate almost quantitatively establish optimal conditions for assay. At present to NH; which accumulated because it could not such studies are few. Substantial US and GOGAT be assimilated (Ramos et al. 1980b). MS0 at a activity has been measured in Platymonas (Edge concentration of 2.0 ,itM MSO was sufficient to and Ricketts 1978), Chlorella fusca (Lea and Miflin prevent growth with NO :', , NH it- , glutamate, or alanine 1975), and Chlatnydomonas reinhardii (Cullimore as N source; normal growth occurred in the presence and Sims 1981; S. A. Everest unpublished data). In of 2.0 p,M MS0 and 1 mM glutamine (Ramos et Platymonas and Chlatnydomonas , GDH activities, in al. 1980a). When illuminated Anabaena , which contrast, were very low. In other algae, e.g. Chlorella fixes N., , was treated with MSO, the nitrogen fixed fusca, GDH is present with a high activity which is accumulated as NH; (Stewart and Rowell 1975). higher in NH ; -grown cells than in NO!rgrown ones Meeks et al. (1978) showed with three other species and highest of all in N-deficient cells (Morris and of blue-green algae that 1 mM MSO almost com- Syrett 1965). This strain of Chlorella (8p) contains pletely inhibited incorporation of '3 N1-1,3 (°N is a only a NADP- dependent GDH. Shatilov et al. (1978) short-lived radioactive isotope). have shown the presence in a thermophilic strain of Rigano et al. (1979, 1980) have made similar Chlorella 82T of two glutamic dehydrogenases, one observations with the eucaryotic alga Cyanidium that is constitutive and functional with either NADH or

187

NADPH, and another that is present only in NH ; - found in eucaryotic algae; it resembles higher plant grown cells and specific for NADPH. The Russian nitrate reductases in many respects. It is found in authors favor the view that in Chlorella assimilation of the soluble portion of cell-free extracts and although NH ; at high concentrations is mainly through gluta- there have been suggestions for an association of mic dehydrogenase while at low concentrations or in it with chloroplasts, possibly in between the inner NQ3 -grown cells it is through GS /GOGAT (Shatilov and outer membranes (Grant et al. 1970), convincing et al. 1978). Conclusive evidence for this view is evidence is lacking. lacking. The enzyme catalyzes:

ALANINE DEHYDROGENASE nitrate reductase NICt + NAD(P)H + H+ Of other pathways of NHT assimilation, the NC), + NAD(P) + + H2 0 direct formation of alanine from NH; and pyruvic acid is significant in some blue-green algae, those The most highly purified and best characterized that lack GDH activity (Neilson and Doudoroff algal enzyme is from Chlore/la (Solomonson et al. 1973). Rowell and Stewart (1976) characterized the 1975; Solomonson 1979; Gin i and Ramadoss 1979) but enzyme from Anabaena cylindrica; they suggest it has also been studied in detail after 200-fold purifica- that because of its high K„, (> 8 mM) it is likely tion from the diatom, Thalassiosira (Amy and Garrett to be of less importance in ov.erall ammonium assimi- 1974). The Chlore/la enzyme has a molecular mass of lation than OS. It is interesting that of three species about 350 000 and a complex structure. It contains of blue-greens studied by Meeks et al. (1978), incor- molybdenum, haem, and flavin adenine dinucleotide, poration of 13NH3 into alanine was inhibited greatly most probably two molecules of each in each molecule by MS0 in two species but not at all in the third, of enzyme. Molybdenum appears to be at or close to Cylindrospennum lichenifonne , suggesting that in the site at which NO ,'3 is reduced and the enzyme works this organism alanine is formed directly and not via by molybdenum being reduced by electrons derived the GS /GOGAT pathway as in the two inhibited from NADH, possibly to Me', and then reoxidized, species. Alanine dehydrogenase is also present in possibly to MO'', by NO,' which is consequently Chlamydomonas where, during the life cycle, its reduced to N0'2 . Substitution of molybdenum by activity changes in an inverse way to that of glutamic tungsten produces an enzyme which cannot reduce dehydrogenase (Kates and Jones 1964). NO ;i but still retains its ability to reduce cytochrome c (see below) (Vega et al. 1971; Paneque et al. 1972). CARBAMYL PHOSPHATE Growth of algae on a medium with tungstate present to Another subsidiary pathway of NElt assimilation produce an inactive nitrate reductase has proved a most is that which results in carbamyl phosphate and hence useful experimental tool (e.g. Solomonson and Spehar incorporation into citrulline (and then arginine), 1977; Serra et al. 1978b). With the higher plant en- pyrimidines (including thiamine), and biotin; it is zyme there is evidence that the molybdenum is con- probable that in plants the ammonia for carbamyl tained in a small complex (molar mass less than phosphate synthesis is derived from the amide group 30 000) that can be fairly easily separated from the of asparagine or glutamine rather than from free bulk of the enzyme (Notton and Hewitt 1979). Haem is NM- ions (Beevers 1976). probably present as a b-type cytochrome which is reduced by NADH and reoxidized by NO:; cyanide Assimilation of Nitrate (NOD stops its reoxidation by NQ'i but not its reduction by and Nitrite (NO) NADH (Solomonson 1979). Solomonson's tentative picture of the structure of Chlore/la nitrate reductase is Assimilation of takes place by reduction shown in Fig. 3 although this does not show molyb- of NO!i to NO followed by a reduction of NO .'2 to denum in a component of fairly small molecular NHI which is then converted to organic compounds. weight. This figure shows a characteristic feature of The first step requires two electrons and the second, nitrate reductase frotn eucaryotic cells, namely, that it six.

2e 6e > NO.> NH; nitrate reductase - nitrite reductasé NADH 3 (NR) (NiR)

NITRATE REDUCTION MV

There are two types of nitrate reductase known FIG. 3. Structure of Chlore/la nitrate reductase according in algae. The first, and better known enzyme, is to Solomonson (1979).

188 TABLE 3. Pyridine nucleotide specificity of nitrate reductase from eucaryotic algae. (All enzymes studied can utilize NADH for nitrate reduction. The table therefore lists those able or not able to utilize NADPH.)

Utilizes NADPH Does not utilize NADPH

Rhodophyceae Cyanidium caldarium" Porphyridiwn aerugineum (1380/2) 1' Porphyridium cruentwn (1380/12)e

Haptophyceae Isochrysis (927/ De Coccolithus huxleyi"

Dinophyceae Gonyaulax polyedra"

Prasinophyceae Platymonas subcordiformis (161/12)e

Bacillariophyceae Phaeodactylum tricornututn (1052/6)e Cyclotella nana" Ditylwn brightwellii" Thalassiosira pseudonanag Skeletonema costatum f (NADPH used poorly)

Chlorophyceae Chlorella variegata (211/10a)c Chlorella (211/8p)e Chlorella variegata (211/10d)e Chlorella fusca (211/15)c Ankistrodesmus braunii (202/7c)c Chlorella vulgaris (211/11b)C Chlamydomonas reinhardii (2192)C Chlorella stigmatophora (211/20)C Dunaliella primolecta (11/34)c Scenedesmus obliquus (276/3a)r Dunaliella tertiolecta (19a)e Scenedesmus obliquus (276/3b)c Dunaliella parvag

uRigano (1971); bRigano et al. (1979b); cHipkin et al. (1979); 'Eppley et al. (1969a); eAmy and Garrett (1974); Serra et al. (1978c); "'Helmer (1976).

has three enzymic activities that can be measured and logical electron donor. There is no convincing evi- distinguished experimentally. The first is the ability to dence that, in algae, bispecificity of pyridine nucleo- reduce nitrate with NAD(P)H; this activity requires tide utilization is due to the presence of two different the full functional enzyme. The second is the ability to nitrate reductases as is so in some higher plants reduce NOr, with electron donors such as reduced (Campbell 1976). Attempts to separate the two methyl viologen or reduced flavin mononucleotide activities in extracts of Dunaliella parva were un- (FMN). The third is cytochrome c reductase activity, successful (Heimer 1976). i .e. the ability to reduce cytochrome c with NAD (P)H. The second type of nitrate reductase is found In some mutants of Chlamydomonas the enzyme has in procaryotic cells and, in the algae, in the blue- lost the ability to utilize NAD(P)H but it can still greens (Cyanophyceae). In broken cell preparations reduce NO with methyl viologen or FMN (Nichols et of blue-green algae the enzyme is associated with al. 1978). However, such mutants cannot grow with chlorophyll-containing particles but it can be solu- NOr, as a N source which suggests that, in vivo, the bilized and purified (Manzano et al. 1978; Losada enzyme functions only with NAD(P)H as electron and Guerrero 1979). The enzyme is smaller than the donor for nitrate reduction. eucaryotic one, with a molecular mass of about The pyridine nucleotide specificity of the en- 75 000. It contains molybdenum but not flavin or zyme differs in different algae (Table 3). It can be cytochrome (Losada and Guerrero 1979). The im- seen that the enzyme is always active with NADH portant difference from the enzyme of eucaryotes is as electron donor and, in many algae, only with that it does not use pyridine nucleotide as electron this. But in Cyanidium and in several species of donor but reduced ferredoxin. It therefore catalyzes Chlorophyceae it is also active .with NADPH and the reaction: for some, e.g. Chlamydomonas and Dunaliella , one suspects that NADPH may be the natural physio- NO(, + 2Fdi1.d + 2H+ --> NO2 + 2Fd„, + H2 0

189 NITRITE REDUCTION A number of algae that grow well on urea do not contain this enzyme. Instead they contain a urea to NM- is catalyzed by The reduction of NO carboxylase and also allophanate hydrolase. Together nitrite reductase. This enzyme appears ferredoxin these enzymes catalyze: to be much the same in algae and in leaves of higher plants. It is a small molecule with a molecular mass Mg2+ K+ of 60 000-70 000. In algae it has now been studied in urea + ATP + HCCIfi urea carboxylase Anabaena (Hattori and Uesugi 1968), Dunaliella (Grant 1970), Ditylunt (Eppley and Rogers 1970), allophanate + ADP + Pi Chlorella (Zumft 1972), Porphyra (Ho et al. 1976), (Llama et al. 1979). The enzyme allophanate 2NH3 + 2CO2 and Skeletonetna allophanate hydrolase contains sirohaem which is an iron tetrahydropor- phyrin (Murphy et al. 1974) and where NO .'2 probably The overall reaction catalyzed is: attaches, it also contains an iron—sulphur center which participates in electron transport (Losada and Guerrero CO(NH2 )2 + ATP + H2 0 1979). The reaction catalyzed is: CO2 + 2NH3 + ADP + nitrite reductase NO:; + 6Fe, + 8H+ This result is the same as that achieved by the NH ; + 6Fe1,, + 2H2 0 urease reaction but at the expense of a molecule of ATP per molecule urea. The overall reaction is With the purified enzymes reduced methyl viologen referred to as the urea amidolyase (UAL-ase) reaction. will also act as an electron donor but pyridine nucleo- After its initial description by Roon and Levenberg tides are inactive. Despite the reduction of NOE.; (1968) it has been studied in detail in yeast (Whitney to NHI- requiring six electrons per NO ion, no and Cooper 1972, 1973), Chlorella (Thompson and intermediates have been detected in this reaction. Muenster 1971, 1974), and Chlawydomonas (Hodson In the leaves of higher plants, nitrite reductase et al. 1975). appears to be located in the chloroplasts (Miflin The reaction catalyzed by urea carboxylase is 1974); information about its localization in algal a biotin-dependent carboxylation which, like other cells is lacking but the quenching of chlorophyll such carboxylations, is inhibited in cell-free prepa- fluorescence following addition of NO.; (but not of rations by avidin: the inhibition is relieved by addition NO: ) to cells of Chlorella and Ankistrodesnuts of biotin. This reaction is not much inhibited by (Kessler and Zumft 1973) suggests a close linkage hydroxyurea whereas the urease reaction is. These between nitrite reduction and the photochemical differences in inhibition properties, together with reactions in the chloroplasts. the requirement of the UAL-ase reaction for ATP, allows the UAL-ase and urease reactions to be dis- tinguished in crude cell-free extracts and using these methods Leftley and Syrett (1973) and Bekheet and Metabolism of Urea Syrett (1977) surveyed the distribution of the urea- degrading enzymes in a number of algae. The result Urea is an excellent nitrogen source for algae is fairly clear-cut. If an alga can metabolize urea in culture and may well be an important natural it has one enzyme or the other, not both, although source in seawater (McCarthy 1972). Nitrogen makes it must be said that the methods would not detect up almost half of the mass of the urea molecule, a low activity of one enzyme in the presence of CO(NH., ) 2 . The work of Allison et al. (1954) on the other. The occurrence of UAL-ase is restricted feeding [ "C]urea to Nostoc suggested that carbon to certain orders of the Chlorophyceae namely the from urea was assimilated by algae only after its con- Volvocales, Chlorococcales, Chaetophorales, and version to CO 2 and most subsequent workers agree Ulotrichales. The other chlorophytes examined with this view (e.g. Hattori 1960). There are, how- contained urease as did members of all the other ever, suggestions that urea N may be assimilated into algal classes. As far as phytoplankton is concerned organic N without prior conversion to NEI :t (Kitoh and a generalization based on the limiteél evidence avail- Hori 1977). able at present is that the chlorophytic phytoplankters Two enzymes which metabolize urea are now are likely to contain UAL-ase and the others, urease. known in algae. The first is urease , an enzyme The hydrolytic reaction catalyzed by urease goes which is widespread in the plant kingdom. It cata- to completion and it is not clear what biological lyzes: advantage, if any, is gained by using ATP to ac- complish the sanie overall reaction. Leftley showed CO(NH2)2 + H2 0 CO2 + 2NH:3 that whole cells of Phaeodactyluni (containing urease)

190 ▪

could metabolize urea at low concentrations just as It has recently been shown that purified urease well as cells of Dunaliella (with UAL-ase): both had a from jack bean contains nickel (Dixon et al. 1975). half-saturation value of about 1.5 btAl urea (Syrett and Rees and Bekheet (unpublished results) have shown Leftley 1976). One advantage would be if the interme- a strong dependence on nickel for the development diate allophanate could be assimilated into metabolism of urease activity in Phaeodactylum and Tetraselmis . but evidence for this is lacking. Another possibility is In the absence of nickel these organisms appear to that UAL-ase levels can be regulated more readily than overproduce a urease protein because the restoration those of urease. Certainly in the few algae that have of nickel after a period of deprivation leads to a very been examined UAL-ase behaves as an inducible/ rapid "overproduction" of urease activity (Fig. 4). repressible enzyme in a way that urease does not (Syrett and Leftley 1976).

0 Tetraselmis subcordiformi s

a) 40 H à No cycl oh ex i mi de .2+ o with Ni s- e) at 24 h A cycl oheximi de z ..— 2+ I( ' added wi th Ni at 24 h eL

-o

rd 0.1 20 -4 71;

Ni 2+ 10 citrate + Ni 2+ Arm, 4

xi) , .2+ à LL1 citrate - Ni J O-

12 24 36 48 HOURS

FIG. 4. The effect of Ni" ions on urease activity in Tetraseltnis subcordifortnis. Cultures were grown with added citrate and either ± Ni". After 24 h Ni" was added to a —Ni" culture. Urease activity was determined in cell-free extracts at the times shown. Note that addition of cycloheximide with Ni" after 24 h had no effect on the marked and rapid increase in urease activity (I.A. Bekheet and T. A. V. Rees unpublished data).

191

NH 3 o to allantoic acid is widely distributed in algae (Villeret H 1955, 1958). Two enzymes are known in animais FIN" H and bacteria, which metabolize allantoic acid. One HLtsr-CNN is allantoicase catalyzing the conversion of allantoic H H acid to one molecule of urea and one of ureido- Adenine Guanine glycolate; the other is allantoate amidohydrolase 1-N H3 I - N H 3 which catalyzes the formation of 2NH 3 , 1CO2 , and 0 o 1 ureidoglycolate. The enzyme ureidoglycolase

HA\ converts ureidoglycolate into urea plus glyoxylic Ts-'2 I I II \ CH CH acid. Little is known of these enzymes in algae. xanthine oxidase /C."N" N H H H Villeret's studies (1955, 1958) suggested a rather Hypoxanthine limited distribution of allantoicase among algae. Xanthine 1+02 In fungi some of these enzymes, e.g. uricase xanthine oxidase 0 OH and allantoinase, are inducible and their formation is repressed by NH; (Vogels and van der Drift 1976). HNC Information is lacking for algae about regulation II c—o 1 0 C—OH HO N H of these enzymes but the long periods of adaptation H H necessary before some algae can grow on hypo- Uric acid (keto form) Uric acid lenol form) xanthine (Antia et al. 1975) strongly suggests that I I uricase CO2 +03 induction of new enzyme systems is necessary. 0 H3 NI NI H2 0H I I H 2 N C.- N\ PYRIMIDINES allantoinase 0=C COON C =0 0..=.j / C oe° ---I- 71- 27), .- I I I These are not good nitrogen sources for algal

allantoicase growth but uracil is degraded by Chlorella to CO, Allantoin H 0 and 0-alanine with dehydrouracil and p-ureidopro- Allantoic acid pionic acid as intermediates (Knutsen 1972). These 2H 2 N — C — NH 2 + HC—COOH are the intermediates expected from the reductive Urea Glyoxylate route of pyrimidine degradation as established in some bacteria and fungi (Vogels and van der Drift FIG. 5. Biochemistry of purine breakdown in aerobic 1976). microorganisms (from Beevers 1976). Uptake of Nitrogenous Compounds Metabolism of Purines and Pyrimidines NECESSITY FOR UPTAKE MECHANISMS PURINES The concentrations of nitrogenous compounds The aerobic pathway of purine degradation in seawater are often low, i.e. 0-10 ,u/1/1. More appears to be much the saute in animais and micro- than 10 yr ago, Dugdale (1967), with natural popula- organisms (Vogels and van der Drift 1976). It is tions of marine phytoplankton, and Eppley et al. summarized in Fig. 5. Although it is possible that (1969b), with pure cultures, showed that when rate of an alga might gain nitrogen only from the side -NH, nitrate (or ammonium) uptake was plotted against of adenine or guanine, this restricted metabolism of concentration of nitrate (or ammonium) the result was purines is not known to occur. The best evidence a hyperbolic curve resembling the familiar Mi- for the occurrence of the reactions of Fig. 5 in algae chaelis—Menten velocity/substrate curve for an comes from the studies of Lynch and her co-workers. enzyme-catalyzed reaction. Moreover, they showed Chlorella cells took up uric acid, xanthine, hypo- that the concentration of nitrate (or ammonium), xanthine, or adenine (the latter two only after long generally called the Ks value, that supported half the lag periods). Uric acid and xanthine were initially maximum rate of nitrate (or ammonium) uptake was taken up faster than they were utilized so that they 0.1-10 ,u/l4 with the lower values being more accumulated in the cells (Ammann and Lynch 1964). characteristic of species from open oceans where When utilization commenced analysis showed the environmental concentrations are lowest. These half- presence of the expected intermediates. Thus hypo- saturation concentrations are considerably lower than xanthine was produced from adenine, xanthine the concentrations necessary to half-saturate the from hypoxanthine, and uric acid and allantoin from enzyme, nitrate reductase, where K„, is often about xanthine. The associated enzymes were not studied 150 p,M although it may be lower (about 50 ,uM) in in detail but uricase was shown to be very active. some marine phytoplankters (Packard 1979). Even the The enzyme, allantoinase, which converts allantoin K„, for glutamine synthetase (20-30 ,u/1/) looks

192

TABLE 4. Some examples of uptake systems for nitrogenous compounds in microalgae which result in substantial accumulation of substrate within the cells.

Inhibition by K, DNP, CCCP, Organism Compound (ILAI) or FCCP References

Ch'arena fusca Phenylalanine 5.0 + Pedersen and Knutsen (1974) Uracil 0.25 + Knutsen (1972) Guanine 1.0 + Pettersen and Knutsen (1974)

Chlorella pyrenoidosa Methylammonium 4.0 NT." Pelley and Bannister (1979)

Chlamydomonas reinhardii Urea 5.1 Williams and Hodson (1976)

Cyclotella cryptic(' Arginine 3.2 Liu and Hellebust (1974) Glutamic acid 36.0 Methylammonium 5.0 NT." Wheeler (1980)

Ditylum brightwellii NHT 1.1 NT." Eppley and Rogers (1970) NO.'2 4.0 NT." 1 NO( 0.6 NT." Of

Phaeodactylum tricornutum Urea 0.6 + Rees and Syrett (1979) NO,; 10.0 + Cresswell and Syrett (1981) Methylammonium 30.0 + Wright (unpublished)

"NT. = flot tested.

somewhat high against natural concentrations of NH; nium has been much used in the study of ammonium in seawater. A similar situation exists for urea meta- uptake mechanisms, e.g. in yeast (Dubois and Gren- bolism where McCarthy (1972) found Ks values for son 1979), in Chlorella (Pelley and Bannister 1979), uptake in several species of marine phytoplankton in Phaeodactylum (Wright and Syrett unpublished ranged from 0.2 to 0.8 p,M whereas the K,,, of urease is data), and in Cyclotella (Wheeler 1980). A parti- considerably higher, e.g. 460 p,M for the urease from cularly interesting finding is the great increase in the Phaeodactylum (Syrett and Leftley 1976). These rate of [" C] methylammonium transport which occurs findings clearly suggest the existence at the surface of when cells of Chlorella (Pelley and Bannister 1979), the cells of uptake mechanisms with a high affinity for Cyclotella (Wheeler 1980), or Phaeodactylum (Fig. 6) the substrate which will lead to the concentration of the are deprived of nitrogen for growth. substrate within the cell . Such mechanisms for the A radioactive analogue of nitrate that has been uptake of ammonium, nitrite, nitrate, urea, amino used in a limited way is [Cl] chlorate (Tromballa and acids , pyrimidines, and purines have now been studied Broda 1971; Shehata 1977). Chlorate competes with in algae. Some of them are listed in Table 4. nitrate for both the uptake mechanism and for nitrate reductase. In Chlorella with which it has been used, STUDY OF UPTAKE MECHANISMS much of the chlorate is reduced to chlorite but in The more detailed study of uptake mechanisms organisms known to accumulate nitrate, it might be an has been much aided by the use of radioactive com- effective tool for measurement of the uptake system. pounds. With nitrogenous compounds, radioactivity Several of the uptake systems listed in Table 4 can usually be introduced only when the compound share common features. First, the substrate is ac- is organic; thus ["C] urea and "C-labeled amino acids cumulated within the cells sometimes reaching a have proved useful. It is also helpful to use an analogue concentration 10 2 or 108 higher than the external of the compound which is not metabolized in the cells. concentration. Second, the uptake is inhibited by The uptake and accumulation of the analogue can then substances such as 2 ,4-dinitrophenol (DNP), carbonyl readily be measured, particularly if it is radioactive. cyanide in -chlorophenyl hydrazone (CCCP), or p- [HC]Thiourea has been used to study urea uptake trifluoromethoxy carbonyl cyanide phenylhydrazone (Syrett and Bekheet 1977) and [NC] methylammo- (FCCP). 193 3.0

2.5 • C14 C]Methylammonium uptake by Phaeodactylum

QI) 2.0 _Y UD

CL = 4— I) 1. 5 0

•, 0:3 • CC 1.0

0 r8 1 2 4 6 24 Hours of N deprivation FIG. 6. Increase in rate of uptake of r 4C1methylammonium by Phaeodactylum as a consequence of N deprivation. The culture (2 x 10" cells/mL) was suspended in N-free medium, illuminated, and aerated with air containing 0.5% CO 2 . Samples were removed at the times shown for measurement of [nC]methylammonium uptake (S. Wright unpublished data).

A MODEL SYSTEM — HEXOSE UPTAKE transport of protons (H+ ) (or a counter transport of The best indication of how such uptake systems OH- ions) as proposed by Mitchell (1967). In this may work in microalgae comes from Tanner and proposal the driving force for uptake is a proton Komor's detailed study of the mechanism of hexose- gradient across the cell boundary which is maintained sugar uptake by Chlorella . The uptake system is by metabolism; the theory can be represented dia- inducible, i.e. it develops during 1-h incubation grammatically as in Fig. 7 (Eddy 1978). On the left of with glucose or with nonmetabolizable sugars such Fig. 7 is an ATP-ase whose function is to pump out as 3-0-methylglucose or 6-deoxyglucose. Nonme- protons with a stoichiometry of ,n H+ per ATP hydro- tabolizable sugars may be taken up and concentrated lyzed and on the right a cotransport mechanism in in the cells by as much as 1600 x . The uptake can which the transport of one substrate molecule, S, is be driven by cyclic photophorylation (i.e. it takes linked to the cotransport of n protons. In the glucose place in illuminated anaerobic cells in which uptake system n = 1. evolution has been inhibited by DCMU) and it is inhibited by 50 ,I.LM FCCP. At pH 6.0-6.5 the K, outer cell membrane for uptake of glucose (and deoxyglucose) is 0.3 mM (Komor and Tanner 1971) but at higher pH values, the uptake system shifts to one of lower affinity, e.g. with a K, of about 30 mM at pH 8.4 (Komor and Tanner 1975). The simultaneous occurrence of dual uptake systems for the same substrate, one of high affinity and one of low affinity, has often been noted (Nissen 1974). When deoxyglucose is added to induced cells at pH 6.5 it is immediately taken up and at the same time the external medium becomes more alkaline; the transient pH rise lasts for about 60 s and the stoichiometry is such that one OH- ion is produced for each molecule of deoxy- glucose taken up (Komor 1973). These results are consistent with a mechanism of FIG. 7. Model for substrate (S) uptake by proton-linked uptake in which hexose transport is linked to a co- cotransport (Eddy 1978). See text for explanation. 194 If the transported complex is indeed S - H÷ as Chroomonas (Falkowski 1975). If such ATP-ases shown in Fig. 7, then the complex is charged and are important in uptake and are specifically stimulated the driving force for its movement can have two by substrates in cell-free preparations the linkage components: (i) the pH gradient across the membrane between the ATP-ase and the substrate uptake mecha- and (ii) the difference in charge across the membrane, nism must be much closer than as depicted in Fig. 7. i.e. the membrane potential. The role of ATP which is generated by metabolic reactions is to maintain H+ AND Na + COTRANSPORT these gradients across the membrane. Komor and Nevertheless, evidence is growing that the Tanner (1974, 1976) have estimated both the pH uptake of substrates other than sugars may be by a gradient and the membrane potential of Chlorella proton-linked cotransport (or OH --ion linked counter by measuring equilibrium distributions between cells transport). For example, thiourea uptake by Chlore/la and medium of radioactive dimethyloxazolidinedione is accompanied by a 1.1 uptake of protons (T. A. V. (distribution determined by difference in pH) or Rees unpublished data). radioactive tetraphenylphosphonium (distribution It has long been known that the uptake of nitrate determined by difference in potential). With an ex- and ammonium by algal cultures is accompanied ternal pH of 6.0, they find an internal pH of 7.1 by pH changes, the medium becoming alkaline when and a membrane potential of - 135 m V (i.e. cell nitrate is used and acidic when ammonium is taken interior is negatively charged with respect to external up. That such changes in H+ and OH- must inevitably medium). Moreover, the membrane potential is accompany the conversion of NO : 3 or NH; N into shown to rise (i.e. become less negative) when cellular nitrogen was well shown by the equations FCCP is added. It also rises when 6-deoxyglucose that Cramer and Myers (1948) were able to write is added and taken up as the theory depicted in Fig. 7 for Chlorella growth. It is, however, only recently requires. Komor and Tanner (1976) calculated the that the changes in pH have been linked mecha- free energy requirement to maintain a 1600-fold nistically to the uptake of the ions by algae (Raven accumulation of 6-deoxyglucose. They also calcu- and De Michelis 1979; Ullrich 1979). lated that, of this energy requirement, the measured In animal cells transport of substrate apparently difference in pH can contribute about one-third and takes place by cotransport with Na ions rather than the membrane potential about two-thirds, giving with protons (Eddy 1978). It is possible that the a sufficent total to meet the free energy requirement. same is true in microscopic marine plants. There are The agreement between the measured pH and now a number of reports of sodium-dependent trans- membrane potential gradients and those predicted port, e.g. of phosphate by the marine fungus Thrau- by the Mitchell theory as necessary to maintain the stochytrium roseum (Siegenthaler et al. 1967), of measured concentration gradient of 6-deoxyglucose glucose, amino acids, and methylammonium by is impressive. It may be partly fortuitous for, after Cyclotella (Hellebust 1978; Wheeler and Hellebust all , a Chlorella cell is a complex structure with 1981), and of urea and nitrate by Phaeodactylum organelles and small vacuoles: one does not know (Rees et al. 1980). It is tempting to speculate that what internal gradients of pH or of membrane po- in such plants which live in an environment where tential exist, nor does one know that the deoxyglucose the concentration of Na is high but that of H+ is is uniformly distributed within the cell; it seems low, substrate uptake may take place by systems unlikely that it is. Nevertheless the model depicted resembling that of Fig. 7 but with Na replacing H+. in Fig. 7 clearly has experimental support. One of its chief features is that the pump driven by ATP is not particularly closely linked to the uptake mecha- Regulation of Nitrate Reduction: nism for the substrate. Rather, the role of metabolism, Effects of Ammonium through ATP, is to maintain a differential between cell interior and the outside from which the uptake It is now well known that during growth in of the substrate follows. The system very much batch culture with ammonium nitrate as N source, resembles that studied by Jacques (1938-39) many the NH; is assimilated first, and only when it has years ago where distribution of ammonium between gone is NO 3 utilized (Ludwig 1938; Harvey 1953). seaweed cells and seawater was determined by dif- There are several reasons for this preferential assimi- ferences in pH and the NH 3 was thought to combine lation of ammonium. Active nitrate reductase is not with something in the cell surface which played a formed in the presence of NH; nor is the NO( uptake part in transport. system. And even if active nitrate reductase and an More work is needed to determine whether the NO (3 uptake system are present, the addition of NH; model of Fig. 7 is correct. It does not explain the can lead to a rapid cessation of NO,' utilization. occurrence of a NO : , Cl' stimulated ATP-ase in These effects of NH; have now been observed with membrane preparations from Skeletonema and a variety of micro algae and they have been subject

195 to considerable laboratory investigation in recent The regulation of the formation of nitrite re- years. ductase has received much less attention. Like NR, its formation in diatoms (Eppley and Rogers 1970; REGULATION OF NITRATE Llama et al. 1979), P/atyinonas (Ricketts and Edge REDUCTASE FORMATION 1978), Chlore/la (Losada et al. 1970), and Chia- mydomonas (Herrera et al. 1972) is repressed in the Active nitrate reductase (NR) is not formed presence of ammonium. while NH; is present as a major source of nitrogen. Active NR is formed when NH -grown cells are DISAPPEARANCE OF NR ACTIVITY transferred to nitrogen-free medium (Morris and Syrett 1965; Gesterheld 1971; Hipkin and Syrett There are at least three mechanisms by which 1977a). It has been argued that cells suspended in NR activity can disappear from cells. These include N-free medium produce small quantities of NO( two sorts of reversible inactivation phenomena and which then lead to induction of NR (Kessler and an irreversible loss of the enzyme due, presumably, Gesterheld 1970; Spiller et al. 1976) but others to degradation. dispute this (Syrett and Hipkin 1973). No NO,' for- Losada et al. (1970) described an inactivation mation could be detected in nitrogen-deprived cul- of NR after addition of NW to Chlore/la cells assimi- tures of a Chlamydomonas mutant lacking the ability lating NO. Inactivation was almost complete in to assimilate NO :' yet during nitrogen deprivation 90 min and was reversed following removal of NH; the cultures formed a modified form of NR (Hipkin and its replacement by N(I). In cell-free extracts, the et al. 1980). It seems most likely therefore that the inactivated NR still retains its cytochrome reductase major effect of nitrogen deprivation is to remove activity (see Fig. 3) and the NR activity can be NH; repression of the formation of NR. It is unlikely restored by incubation with the mild oxidant, po- that the repressor is NH; per se. In the fungus Neu- tassium ferricyanide. At about the same time Ven- rospora where a rather similar control operates, nesland and Jetschmann (1971) described an inactive there is good evidence that the repressor of NR for- nitrate reductase in another strain of Ch/ore//a. The mation is glutamine (Premakumar et al. 1979; Dunn- NR activity in cell-free extracts as normally prepared Coleman et al. 1979). Nevertheless although NR was only partial but full activity developed during is formed in nitrogen-deprived algae in the absence storage of the extract at 4°C in phosphate buffer. of NO!, , addition of NO,', generally leads to more Later studies showed that the cytochrome reductase rapid production of NR and a greater maximum activity was always fully developed and that full activity. This may be partly because the NO: 1 N is NR activity was restored by oxidation with ferri- assimilated and used for protein synthesis including cyanide (Pistorius et al. 1976). Much work has now synthesis of NR. This cannot be the whole story, been done by the Losada and Vennesland schools however, because with the Chlamydomonas mutant to elucidate the details of this reversible inactivation. lacking NAD(P)H—NR, which is unable to reduce Both agree that reduction of the enzyme, probably nitrate, although addition of nitrate has no effect on at the molybdenum site, is necessary for inactivation. the initial formation of the altered NR activity that The presence of NO rl protects the enzyme. The chief this mutant forms, it does have a large effect later, difference between their views is that the Losada possibly by preventing degradation of the enzyme group hold that reduction of the enzyme by NAD(P)H, (Hipkin et al. 1980). with the involvement of oxygen and ADP, is suf- The formation of NR is inhibited by cyclo- ficient to inactivate the enzyme in vitro (Losada and heximide, an inhibitor of protein synthesis, but not Guerrero 1979). The Vennesland group hold that a by 6-methyl purine which inhibits RNA synthesis small amount of cyanide (10' it,M) must also be (Hipkin and Syrett 1977b). This suggests that the present which combines with the enzyme (Lorimer formation of a messenger RNA is not necessary. et al. 1974; Pistorius et al. 1976). Moreover, it is now known that NH; -grown cells of The role of this reversible inactivation in vivo Chlorella contain substantial quantities of a protein is not easy to assess. There is no doubt that the which is immunologically very close to NR (Funk- reversible inactivation of NR following NH; addition houser and Ramadoss 1980). Thus formation of occurs but most workers find it less rapid and dramatic active NA after removal of NH; does not necessitate than as originally described by Losada et al. (1970) de novo synthesis of the whole NR molecule but and Herrera et al. (1972) (e.g. Rigano and Violante only of part of it. As the formation of active enzyme 1972; Pistorius et al. 1978). Rigano's work is is prevented by tungstate (Solomonson and Spehar with Cyanidium where the NR after extraction is 1977) it is probable that one part that has to be syn- partially inactive with activity being fully restored thesized contains Mo; it may be that a small molecular by heat treatment; Dunaliella also has a heat-activated weight Mo-containing peptide has to be formed and NR activity (Heimer 1975). The in vivo reversible added to the preformed major protein. inactivation seems greatest with high 02 and low

196 CO2 concentrations (Pistorius et al. 1976) and this no NH; inhibition of NO', uptake and, as with C- suggests a link with photorespiration. Solomonson deficient cells, NO is reduced to NH; which accu- and Spehar (1977) have developed a model for the mulates (Rigano et al. 1979a; Cullimore and Sims regulation of NR activity in vivo which suggests 1981). As MSO stops glutamine formation one might that intracellular cyanide plays an important part, suggest that glutamine is the inhibitor of NO uptake. it being formed by an interaction between glyoxylate However, with Anacystis, NH; does not inhibit NOt, (from photorespiration) and hydroxylamine (from uptake in the presence of either MSO or azaserine NO., reduction). which inhibits GOGAT and hence does not prevent The other type of reversible inactivation is best glutamine formation (Flores et al. 1980). Hence, if the seen in studies of Chlorella cultures synchronized inhibitor is organic, it is not glutamine, but something by light/dark alternation (Hodler et al. 1972; Tischner formed from it. Conway et al. (1976) suggested that 1976; Griffiths 1979). Activities of NR and nitrite inhibition was caused by the total internal amino acid reductase increase rapidly during the light period. pool. The activities decrease during the dark period al- Another possibility is that the inhibition results though they may begin to fall before the dark period not from an organic product of NH ; assimilation starts. Tischner and Hütterman (1978) showed that but from the production of protons as NH ; incor- the rapid increase of NR activity following illumi- poration may be represented as: nation is due to conversion of a preformed macro- molecule into active enzyme but the activation differs NH1 + organic C » organic N + H+ from the oxidation activation discussed above because (Raven and De Michaelis 1979) it is inhibited in vivo by cycloheximide and cannot, in cell-free extracts, be brought about by oxidation Proton production inside the cells could interfere with ferricyanide. Possibly part of the enzyme is with NO uptake if this does take place by H+- or synthesized during this process as appears to happen Ne-linked cotransport but it is difficult to see how when the enzyme is induced in NH; -grown cells such an inhibitory mechanism could be sufficiently (see above). selective. It is also unlikely because NO addition In addition, irreversible degradation of NR prob- has the same inhibitory effect on NO', assimilation ably occurs when, for example, cells are given (Thacker and Syrett 1972a). ammonium or darkened (e.g. Thacker and Syrett The effects of addition of NH; to cells assimi- 1972b) but it is not clear how such losses of activity lating NO', are thus complex. The first effect appears are related to the reversible type studied by Tischner to be an inhibition of NO', uptake but this is followed and Hütterman. by loss of NR (and nitrite reductase) activities. The loss of NR activity will be partly due to a reversible EFFECTS OF AMMONIUM ON inactivation, and partly due to irreversible loss of NITRATE UPTAKE enzyme with the rate of proteolytic breakdown of NO uptake by algal cells generally stops very NR possibly being greater in the presence of NH; quickly when NH; is added and recommences when (Hipkin et al. 1980). At the same time addition of the NH; has disappeared because of its assimilation NH; stops the synthesis of NR. (Syrett and Morris 1963; Eppley et al. 1969a; Conway et al. 1976). This inhibition by NH; is too rapid to be accounted for by the inactivation of nitrate re- SIMULTANEOUS ASSIMILATION OF ductase and there have been several suggestions AMMONIUM AND NITRATE that it is due to an effect on uptake. This effect is most readily demonstrated using diatoms where the Although NH; inhibition of NO", assimilation appearance of NO 4 in the cells can be measured by algae has often been demonstrated both with (Fig. 8). Serra et al. (1978b) demonstrated similar laboratory cultures and in field experiments (Mac- NH; inhibition of NO uptake using Skeletonema Isaac and Dugdale 1969, 1972) it is clear that algae grown with tungstate to produce inactive NR. that are highly nitrogen deficient or growing with a With highly carbon-deficient cells, NH; does limiting nitrogen supply assimilate both NH; and not inhibit uptake; indeed NO', is reduced to NO", simultaneously. This was shown nearly 50 yr NH; which accumulates (Syrett and Morris 1963; ago by Urhan (1932) using nitrogen-starved Chlorella Thacker and Syrett 1972a; Ullrich 1979). These and Scenedesmus. Simultaneous assimilation has findings suggest that it is not NH; that is the inhibitor been demonstrated using nitrogen-limited chemostat but an organic product of NH; assimilation. Fol- cultures, e.g. Bienfang (1975) with Dunaliella, lowing the finding that GS plays a major role in the Caperon and Ziemann (1976) with Pavlova (Mono- assimilation of NH;, it has now been shown that chrysis). In Bienfang's experiments the steady-state algae incubated with MSO (an inhibitor of GS) show NH; concentration never exceeded 0.5 ,w1/ but in 197 Inhibition of NO3 uptake by Phaeodactylum following addition of NH4 .

NHA . to • 300 —1

200

NO3 ' in • o medium

100 —I

\jle\a • O 1 1 1 1 1 1

3.0 —1

11-• NO3 - in • Co cc NH4 . to 2.0 A

cc

1.0 —I o

1 I 1

--I 200 NH4 . in 100 medium

0 1 O 60 120

MINUlES

FIG. 8. Effect of NH1 addition to a Phaeodactylum culture in a NO 4 medium on A (top), NC)!3 disappearance from the medium; B (middle), NO: 1appearance in the cells; C (bottom) shows NH1 disappearance from the medium; biCo uptake recommences after the NHI has gone (2 x 10 7 organisms/mL) (Cresswell and Syrett 1979). the experiments with Pavlova cells taken from a medium at the other. In between, some cultures nitrogen-limited chemostat, simultaneous uptake of grow with N sources such as NO: , urea, or amino NH; and NO: 1occurred with NH; concentrations acids which, while allowing close to maximal growth of 4 ,u,M. rates, nevertheless do not repress the formation of some enzymes as does NH;. In chemostat cultures NITROGEN STATUS OF ALGAL CULTURES various states of nitrogen nutrition can also be ar- Algal cultures can exist in various states of ranged by limiting the N input in the medium. It nitrogen nutrition. With batch cultures these range, may be, however, that nitrogen limitation by chem- at one extreme, from cultures growing with an ample ostat culture produces rather different metabolic states supply of NH; to cultures that are existing in a N-free than does N starvation of batch cultures.

198 The effects of nitrogen deprivation are sum- maximum rates of NO 4 or NW; assimilation and marized in Table 5. Some of the more obvious effects NR activity per cell. In contrast, GDH activity fell are an accumulation of carbon compounds such as markedly; GS activity was not measured. Not only polysaccharides and fats, a reduction in the rate of the activities of metabolic enzymes change but also photosynthesis, and, sometimes, a loss of chlorophyll. the activities of uptalce mechanisms. Thus in Phaeo- When supplied with NO!" and more especially dactylum even a short period of N deprivation leads NW, the N compound is assimilated more rapidly to a marked increase in the rates of uptalce of nitrate, than it is by normal-growing N-replete cells (se,e urea, and methylammonium and in other diatoms Table 6). Enzymic readjustments occur during N- and in Platymonas the activity of uptalce systems deprivation. For example, in N-starved batch culture for amino acids rises steeply (North and Stephens of Ankistrodesmus, there were increases in the ac- 1971; Liu and Hellebust 1974). tivities of NR, nitrite reductase, GDH, GS, and the Conway et al. (1976) used N-limited cultures allophanate hydrolase component of urea amidolyase. of Skeletonema to study the relationship between In contrast, ribulose biphosphate carboxylase activity rate of uptalce of NH:1- and NW- concentration. Rate fell as did the rate of ' 4CO2 incorporation but the of uptalce increased with increasing NHI concen- activity of glucose-6-phosphate dehydrogenase in- tration up to a concentration at which the rate quite creased (Hipkin and Syrett 1977a). In N-limited suddenly became maximal and higher NH_1 con- chemostat cultures similar adjustments are seen. For centrations produced no further increase. The maxi- example, with Thalassiosira (Eppley and Renger mum rate of uptake was dependent on the N content 1974) increased N limitation, by reduction of the of the cells, cells with the lowest N/cell giving the dilution rate, led to a fall in the rate of photosynthesis highest maximum rate. Conway et al. interpret these as measured by allowing samples to fix HCO., at results as showing that rate of N14:1- uptalce vs. Me saturating light intensity, but to little change in concentration follows Michaelis-Menten kinetics chlorophyll. Increasing N limitation increased the but only up to the point at which some internal factor

TABLE 5. Changes in algal cultures consequent upon nitrogen status.

Cultures grown Cultures grown Nitrogen-deficient on N}It on NO :i or urea cultures

accumulation of C reserves

decrease in rate of PS.

loss of RuBP carboxylase activity

loss of chlorophyll

increased proteolysis ea.

repression of some increase in activity of N-assimilatory enzymes N-assimilatory enzymes

repression of uptake systems increase in activity of .mg 1•••

for N compounds . uptake systems for N compounds

induction of gamete formation

199 ▪

imposes a limitation on rate of uptake. This they TABLE 6. Rates of assimilation of MX and N}11- by suggest may be the concentration of the intracellular Chlamydomonas (Thacker and Syrett 1972). (Rates are pool of amino acids (Sims et al. 1968). This may ihg atoms N (mg dry wt) - ' • h-' .) be so but it is clear that we are far from understanding the internal controls that regulate the metabolic readjustments that occur during N deprivation. NO; NW.; Interest now focuses on the glutamine synthetase N-de- N-de- pathway. Glutamine may play an important regulatory Normal prived Normal prived role or the regulator may be, as in the bacterium cells cells cells cells Klebsiella , the glutamine synthetase protein itself . (Tyler 1978). In darkness 0.013 0.14 0.000 0.58

Effects of Light OD Nitrate Assimilation: In light 0.117 0.38 0.127 0.81 Interaction with Carbon Metabolism Stimulation NECESSITY FOR A SUPPLY OF CARBON by light (cif') 800 170 CC 39 Figure 9 and Table 6 show some results from work with Chlamydomonas whiçh illustrate a number low concentration of 3- of points relating to NO and NHt assimilation and so does the addition of a their interrelationships with light and carbon nietabo- (3 ,4-dichloropheny1)-1 ' , 1 '-dimethylurea (DCMU). lism. Essentially similar results have been found Moreover, one notes that in light with CO., the rates of NHt and NO assimilation are about the same. with Dunaliella and other algae (Grant 1967, 1968; Turner 1969). Figure 9 shows that with One might interpret these results as demonstrating Grant and or NHt assimi- a "normal" culture of Chiamydomonas the assimila- an obligatory linlcage between NO; tion of both NO; and NFIt is dependent on photo- lation and light. But that this is not so is shown by synthesis: that is assimilation requires light + CO.; Table 6. In this experiment the Chlamydomonas removal of either of these prevents assimilation and cultures were allowed to photosynthesize overnight in a nitrogen-free medium; the cells then become nitrogen starved and accumulate polysaccharide. 15 Net ■ - Light Such cultures now assimilate either NO or .---..----8- à•— -à -4---à ----_,5--e,5 , • 1-.-11,—Eli:tr—ii--0■...._.„ in darkness. Nlit is assimilated some 4 times faster 12 •— — No than nitrate but even NO is assimilated faster than f ...... _ - CO Light to \ photosynthesizing cells. Never- • 2 it is by "normal" • CO2 to • "e 09 \ \ theless even higher rates of NO or NFlt assimi- \• lation are attained when these nitrogen-starved cells 2: .06 are illuminated with CO., (Table 6). One can conclude then that both NO and NM- assimilation can talce place in darkness provided that NITRATE • • CO, • ASSIMILATION sufficient carbon reserves are available; the extent to which illumination will stimulate assimilation will _ (o) • O depend on the metabolic state of the cells (Grant and „ 1111.1 - CO, Turner 1969). 0 09 _ Figure 9 shows that "normal" cells, when illu- • à - Light minated without CO., do nothing with NO; , that N...‘e."%. Li ghti to • •_ is, they do not reduce it to N11.1- as do cells which are 06 CO, ton carbon-starved as well as being deprived of CO.,. This finding points to the operation of a feed-back o 03 control in "normal" cells when unable to photo- AMMON ION Light ASSIMILATION • CO2 "Z\ • synthesize, which prevents reduction of NO; , pos- •—•—•—t The control is absent from o (bi ) sibly by stopping uptake. 1 1 I 1 carbon-starved cells and probably works through r J I I 2 4 6 8 10 levels of organic N compounds; it is removed when Hours ample carbon is available (see Table 8 for summary).

FIG. 9. Effect of darkness or removal of CO2 on uptake PHOTOGENERAT1ON OF REDUCTANT of (a) NO:;, (b) NW; by Chlamydomonas . All cultures except the — CO2 ones were aerated throughout with air The results of Fig. 9 illustrate one relationship containing 0.5% CO2 (from Thacker and Syrett 1972). between photosynthesis and nitrogen assimilation 200 but ever since Warburg and Negelein's (1920) de- preventing Nlit assimilation either by rigorous monstration of a light stimulation of NO reduction by ' exclusion of carbon (e.g. Syrett and Morris 1963 Chlorella closer connections between nitrate assimila- with Chlorella 8p.; Thacker and Syrett 1972 with tion and light have been sought. Chlamydomonas; Ullrich 1979 with Ankistrodesmus) It should be noted that NO!, assimilation, like or by treatment with MSO (e.g. Ramos et al. 1980a CO2 assimilation, is a reductive process. Losada with Anacystis). The result with Chlorella is worth (1980) has emphasized that reduction of a N atom noting because this 'strain of Chlorella contains an in NO to a N atoin in an amino acid requires 10 NADH-specific nitrate reductase so that presumably electrons. Reduction of one CO. molecule to (CH2 0) NADH must have been generated via the photo- requires 4. As the C:N ratio of phytoplankton chemical reactions in the chloroplasts. These experi- cells is about 5, the ratio of electrons required for ments with whole cells are successful, however, only CO. incorporation to those required for NO N because the conditions of treatment, i.e. deprivation incorporation by growing cells is about 5 x 4:10 of carbon or treatment with MSO, somehow remove or 2:1. This argument does not necessarily imply the normal inhibition of NO assimilation by NH.1"; that the reductant for NO :; assimilation is derived possibly this is because the treatments prevent the directly from photochemical reactions but it does formation of glutamine. They do not necessarily illustrate that considerable amounts of reductant prove that in the normal growing algal cell, reductant have to be generated for NO:; N assimilation. for NO reduction is generated by photochemical The physiological electron donors for NO reactions. Perhaps the best evidence that it may be reduction to N117,' are NADH (and NADPH in some is the demonstration that at saturating light intensity, algae) for eucaryotic NR, and reduced ferredoxin the rate of oxygen evolution by Chlorella was in- for the NR of blue-green algae and for all nitrite creased by addition of NO while CO2 assimilation reductases. Reduced ferredoxin or NADH is also re- was unchanged (van Niel et al. 1953). As the quired for the GOGAT reaction of NH t assimilation. carbon dioxide reduction system was already working Neither NO nor NO4 reduction requires ATP whereas at full capacity, the evolution of extra 02 when NO the assimilation of ammonium by the GS /GOGAT due to the assimilation of was added could not be pathway does requires ATP to drive the synthesis of CO2 produced during a dark reduction of NO!, . This glutamine. Ferredoxin is well known as a component rather clear-cut result of van Niel et al. (1953) of the photochemical reactions in the chloroplast appears not to have been confirmed by workers where it is reduced by electrons derived from Photo- with other algal species (Bongers 1958; Grant and system I. It is likely, therefore, that there will be a Turner 1969). Generally, , addition of NO,; depresses close connection between light and those reactions CO. fixation. This depression could result from a that require reduced ferredoxin to drive them. But direct competition between NO (and NO.; produced we also note (Table 6) that NO assimilation to from it) and CO. for photochemically generated organic N can occur in darkness; therefore there electrons or it could result from an increased rate of must be dark catabolic reactions that can lead to respiration coupled to NO!, reduction. The extent ferredoxin reduction just as there must be for the of the depression of CO2 fixation is, however, often nitrogen fixation system of blue-green algae. In the too great, particularly with nitrogen-deficient cells chloroplast, electrons from reduced ferredoxin can (e.g. Thomas et al. 1976), to make the second alter- be transferred and so reduce pyridine nucleotide. native tenable. But the chloroplast pyridine nucleotide that is so There is more evidence for a direct photo- reduced is NADPH. Although this can act as electron chemical reduction of nitrite in algae (Kessler 1957, donor for some nitrate reductases, all eucaryotic 1964). Nitrogen-deficient cells, with C reserves, algal nitrate reductases can use NADH as electron can, in low illumination or darkness, reduce NO donor and some only utilize this (Table 3). More- the photochemical to NO!, which accumulates because over, the major pathway for NADH reduction in cells reactions to reduce NO.', further are lacking (Carlucci is the TCA cycle reactions of the mitochondrion et al. 1970; Thomas et al. 1976). with the glycolytic triose phosphate dehydrogenase reaction (located in the cytoplasm) playing a sub- INTERRELATIONSHIP WITH MITOCHONDRIAL sidiary role. METABOLISM With cell-free preparations of chloroplasts of The complexities of the relationship between chlorophyll-containing particles a photochemically light and NO!, reduction are well illustrated by recent driven reduction of NO to NFIt can be demonstrated work with higher plant leaves. NO reduction to (Paneque et al. 1969; Candau et al. 1976). Studies NO is strictly light dependent and normally ceases with whole cells are physiologically more relevant as soon as light is removed. However, NO!, reduction and with these a quantitative photochemical con- continues in darkness under anaerobiosis or when version of NO!, to NFIt can be demonstrated by the mitochondria] electron transport chain is inhibited

201 with antimycin A (Canvin and Atkins 1974; Canvin tivation. Several of the enzymes of the Calvin cycle and Woo 1979). The anaerobic dark reduction of in the chloroplasts, e.g. ribulose biphosphate carbo- nitrate is inhibited by malonate and the inhibition xylase and fructosebiphosphatase, are inactive in is removed by addition of fumarate; these obser- darkness but become active following illumination. vations imply that it is the TCA cycle that is sup- Other enzymes of glycolysis (e.g. phosphofructo- plying electrons for NO!, reduction (Ramaroa et al. kinase) and of the pentose phosphate pathway (e.g. 1980) and are consistent with the view that, in leaves, glucose-6-phosphate dehydrogenase) are active in the NADH for nitrate reduction is generated in the darkness but inactivated by light and, in some algae, mitochondrion by TCA-cycle reactions. Under ribulose biphosphate carboxylase becomes inacti- aerobic conditions in darkness this NADH is not vated again at high light intensities (Codd and Stewart used for nitrate reduction but is reoxidized by molec- 1980). Part, at least, of the mechanism of light ular oxygen via the electron chain, i.e. it is used in activation/inactivation involves the reduction of normal aerobic respiration and AT.P is generated the enzyme by electrons derived from reduced fer- by oxidative phosphorylation. If the transport of redoxin and passed to the enzyme via a low molecular electrons down the mitochondrial electron transport weight protein, thiorecloxin (Buchanan 1980). Light chain to oxygen is prevented by (i) antimycin A may, however, activate enzymes more directly. or (ii) lack of oxygen or (iii) a high level of ATP One inactive form of Chlorella nitrate reductase can coming from photophosphorylation, then the NADH be reactivated in vitro in darkness by addition of is not reoxidized by oxygen but by NO!,. Thus the the oxidant, ferricyanide: it can also be activated strict light dependence of NO!, reduction under in vitro by illumination with blue light but not by aerobic conditions is explained although the reactions red light (Aparacio et al. 1976). Yet another type generating electrons for NO reduction are essentially of light activation of NR occurs in light/dark syn- dark respiratory ones (Sawhney et al. 1978). Whether chronized cultures of Chlorella; this has been dis- such a regulatory mechanism for NO reduction cussed above. In these cultures not only NR but occurs in algal cells is not known but it is clear that also glutamine synthetase and GOGAT show light interrelationships between the cellular organelles activation (Tischner 1980). may play a major part in metabolic control.

EFFECTS OF LIGHT QUALITY ON PHOTOGENERAT1ON OF ATP PROTEIN SYNTHESIS Light can have other effects on nitrate metab- Another effect of light on algal nitrogen metabo- olism. For example, CCCP and FCCP are known lism is that of light quality on the nature of the prod- to uncouple the formation of ATP from the electron ucts of photosynthesis (see Kowallik 1970; Raven transport reactions in the mitochondria and the chlo- 1974). In general, blue light favors the production roplasts and to decrease the charge difference across of amino acids and proteins rather than of carbo- the plasma membrane; 10 t.tA/ CCCP inhibits both hydrate with either N04, NH, or urea as N source. NO and NO reduction by illuminated Ankistrodes- Chlorella has been much studied but so have the mus (Ullrich 1974) and 0.2 1.4,M FCCP inhibits NO!, marine microalgae Cyclotella nana and Dunaliella reduction and NO reduction by Scenedesmus at tertiolecta (Wallen and Geen 1971); both these or- pH 6.5 (Andersion and Larrson 1980); at pH 5.0 ganisms grew faster in blue light than in white light NO reduction is much less inhibited by FCCP. The of equal energy content, and a greater proportion explanation here is that NO!, (and NO at pHs around of 14C fixed went into protein. The underlying neutrality) enters the cells by an active uptake mech- mechanism is not understood but the effect may be anism requiring ATP which can come from photo- confined to eucaryotes and Kowallik points to a phosphorylation. Uncouplers which prevent ATP suggestion of Pirson that the explanation may lie synthesis consequently stop NO'3 reduction. This in effects of light quality on intracellular transport mechanism comes close to the suggestion of Warburg of metabolites, across the chloroplast boundary, for and Negelein (1920) that light stimulates NO reduc- instance. tion by Chlorella by increasing the rate of permeation into the cells. SUMMARY Table 7 summarizes some of the ways in which LIGHT ACTIVATION/INACTIVATION . light can affect algal nitrogen assimilation and Table OF ENZYMES 8, the interaction between light and the metabolic Another effect of light just becoming under- state of the cells as these factors alter NO!, and NH:t stood is its effect on enzyme activation and deac- assimilation. 202

TABLE 7. Possible interactions of light with inorganic ni- Relevance of Laboratory Investigations trogen metabolism of algae. to Studies of Natural- Populations

PHOTOSYNTHETIC (CHLOROPLAST) EFFECTS It is never easy to extend to natural populations (i) Generation of reduced ferredoxin which then: the results of laboratory studies, carried out under (a) reduces I\11`24 (and N2 and NO!, in blue-greens). defined conditions with pure cultures of organisms, (b) reduces NAD(P)H and hence NO!, in eucaryotic often of little ecological significance. But there are algae. (c) drives GOGAT reaction of NH; assimilation. several situations where analogies can be drawn (d) activates/inactivates enzymes via thioredoxin. and to conclude attention is drawn briefly to three examples. (ii) Generation of ATP through photophosphorylation First, the well-known diel periodicity of NR which then: activity (and of NQ'i uptake) which has been observed (a) is used for to drive transport mechanisms for with diatoms (Eppley et al. 1971) and with other phy- N04, /‘/O, NH;. toplankters in culture and in natural populations (b) drives GS reaction of NH; assimilation. (Eppley et al. 1970) has a clear relationship to the light (c) stops the reoxidation of mitochondria] NADH activation of NR that has been so well studied in by 02 so making this NADH available for NO,; reduction. synchronous Chlorella cultures. (d) drives N2 fixation in blue-greens. Second, studies in the oceans of the assimi- lation of NH ;, urea, NO, and NO( which show (iii) Photosynthetic fixation of CO 2 makes C acceptors preferential assimilation of NHI followed by urea, available for NH; assimilation thus removing but simultaneous assimilation of all substrates when feed back inhibition by organic N compounds of concentrations are low (McCarthy et al. 1977) must NO (NO) uptake. be considered in relationship to laboratory studies which reveal a variety of effects of ammonium on OTHER EFFECTS nitrate metabolism. These studies also suggest that (i) (red light) effects? the extent of ammonium inhibition of nitrate assimi- (ii) Direct enzyme activation/inactivation by blue light lation will depend both on NH; concentration and possibly mediated through flavoproteins. on the metabolic state of the organisms. (iii) Effects of light quality on protein synthesis. Lastly, one can comment on the relationship of laboratory studies to the vertical distribution of nitrite (NO) in oceans. In late summer in the N. TABLE 8. Interaction of light and metabolic state in deter- Pacific, N4:1; concentration is relatively low down mining NH; or NO assimilation by microalgae such as to a depth of 100 m. It then rises sharply to a peak at Chlamydomonas and Dunaliella 130 m depth and then declines again; this peak coincides with a maximum for chlorophyll. In con- Metabolic state of cells trast, NO;; concentration remains low until 130 m, Carbon- Normal and then begins to rise markedly so that its concen- starved growth fast Nitrogen- tration continues to increase down to 200 m and +CO2 starved beyond (Kiefer et al. 1976). There are two major ways in which NO. may be formed. First, from Storage C compounds Nil Very low High NO, reduction by phytoplankton (Vaccaro and Ryther 1960); second, from bacterial oxidation Rate of assimilation of of NH1- (Wada and Hattori 1971). The relative im- { Light 0 ++++ portance of the two processes may well differ in NH; different circumstances. Nevertheless, laboratory Dark 0 O +++ studies of algal cultures utilizing NO:' show that among the conditions leading to NO .'2 accumulation Light ++ are (i) the presence of nitrogen-deficient organisms ammonium rich in C reserves and (ii) inhibition of photoreactions accumulates NO by either darkness or a chemical inhibitor (DCMU) Dark 0 0 (Carlucci et al. 1970; Thomas et al. 1976). Under . . nitrite these conditions NO(3 can be reduced to NO !2 by accumulates respiratory-linked reactions but as the further reduction of NO is more closely linked to light NH; inhibition of O +a NO 4 uptake reactions, NO.', accumulates. It seems probable, therefore, as Kiefer et al. (1976) have argued, that 'With dependence on NH; concentration. the peak of NO., at 130 m in the N. Pacific is due to

203 carbon-rich organisms sinking to a position where BAKER, J. E., AND J. F. THOMPSON. 1961. Assimilation illumination is low but NO(, concentration high. of ammonia by nitrogen-starved cells of Chlorella When one reflects on advances in knowledge vulgaris . Plant Physiol. 36: 208-212. of the nitrogen metabolism of microalgae over the BEE VERS, L. 1976. Nitrogen metabolism in plants. Edward last 25 yr, one is indeed struck by how many funda- Arnold, London. mental studies have been undertaken, successfully, BEKHEET, I. A., AND P. J. SYRETT. 1977. Urea-degrad- in attempts to explain ecological observations. In- ing enzymes in algae. Br. Phycol. J. 12: 137-143. deed, one suspects that microalgal studies may re- BELMONT, L., AND J. D. A. MILLER. 1965. The uti- present one of the best examples of close and fruitful lization of glutamine by algae. J. Exp. Bot. 16: 318- 324. cooperation between the closed environment of the BIENFANG, P. K. 1975. Steady state analysis of nitrate- laboratory and the open one outside. ammonium assimilation by phytoplankton. Limnol. Oceanogr. 20: 402-411. BIRDSEY, E. C., AND V. H. LYNCH. 1962. Utilization Acknowledgments of nitrogen compounds by unicellular algae. Science (Washington, D.C.) 137: 763-764. I wish to thank the following members of my research BOLLER, T., AND H. KENDE. 1979. Hydrolytic enzymes group for permission to quote their unpublished results: in the central vacuole of plant cells. Plant Physiol. Dr I. A. Bekheet, Ray Cresswell, Sue Everest, Nish Shah, 63: 1123-1132. and Sally Wright. I thank Drs Charles Hipkin and Alwyn BONGERS, L. H. J. 1958. Kinetic aspects of nitrate re- Rees both for results and for helpful criticisms of the duction. Neth. J. Agric. Sci. 6: 70-88. manuscript. BROWN, C. M., D. S. MACDONALD - BROWN, AND J. L. MEERS. 1974. Physiological aspects of microbial inorganic nitrogen metabolism. Adv. Microb. Physiol. References 11: 1-52. BUCHANAN, B. B. 1980. Role of light in the regulation of chloroplast enzymes. Annu. Rev. Plant Physiol. AHMED, S. I., R. A. KENNER, AND T. T. PACKARD. 31: 341-374. 1977. A comparative study of the glutamate dehy- AND JACKSON. 1977. A mechanism drogenase activity in several species of marine phyto- BUTZ, R. G., W. A. plankton. Mar. Biol. 39: 93-101. for nitrate transport and reduction. Phytochemistry 16: 409-417. ALLISON, R. K., H. E. SKIPPER, M. R. REID, W. A. SHORT, AND G. L. HOGAN. 1954. Studies on the CAIN, BROTHER JOSEPH. 1965. Nitrogen utilization in photosynthetic reaction. II. Sodium formate and urea 38 freshwater chlamydomonad algae. Can. J. Bot. feeding experiments with Nostoc muscorum. Plant 43: 1367-1378. Physiol. 29: 164-168. CAMPBELL, W. H. 1976. Separation of soybean leaf AMMANN, E. C. B., AND V. H. LYNCH. 1964. Purine nitrate reductases by affinity chromatography. Plant metabolism by unicellular algae. II. Adenine, hypo- Sci. Lett. 7: 239-247. xanthine, and xanthine degradation by Chlore/la CANDAU, P., C. MANZANO, AND M. LOSADA. 1976. pyrenoidosa Biochim. Biophys. Acta 87: 370-379. Bioconversion of light energy into chemical energy AMY, N. K., AND R. H. GARRETT. 1974. Purification through reduction with water of nitrate to ammonia. and characterization of the nitrate reductase from the Nature (London) 262: 715-717. diatom, Thalassiosira pseudonana. Plant Physiol. CANVIN, D. T., AND C. A. ATKINS. 1974. Nitrate, 54: 629-637. nitrite and ammonia assimilation by leaves: effects ANDERSON, M., AND C. - M. LARRSON. 1980. NO 7 and of light, carbon dioxide and oxygen. Planta 116: NO ,i dependent 03 evolution in Scenedesnuts: inter- 207-224. actions with CO2 reduction, p. 178-179. Confer- CAN VIN, D. T., AND K. C. Woo. 1979. The regulation ence proceedings. II. Congress of Federation of of nitrate reduction in spinach leaves. Can. J. Bot. European Societies of Plant Physiology, Santiago. 57: 1155-1160. ANTIA, N. J., B. R. BERLAND, D. J. BONIN, AND S. Y. CAPERON, J., AND D. A. ZIEMANN. 1976. Synergistic MAESTRINI. 1975. Comparative evolution of certain effects of nitrate and ammonium ion on the growth organic and inorganic sources of nitrogen for photo- and uptake kinetics of Monochrysis lutheri in con- trophic growth of marine microalgae. J. Mar. Biol. tinuous culture. Mar. Biol. 36: 73-84. Assoc. U.K. 55: 519-539. CARLUCCI, A. F., E. O. HARTWIG, AND P. M. BOWES. ANTIA, N. J., AND A. F. LANDYMORE. 1974. Physio- 1970. Biological production of nitrite in sea water. logical and ecological significance of the chemical Mar. Biol. 7: 161-166. instability of uric acid and related purines in sea CARPENTER, E. J., C. C. REMSEN, AND S. W. WATSON. water and marine algal culture medium. J. Fish. Res. 1972. Utilization of urea by some marine phyto- Board Can. 31: 1327-1335. plankters. Limnol. Oceanogr. 17: 265-269. APARACIO, P. J., J. M. ROLDÀN, AND F. CALERO. 1976. CODD, G. A., AND R. STEWART. 1980. Photoinactivation Blue light photoreactivation of nitrate reductase from of ribulose bisphosphate carboxylase from green algae green algae and higher plants. Biochem. Biophys. and cyanobacteria. FEMS Microbiol. Lett. 8: 237- Res. Commun. 70: 1071-1077. 240.

204 CONWAY, H. L., P. J. HARRISON, AND C. O. DAVIS. PALKOWSKI, P. G. 1975. Nitrate uptalce in marine phyto- 1976. Marine diatoms grown in chemostats under plankton: (nitrate-chloride)-activated ATP from silicate or ammonium limitation. II. Transient response Skeletonema costatum (Bacillariophyceae). J. Phycol. of Skeletonema costatum to a single addition of the 11: 323-326. limiting nutrient. Mar. Biol. 35: 187-199. FALKOWSKI, P. G., AND R. B. RIVKIN. 1976. The role CRAMER, M., AND J. MYERS. 1948. Nitrate reduction of glutamine synthetase in the incorporation of am- and assimilation in Chlorella pyrenoidosa . J. Gen. monium in Skeletonema costatum (Bacillariophyceae). Physiol. 32: 93-102. J. Phycol. 12: 448-450. CRESSWELL, R. C., AND P. J. SYRETT. 1979. Ammonium FLORES, E., M. G. GUERRERO, AND M. LOSADA. 1980. inhibition of nitrate uptake by the diatom, Phaeo- Short-term ammonium inhibition of nitrate utilization dactylum tricornutum . Plant Sci. Lett. 14: 321-325. by Anacystis nidulans and other cyanobacteria. Arch. 1981. Uptake of nitrate by the diatom, Phaeo- Microbiol. 128: 137-144. dactylum tricornutum . J. Exp. Bot. 32: 19-26. F000, G. E. 1978. Nitrogen fixation in the oceans, p. 11-19. In V. Granhall [ed.] Environmental role of CULLIMORE, J. V., AND A. P. Sims. 1981. An asso- ciation between photorespiration and protein catabo- nitrogen fixing blue-green algae and asymbiotic lism: studies with Chlamydomonas . Planta. (In press) bacteria. Ecol. Bull. 26. Swedish Natural Science Research Council. CZYGAN, F. C. 1965. Zur Frage der Zwischenprodukte FOGG, G. E., AND D. M. COLLYER. 1953. The accumu- der Nitratreduktion bei Griinalgen. Planta 64: 301-311. lation of lipides by algae, chap. 12, p. 177-182. In DIXON, N. E., C. GAZZOLA, R. L. BLAKELEY, AND J. S. Burlew [cd.] Algal culture: from laboratory to B. ZERNER. 1975. Jack bean urease (EC 3.5.1.5). pilot plant. Carnegie Institution of Washington A metalloenzyme. A simple biological role for nickel? Publication 600, Washington, DC. J. Am. Chem. Soc. 97: 4131-4133. FUNKHOUSER, E. A., AND C. S. RAMADOSS. 1980. DuBois, E., AND M. GRENSON. 1979. Methylamine/ Synthesis of nitrate reductase in Chlorella. II. Evidence ammonia uptake systems in Saccharomyces cerevisiae: for synthesis in ammonia-grown cells. Plant Physiol. multiplicity and regulation.. Mol. Gen. Genet. 175: 65: 944-948. 67-76. GALLON, J. R. 1980. Nitrogen fixation by photoautotrophs, DUGDALE, R. C. 1967. Nutrient limitation in the sea: p. 197-238. In W. D. P. Stewart and J. R. Gallon dynamics, identification, and significance. Limnol. [ed.] Nitrogen fixation. Annu. Proc. Phytochem. Oceanogr. 12: 685 - 695. Soc. Europe, Vol. 18. Academic Press, London and DUNN - COLEMAN, N. S., A. B. TOMSETT, AND R. H. New York. GARRETT. 1979. Nitrogen metabolite repression of GIRI, L., AND C. S. RAMADOSS. 1979. Physical studies nitrate reductase in Neurospora crassa: effect of the on assimilatory nitrate reductase from Chlorella gin-la locus. J. Bacteriol. 139: 697-700. vulgaris. J. Biol. Chem. 254: 11703-11712. EDDY, A. A. 1978. Proton-dependent solute transport GRANT, B. R. 1967. The action of light on nitrate and in microorganisms, p. 280-360. 1/1 F. Bronner and nitrite assimilation by the marine Chlorophyte, Duna- A. Kleinzeller [ed.] Current topics in membranes and liella . J. Gen. Microbiol. 48: 379-389. transport. Vol. 10. Academic Press, New York and 1968. Effect of carbon dioxide concentration and London. buffer system on nitrate and nitrite assimilation in EDGE, P. A., AND T. R. RICKETTS. 1978. Studies on Dunaliella tertiolecta. J. Gen. Microbiol. 54: 327- ammonium-assimilating enzymes of Platymonas 336. striata Butcher (Prasinophyceae). Planta 138: 123- 1970. Nitrite reductase in Dunaliella tertiolecta: 125. isolation and properties. Plant Cell Physiol. 11: 55- EPPLEY, R. W., J. L. COATSWORTH, AND L. SOLÔR- 64. ZANO. I969a. Studies of nitrate reductase in marine GRANT, B. R., C. A. ATKINS, AND D. T. CANVIN. phytoplankton. Limnol. Oceanogr. 14: 194-205. 1970. Intracellular location of nitrate reductase and EPPLEY, R. W., T. T. PACKARD, AND J. J. MACISAAC. nitrite reductase in spinach and sunflower leaves. 1970. Nitrate reductase in Peru current photoplankton. Planta 94: 60-72. Mar. Biol. 6: 195-199. GRANT, B. R., AND I. M. TURNER. 1969. Light stimulated EPPLEY, R. W., AND E. H. RENGER. 1974. Nitrogen nitrate and nitrite assimilation in several species of assimilation of an oceanic diatom in nitrogen-limited algae. Comp. Biochem. Physiol. 29: 995-1004. continuous culture. J. Phycol. 10: 15-23. GRIFFITHS, D. J. 1979. Factors affecting nitrate reductase EPPLEY, R. W., AND J. N. ROGERS. 1970. Inorganic activity in synchronous cultures of Chlorella. New nitrogen assimilation of Ditylum brightwellii . J. Phy- Phytol. 82: 427-438. col. 6: 344-351. GUILLARD, R. R. L. 1963. Organic sources of nitrogen EPPLEY, R. W., J. N. ROGERS, AND J. J. MCCARTHY. for marine centric diatoms, p. 93-104. In C. H. 1969b. Half saturation constants for uptake of nitrate Oppenheimer [ed.] Marine microbiology. C. C. and ammonia by marine phytoplankton. Limnol. Thomas, Springfield, IL. Oceanogr. 14: 912-920. HARVEY, H. W. 1953. Synthesis of organic nitrogen and EPPLEY, R. W., J. N. ROGERS, J. J. MCCARTHY, AND chlorophyll by Nitzschia closteriutn . J. Mar. Biol. Res. A. So URNIA. 1971. Light/dark periodicity in nitrogen Assoc. U.K. 31: 477-487. assimilation of the marine phytoplankters Skeletonetna HATTORI, A. 1958. Studies on the metabolism of urea costatum and Coccolithus huxleyi in N-limited chem- and other nitrogenous compounds in Chlorella ellip- ostat culture. J. Phycol. 7: 150-154. soidea. II. Changes in levels of amino acids and amides

205 during the assimilation of ammonia and urea by ni- KATES, J. R., AND R. F. JONES. 1964. Variation in trogen-starved cells. J. Biochem. (Tokyo) 45: 57-64. alanine dehydrogenase and glutamate dehydrogenase 1960. Studies on the metabolism of urea and other during the synchronous development of Chlamy- nitrogenous compounds in Chlorella ellipsoidea. III. donzonas Biochim. Biophys. Acta 86: 438-447. Assimilation of urea. Plant Cell Physiol. 1: 107-115. KESSLER, E. 1957. Untersuchungen zum Problem der HATTORI, A., AND I. UESUGI. 1968. Purification and photochemische Nitratreduktion in Grünalgen. Planta properties of nitrite reductase from the blue-green 49: 505-523. alga Anabaena cylindrica . Plant Cell Physiol. 9: 1964. Nitrate assimilation by plants. Annu. Rev. 689-699. Plant Physiol. 15: 57-72. HEIMER, Y. M. 1975. Nitrate reductase of Dunaliella KESSLER, E., AND H. OESTERHELD. 1970. Nitrification parva . Electron donor specificity and heat activation. and induction of nitrate reductase in nitrogen-deficient Arch. Microbiol. 103: 181-183. green algae. Nature (London) 228: 287-288. 1976. Specificity for nicotinamide adenine dinu- KESSLER, E. AND W. G. ZUMFT. 1973. Effect of nitrite cleotide and nicotinamide adenine dinucleotide phos- and nitrate on chlorophyll fluorescence in green algae. phate of nitrate reductase from the salt-tolerant alga Planta 111:41-46. Dunaliella parm . Plant Physiol. 58: 57-59. KIEFER, D. A., R. J. OLSON, AND O. HOLM - HANSEN. HELLEBUST, J. A. 1978. Uptake of organic substrates 1976. Another look at the nitrite and chlorophyll by Cyclotella cryptica (Bacillariophyceae): effects of maxima in the central North Pacific. Deep-Sea Res. ions, ionophores and metabolic and transport inhib- 23: 1199-1208. itors. J. Phycol. 14: 79-83. KITOH, S., AND S. HORI. 1977. Metabolism of urea in HELLEBUST, J. A., AND J. LEWIN. 1977. Heterotrophy, Chlorella ellipsoidea . Plant Cell Physiol. 18: 513- chap. 6. In D. Werner [ed.] The biology of diatoms. 519. Blackwell Scientific Publications, Oxford. KNUTSEN, G. 1972. Uptake of uracil by synchronous HERRERA, J., A. PANEQUE, J. M. MALDONADO, J. L. Chlore/la flaw. Physiol. Plant. 27: 300-309. BAREA, AND M. LOSADA. 1972. Regulation by KOMOR, E. 1973. Proton-coupled hexose transport in ammonia of nitrate reductase synthesis and activity Chlorella vulgaris FEBS Lett. 38: 16-18. in Chlanzydomonas reinhardii. Biochem. Biophys. KOMOR, E., AND W. TANNER. 1971. Characterization Res. Commun. 48: 996-1003. of the active hexose transport system of Chlorella HIPKIN, C. R., B. A. AL - BASSAM, AND P. J. SYRETT. vulgaris Biochim. Biophys. Acta 241: 170-179. 1980. The roles of nitrate and ammonium in the 1974. The hexose-proton cotransport system of regulation of the development of nitrate reductase in Chlorella . J. Gen. Physiol. 64: 568-581. Chlaznydomonas reinhardii . Planta 150: 13-18. 1975. Simulation of a high and low affinity sugar HIPKIN, C. R., AND P. J. SYRETT. 1977a. Some effects uptake system in Chlore/la by a pH-dependent change of nitrogen-starvation on nitrogen and carbohydrate in the K„, of the uptake system. Planta 123: 195-198. metabolism in Ankistrodesmus braunii . Planta 133: 1976. The determination of the membrane po- 209-214. tential of Chlore/la vulgaris. Evidence for electrogenic 1977b. Post-transcriptional control of nitrate sugar transport. Eur. J. Biochem. 70: 197-204. reductase forrnation in green algae. J. Exp. Bot. 28: KOWALLIK, W. 1970. Light effects on carbohydrate and 1270-1277. protein metabolism in algae, p. 165-186. In P. Han-

HIPKIN, C. R., P. J. SYRETT, AND B. A. AL - BASSAM. dal [ed.] Photobiology of microorganisms. Wiley- 1979. Some characteristics of nitrate reductase in Interscience, London and New York. unicellular algae, p. 309-312. In E. J. Hewitt and KRATZ, W. A., AND J. MYERS. 1955. Nutrition and C. V. Cutting [ed.] Nitrogen assimilation of plants. growth of several blue-green algae. Am. J. Bot. 42: Academic Press, London and New York. 282-287. Ho, C-H., T. IKAWA, AND K. NISIZAWA. 1976. Puri- LEA, P. J., AND B. J. MIFLIN. 1975. The occurrence of fication and properties of a nitrite reductase from glutamate synthase in algae. Biochem. Biophys. Res. Porphyra yezoensis Ueda. Plant Cell Physiol. 17: Commun. 64: 856-859. 417-430. LEFTLEY, J. W. 1980. Studies on urea metabolism in HODLER, M., J-J. MORGENTHALER, W. EICHENBERGER, some unicellular algae. Ph.D. thesis, University of AND E. C. GROB. 1972. The influence of light on Wales, Wales, the activity of nitrate reductase in synchronous cultures LEFTLEY, J. W., AND P. J. SYRETT. 1973. Urease and of Chlorella pyrenoidosa FEBS Lett. 28: 19-21. ATP: urea amidolyase activity in unicellular algae. HODSON, R. C., S. K. WILLIAMS, AND W. R. DAVIDSON. J. Gen. Microbiol. 77: 109-115. 1975. Metabolic control of urea catabolism in Chia- Lw, M. S., AND J. A. HELLEBUST. 1974. Uptake of mydomonas reinhardii and Chlorella vulgaris . J. amino acids by the marine centric diatom Cyclotella Bacteriol. 121: 1022-1035. cryptica . Can. J. Microbiol. 20: 1109-1118. JACQUES, A. G. 1938-39. The kinetics of penetration. LLAMA, M. J., J. M. MACARULLA, AND J. L. SERRA. XVI. The accumulation of ammonia in light and 1979. Characterization of the nitrate reductase activity darkness. J. Gen. Physiol. 22: 501-520. in the diatom Skeletonema costatum . Plant Sci. Lett. KAPP, R., S. E. STEVENS, AND J. L. Fox. 1975. A 14: 169-175. survey of available nitrogen sources for the growth LORIMER, G. H., H- S. GEWITZ, W. VeILKER, L. P. of the blue-green alga, Agmenellum quadmplicatum SOLOMONSON, AND B. VENNESLAND. 1974. The Arch. Microbiol. 104: 135-138. presence of bound cyanide in the naturally inactivated

206 form of nitrate reductase of Chlorella vulgaris . J. MoRTENsoN, L. E., AND R. N. F. THORNELEY. 1979. Biol. Chem. 249: 6074-6079. Structure and function of nitrogenase. Annu. Rev. LOSADA, M. 1980. The photosynthetic assimilation of Biochem. 48: 387-418. nitrate, p. 112-113. Conference proceedings. II. MURPHY, M. J., L. M. SIEGEL, S. R. TOVE, AND H. Congress of Federation of European Societies of Plant KAmiN. 1974. : a new prosthetic group Physiology, Santiago. participating in six-electron reduction reactions cata- LOSADA, M., AND M. G. GUERRERO. 1979. The photo- lysed by both sulfite and nitrite reductases. Proc. Natl. synthetic reduction of nitrate and its regulation, chap. Acad. Sci. U.S.A. 71: 612-616. 12, p. 366-408. In J. Barber [ed.] Photosynthesis NAYLOR, A. W. 1970. Phytogenetic aspects of nitrogen in relation to model systems. Elsevier/North-Holland metabolism in the algae. Ann. N.Y. Acad. Sci. 170: Biomedical Press. 511-523. LOSADA, M., A. PANEQUE, P. J. APARACIO, J. M. NEILsoN, A. H., AND M. DOUDOROFF, 1973. Ammonia VEGA, J. CÀRDENAS, AND J. HERRERA. 1970. assimilation in blue-green algae. Arch. Mikrobiol, Inactivation and repression by ammonium of the nitrate 89: 15-22. reducing system in Chlorella . Biochem. Biophys. Res. NEILSON, A. H., AND T. LARSSON. 1980. The utilization Commun. 38: 1009-1015. of organic nitrogen for growth of algae: physiological LUDWIG, C. A. 1938. The availability of different forms aspects. Physiol. Plant. 48: 542-553. of nitrogen to a green alga (Chlorella). Am. J. Bot. NicHoLs, G. L., S. A. M. SHEHATA, AND P. J. SYRETT. 25: 448-458. 1978. Nitrate reductase deficient mutants of Chlatny- LYNCH, V. H., AND G. G. GILLMOR. 1966. Utilization domonas reinhardii . Biochemical characteristics. J. of glutamine and glutamic acid by Chlorella pyrenoi- Gen, Microbiol. 108: 79-88. dosa Biochim. Biophys. Acta 115: 253-259. NISSEN, P. 1974. Uptake mechanisms: inorganic and MCCARTHY, J. J. 1972. The uptake of urea by natural organic. Annu. Rev. Plant Physiol. 25: 53-79. populations of marine phytoplankton. Limnol. Ocea- NoRTH, B. B., AND G. C. STEPHENS. 1971. Uptake nogr. 17: 738-748. and assimilation of amino acids by Platymonas . 2. MCCARTHY, J. J., W. R. TAYLOR, AND J. L. TAFT. Increased uptake in nitrogen-deficient cells. Biol. Bull. 1977. Nitrogenous nutrition of the plankton in the 140: 242-254. Chesapeake Bay. 1. Nutrient availability and phyto- NOTTON, B. A., AND E. J. HEWITT. 1979. Structure plankton preferences. Limnol. Oceanogr. 22: 996- and properties of higher plant nitrate reductase, 1011. especially Spinacea oleracea, p. 227-244. In E. J. MACISAAC, J. J., AND R. C. DUGDALE. 1969. The Hewitt and C. V. Cutting [ed.] Nitrogen assimilation kinetics of nitrate and ammonium uptake by natural of plants. Academic Press, London and New York. populations of marine phytoplankton. Deep-Sea Res. ÔESTERHELD, H. 1971. Das Verhalten von Nitratreductase, 16: 45-57. Nitritreductase, Hydrogenase und anderen Enzymen 1972. Interactions of light and inorganic nitrogen von Atzkistrodestnus braunii bei Stickstoffmangel. in controlling nitrogen uptake in the sea. Deep-Sea Arch. Mikrobiol, 79: 25-43. Res. 19: 209-232. PAASCHE, E. 1971. Effect of ammonia and nitrate on MANZANO, C., P. CANDAU, AND M. G. GUERRERO. growth, photosynthesis and ribulosediphosphate 1978. Affinity chromatography of Anacystis nidulans carboxylase content of Dunaliella tertiolecta. Physiol. ferredoxin-nitrate reductase and NADP reductase on Plant. 25: 294-299. reduced ferredoxin-Sepharose. Anal. Biochem. 90: PACKARD, T. T. 1979. Half-saturation constants for nitrate 408-412. reductase and nitrate translocation in marine phyto- MEEKS, J. C., C. P. WOLK, W. LOCKAU, N. SCHILLING, plankton. Deep-Sea Res. 26A: 321-326. P. W. SHAFFER, AND W-S. CHIEN. 1978. Pathways PANEQUE, A., P. J. APARICIO, J. CÀRDENAS, J. M. of assimilation of ('"N) N2 and "NH; by cyano- VEGA, AND M. LOSADA. 1969. Nitrate as a Hill bacteria with and without heterocysts. J. Bacteriol. reagent in a reconstituted chloroplast system. FEBS 134: 125-130. Lett. 3: 57-59. MIFLIN, B. J. 1974. The location of nitrite reductase and PANEQUE, A., J. M. VEGA, J. CÀRDENAS, J. HERRERA, other enzymes related to amino acid biosynthesis in P. J. APARICIO, AND M. LOSADA. 1972. 185W- the plastids of root and leaves. Plant Physiol. 54: labelled nitrate reductase from Chlorella. Plant Cell 550-555. Physiol. 13: 175-178. MIFLIN, B. J., AND P. J. LEA. 1976. The pathway of PARSONS, T. R., M. TAKAHASHI, AND B. HARGRAVE. nitrogen assimilation in plants. Phytochemistry 15: 1977. Biological oceanographic processes, 2nd ed. 873-885. Pergamon Press, Oxford, New York. MITCHELL, P. 1967. Active transport and ion accumu- PEDERSEN, A. G., AND G. KNUTSEN. 1974. Uptake of lation, chap. IV, p. 167-198. In M. Florkin and L-phenylalanine in synchronous Chlorella fusca; E. H. Stotz [ed.] Comprehensive biochemistry, Vol. characterization of the uptake system. Physiol. Plant, 22. Elsevier Publishing Company, Amsterdam, 32: 294-300. London, New York. PELLEY, J. L., AND T. T. BANNISTER. 1979. Methyl- MORRIS, I., AND P. J. SYRETT. 1965. The effect of amine uptalce in the green alga Chlorella pyrenoidosa. nitrogen starvation on the activity of nitrate reductase J. Phycol. 15: 110-112. and other enzymes in Chlorella. J. Gen. Microbiol. PETTERSEN, R., AND G. KNUTSEN. 1974. Uptake of 38: 21-28. guanine by synchronised Chlorella fusca: charac-

207 terization of the transport system in autospores. Arch. of Federation of European Societies for Plant Physi- Mikrobiol. 96: 233-246. ology, Santiago. PISTORIUS, E. K., E. A. FUNKHOUSER, AND H. Voss. RIGANO, C., V. D. M. RIGANO, V. VONA, AND A. 1978. Effect of ammonium and ferricyanide on nitrate FUGGI. 1979a. Glutamine synthetase activity, am- utilization by Chlore/la vulgaris. Planta 141: 279- monia assimilation and control of nitrate reduction in 282. the unicellular red alga Cyanidium caldariwn . Arch. PISTORIUS, E. K., H-S. GEWITZ, H. VOSS, AND B. Microbiol. 121: 117-120. VENNESLAND. 1976. Reversible inactivation of nitrate RIGANO, C., AND U. VIOLANTE. 1972. Effect of heat reductase in Chlorella vulgaris in vivo. Planta 128: treatment on the activity in vitro of nitrate reductase 73-80. from Cyanidium caldarium. Biochim. Biophys. Acta PLATT, T., AND D. V. SUBBA RAO. 1975. Primary 256: 524-532. production of marine microphytes, p. 249-280. In RIGANO, C., V. VONA, V. DI MARTINO RIGANO, AND J. P. Cooper [ed.] Photosynthesis and productivity A. FUGGI. 1979b. Nitrate reductase and glutamate in different environments. Cambridge University dehydrogenase of the red alga Porphyridi uni aeru- Press, Cambridge. ginewn . Plant Sci. Lett. 15: 203-209. POCKLINGTON, R. 1971. Free amino-acids dissolved in ROON, R. J., AND B. LEVENI3ERG. 1968. An adenosine North Atlantic ocean waters. Nature (London) 230: triphosphate-dependent avidin-sensitive enzymatic 374-375. cleavage of urea. J. Biol. Chem. 245: 4593-4595. PREMAKUMAR, R., G. J. SORGER, AND D. GOODEN. ROWELL, P., AND W. D. P. STEWART. 1976. Alanine 1979. Nitrogen metabolite repression of nitrate re- dehydrogenase of the N2 -fixing blue-green alga, ductase in Neurospora crassa. J. Bacteriol. 137: Anabaena cylindrica . Arch. Microbiol, 107: 115-124. 1119-1126. SAMEJIMA, H., AND J. MYERS. 1958. On the hetero- RAMAROA, C. S., SRINIVASAN, AND M. S. NAIK. 1980. trophic growth of Chlorella pyrenoidosa. J. Gen. Origin of reductant for in vivo reduction of nitrate Microbiol. 18: 107-117. and nitrite in rice and wheat leaves. New Phytol. SAWHNEY, S. K., M. S. NAIK, AND D. J. D. NICHOLAS. RAMOS, J. L., E. FLORES, AND M. G. GUERRERO. 1978. Regulation of nitrate reduction by light, ATP 1980a. Glutamine synthetase - glutamate synthase: and mitochondrial respiration in wheat leaves. Nature the pathway of ammonium assimilation in Anacystis (London) 272: 647-648. nidulans, p. 579-580. Conference proceedings. II. SERRA, J. L., M. J. LLAMA, AND E. CADENAS. 1978a. Congress of Federation of European Societies of Nitrate utilization by the diatom Skeletonema costal mn. Plant Physiology, Santiago. I. Kinetics of nitrate uptake. Plant Physiol. 62: 987- RAMOS, J. L., M. G. GUERRERO, AND M. LOSADA. 990. 1980b. Photosynthetic production of ammonia by blue- 19786. Nitrate utilization by the diatom Skele- green algae, p. 581-582. Conference proceedings. tonema costatum. II. Regulation of nitrate uptake. II. Congress of Federation of European Societies of Plant Physiol. 62: 991-994. Plant Physiology, Santiago. 1978e. Characterization of the nitrate reductase RAVEN, J. A. 1974. Carbon dioxide fixation, chap. 15, activity in the diatom Skeletonema costatum. Plant p. 434-435. In W. D. P. Stewart [ed.] Algal physi- Sci. Lett. 13: 41-48. ology and biochemistry. Blackwell Scientific Publi- SHATILOV, V. R., A. V. SoF'IN, T. M. ZABRODINA, cations, Oxford, London. A. A. MUTUSKIN, K. V. PSHENOVA, AND V. L. KRETOVICH. 1978. Ferredoxin-dependent glutamate RAVEN, J. A., AND M. I. DE MICHELIS. 1979. Acid- synthase from Chlorella. Biokhimiya 43: 1492-1495. • base regulation during nitrate assimilation in Hydro- SHEHATA, S. A. 1977. Mechanisms for the uptake and &molt africanum. Plant Cell Environ. 2: 245-257. metabolism of nitrate by Chlatnydomonas. Ph.D. REES, T. A. V., AND P. J. SYRETT. 1979. The uptake thesis, University of Wales. Phytol. of urea by the diatom Phaeodactylum. New SIEGENTHALER, P. A., M. M. BELSKY, AND S. GOLD- 82: 169-178. STEIN. 1967. Phosphate uptake in an obligately marine REES, T. A. V., R. C. CRESSWELL, AND P. J. SYRETT, fungus: a specific requirement for sodium. Science 1980. Sodium-dependent uptake of nitrate and urea (Washington, D.C.) 155: 93-94. by a marine diatom. Biochim. Biophys. Acta 596: Sims, A. P., B. F. FOLKES, AND A. H. HUSSEY. 1968. 141-144. Mechanisms involved in the regulation of nitrogen REISNER, G. S., R. K. GERING, AND J. F. THOMPSON, assimilation in microorganisins and plants, p. 91-114. 1960. The metabolism of nitrate and ammonia by In E. J. Hewitt and C. V. Cutting [ed.] Recent aspects Chlorella . Plant Physiol. 35: 48-52. of nitrogen metabolism in plants. Academic Press, RICKETTS, T. R., AND P. A. EDGE. 1978. Nitrate and London and New York. nitrite reductases in Platymonas striata, Butcher SOLOMONSON, L. P. 1979. Structure of Chlorella nitrate (Prasinophyceae). Br. Phycol. J. 13: 167-176. reductase, p. 199-205. In E. J. Hewitt and C. V. RIGANO, C. 1971. Studies on nitrate reductase from Cutting [ed.] Nitrogen assimilation of plants. Academic Cyanidium caldarium. Arch. Mikrobiol. 76: 265-276. Press, New York and London. RIGANO, C., V. DI MARTINO RIGANO, A. FUGGI, AND SoLomoNsoN, L. P., G. H. LORIMER, R. L. HALL, V. VONA. 1980. Control of the assimilatary nitrate R. BORCHERS, AND J. L. BAILEY. 1975. Reduced reduction in a unicellular alga and the possible effector, nicotinamide adenine dinucleotide-nitrate reductase p. 595-596. Conference proceedings. II. Congress of Chlorella vulgaris . Purification, prosthetic groups

208 and molecular properties. J. Biol. Chem. 250: 4120- 1980. The effect of light on nitrogen metabolism 4127. of Chlorella sorokiniana , p. 659-660. Conference SOLOMONSON, L. P., AND A. M. SPEHAR. 1977. Model proceedings. II. Congress of Federation of European for the regulation of nitrate assimilation. Nature Societies for Plant Physiology, Santiago. (London) 265: 373-375. TISCHNER, R., AND A. HüTTERMAN. 1978. Light- SPILLER, H., E. DIETSCH, AND E. KESSLER. 1976. mediated activation of nitrate reductase in synchronous Intracellular appearance of nitrite and nitrate in ni- Chlorella . Plant Physiol. 62: 284-286. trogen-starved cells of Ankistrodesmus braunii . Planta TROMBALLA, H. W., AND E. BRODA. 1971. Das Verhalten 129: 175-181. von Chlorella fusea gegenüber Perchlorat und Chlorat. STEWART, W. D. P., AND P. ROWELL. 1975. Effects Arch. Milcrobiol. 78: 214-223. of L-methionine-DL-sulphoximine on the assimilation TYLER, B. 1978. Regulation of the assimilation of nitrogen of newly fixed NH2 , acetylene reduction and hetero- compounds. Annu. Rev. Biochem. 47: 1127-1162. ULLR1CH, W. R. 1974. cyst production in Anabaena cylindrica. Biochem. Der nitrat-und nitritâbhangige Biophys. Res. Commun. 65: 846-856. photosynthetische 0 2-Entwicklung in N2 bei Ankistro- destnus braunii. Planta 116: 143-152. SYRETT, P. J. 1953a. The assimilation of ammonia by 1979. Die Nitritaufnahme bei Grünalgen und ihre nitrogen-starved cells of Chlorella vulgaris . I. The Regulation durch ütissere Faktoren. Ber. Dtsch. Bot. correlation of assimilation with respiration. Ann. Bot. Ges. 92: 273-284. (London) 17: 1-19. URBAN, 0. 1932. Beitrâge zur Kenntis der Stickstoff- 1953b. The assimilation of ammonia by nitrogen- assimilation von Chlorella und Scenedestnus Jahrb. starved cells of Chlorella vulgaris . II. The assimila- Wiss. Bot. 75: 1-44. tion of ammonia to other compounds. Ann. Bot. VACCARO, R. F. 1965. Inorganic nitrogen in sea-water, (London) 17: 20-36. chap. 9, p. 365-408. In J. P. Riley and G. Skirrow SYRETT, P. J., AND I. A. BEKHEET. 1977. The uptake [cd.] Chemical oceanography. Vol. 1. Academic of thiourea by Chlorella . New Phytol. 79: 291-297. Press, London and New York. SYRETT, P. J., AND C. R. HIPKIN. 1973. The appearance VACCARO, R. F., AND J. H. RYTHER. 1960. Marine of nitrate reductase activity* in nitrogen-starved cells phytoplankton and the distribution of nitrite in the of Ankistrodesmus braunii . Planta 111: 57-64. sea. J. Cons. Cons. Perm. Int. Explor. Mer 25: 260- SYRETT, P. J., AND J. W. LEFTLEY. 1976. Nitrate and 271. urea assimilation by algae, p. 221-234. In N. Sun- VAN BAALEN, C. 1962. Studies on marine blue-green derland [cd.] Perspectives in experimental biology. algae. Bot. Mar. 4: 129-139. Vol 2. Pergamon Press, Oxford and New York. VAN NIEL, C. B., M. B. ALLEN, AND B. E. WRIGHT. SYRETT, P. J., AND I. MORRIS. 1963. The inhibition 1953. On the photochemical reduction of nitrate by of nitrate assimilation by ammonium in Chlorella . algae. Biochim. Biophys. Acta 12: 67-74. Biochim. Biophys. Acta 67: 566-575. VEGA, J. M., J. HERRERA, P. J. APARICIO, A. PANEQUE, TEMPEST, D. W., J. L. MEERS, AND C. M. BROWN. AND M. LOSADA. 1971. Role of molybdenum in 1970. Synthesis of glutamate in Acrobacter aerogenes nitrate reduction by Chlorella . Plant Physiol. 48: by a hitherto unknown route. Biochem. J. 117: 405- 294-299. 407. VENNESLAND, B., AND C. JETSCHMANN. 1971. The THACKER, A., AND P. J. SYRETT. 1972a. The assimilation nitrate reductase of Chlorella pyrenoidosa Biochim. of nitrate and ammonium by Chlatnydotnonas reinhar- Biophys. Acta 227: 554-564. dii. New Phytol. 71: 423-433. VILLERET, S. 1955. Sur la présence des enzymes des 1972b. Disappearance of nitrate reductase activity uréides glyoxyliques chez les Algues d'eau douce. from Chlamydotnonas reinhardii. New Phytol. 71: C. R. Acad. Sci. Ser. D: Sci. Nat. 241: 90-92. 435-441. 1958. Recherches sur la présence des enzymes THOMAS, R. J., C. R. HIPKIN, AND P. J. SYRETT. 1976. des uréides glyoxyliques chez les Algues marines. The interaction of nitrogen assimilation with photo- C. R. Acad. Sci. Ser. D: Sci. Nat. 246: 1452-1454. synthesis in nitrogen deficient cells of Chlorella . VOGELS, G. D., AND C. VAN DER DRIFT. 1976. Degra- Planta 133: 9-13. dation of purines and pyrimidines by microorganisms. THOMAS, W. H. 1968. Nutrient requirements and utili- Bacteriol. Rev. 40: 403-468. zation; algae, p. 210-228. In P. L. Altman and D. S. WADA, E., AND A. HATTORI. 1971. Nitrite metabolism Dittmer [cd.] Metabolism. Biological Handbooks. in the euphotic zone of the Central Pacific ocean. Federation of America Societies for Experimental Limnol. Oceanogr. 16: 766-772. Biology, Bethesda, MD, USA. WALLEN, D. G., AND G. H. GEEN. 1971. Light quality THOMPSON, J. F., AND A.-M. MUENSTER. 1971. Sepa- in relation to growth, photosynthetic rates and carbon ration of the Chlorella ATP: urea amidolyase into metabolism in two species of marine plankton-algae. two components. Biochem. Biophys. Res. Commun. Mar. Biol. 10: 34-43. 43: 1049-1055. WARBURG, O., AND E. NEGELE1N, 1920. Über die 1974. ATP-dependent urease: characteristics Reduktion der Salpetersâure in grünen Zellen. Bio- of a control in Chlorella; the search for it in higher chem. Z. 110: 66-115. plants. Bull. R. Soc. N. Z. 12: 91-97. WHEELER, P. A. 1980. Use of methylammonium as an TISCHNER, R. 1976. Zur Induktion der Nitrat-und Nitrit- ammonium analogue in nitrogen transport and assimi- reduktase in vollsynchronen Chlorella Kulturen. lation studies with Cyclotella ctyptiea (Bacillario- Planta 132: 285-290. phyceae). J. Phycol. 16: 328-334.

209 WHEELER, P. A., AND J. A. HELLEBUST. 1981. Uptake in Saccharomyces cereyisiae . J. Biol. Chem. 247: and concentration of alkylamines by a marine diatom: 1349-1353. effects of H+ and K+ and implications for the transport 1973. Urea carboxylase from Saccharomyces and accumulation of weak bases. Plant Physiol. 67: cerevisiae . Evidence for a minimal two-step reaction 367-372. sequence. J. Biol. Chem. 248: 325-330. WHEELER, P. A., B. B. NORTH, AND G. C. STEPHENS. WILLIAMS, S. K., AND R. C. HODSON. 1977. Transport 1974. Amino acid uptake by marine phytoplankters. of urea at low concentrations in Chlamydomonas Limnol. Oceanogr. 19: 249-295. reinhardii. J. Bacteriol. 130: 266-273. WHITNEY, P. A., AND T. G. COOPER. 1972. Urea ZUMFT, W. G. 1972. Ferredoxin: nitrite oxidoreductase carboxylase and allophanate hydrolase. Two com- from Chlorella. Purification and properties. Biochim. ponents of adenosine triphosphate: urea amidolyase Biophys. Acta 276: 363-375.

210 product substrate complex reversibly bindswith port brightwellii (2) of and the netics. expected tofollowso-calledMichaelis-Mentenki- can besupported as anapproximationofthe In mine maximal rateof to (S) phosphate Harvey's (1963) dimensions ofN (1) is results from phytoplankton on the unit time,or which Dugdale (1967)and activity beganwith investigations inthe ton nutrientuptakehasbeen AND KINETIC MODELS studies thatutilize''N-labeledsubstratestodeter- uptake a the concentration kinetics with V„,„, In theMichaelis-Mentenmodel, The useofthe of hyperbolic function An rates PHYSIOLOGICAL shed The useof potential role nutrients across Assessment (P) S understanding the half-saturationconstant need tobeknown for of nutrientuptakes, uptake, and Assimilation + of new light,boththeoretical and freeenzyme. to demonstratethatnutrientuptake(y) simply reciprocal any given nitrate E uptake physiological ecology.Dugdaleused data for uptake in taken up this model K = V K and theory, AND (ES) the enzyme 1 Eppley rectangular hyperbola \ in necessary toachievehalf uptake studieswith of of last decade. ma of the Eppley BASES that onlytwoparameters, (V,„„ ES concentration of the plasmalemma Phaeodadylum tricornutum nearly coincidentpapers K.+S Nutrient Uptake nutrient uptakekinetics THEIR BIOCHEMICAL of that subsequentlyyields per unit in as simplifies the The and Coatsworth(1968), the u \ (E) substrate order tospecify kinetics carrier-mediated trans- ). observed relationship isusuallyexpressed and Coatsworth time. Kinetics subject to form particulate N Kinetics E +P Museum This (K of phytoplank- and concentration substrate Cambridge, s of ) an substrate. JAMES analysis equivalent empirical, period numerous might be enzyme- Harvard (Fig. 1) Ditylum the rate of used of Comparative Zoology, (S) per the K J. of of of in in s Nutrient Utilization MA 02138,USA

MCCARTHY University, describing phytoplanktonnutrientuptakerequires -2.8 reaction the unable togrow For result depletion sufficiently measured atdifferentstarting of case ofmultisubstrate tration FIG. grown Either kinetics i.e. reaction al. plot all othersubstrates 1969.) following primaryassumptions: linear example, The It needstobestated,however,that 2) 3) 1) 1. NO UPTAKE RATE (jj in

(S) M/h) NO

a saturabletransportor 3- culture The medium There isonly Only tnily do and kinetics thatcanbedescribedby Eq.(1). transformation of derivation of for : i not necessarilyimplyproduct short substrate at uptake feedback rates are Coscinodiscus lineatus Ditylwn brightwellii(D.bright.) followed on NO 3 to precludebothsignificantback - urea, CONCENTRATION(pM) rate initial of is mixedsufficientlytoprevent are observed over inhibition. a single (v) reactions, Eq. (1) saturation but the S/v held as a rates of uptake cell vs. function of constant. concentrations of uptake systemcould substrate for the S surface. (•), (x). the concentration kinetics (Fig.2). of reaction (y) a (From Eppley urea byNO7- and theWoolf time interval NO purposes formation. or, inthe saturable i• concen- 10 0

was 211 are S, of et 1600

1400

Ditylum brighlwelhl

0.012 I- 1200

0.010 I- 1000

800 0.008 I- cn

600

400

200

1 -0.41 2 4 6 8 10 12 14 16 S (p.g - at N.1- -I )

FIG. 2. Urea uptake rate (v) as a function of urea concentration (S) for Dity/um brig/me/Hi (°), and the Woolf plot linear transformation of S/v vs. S (0). (From McCarthy 1972.)

212 In an alternate formulation used to describe nutrient-limited, steady-state growth (Droop 1974; 0.8 Dugdale 1977), p and p„, are the rate of nutrient uptake and maximal rate of uptake per individual f- 0.6 unit of the population (i.e. per cell), respectively. o K, is the same parameter represented in Eq. (1). -c; — 04

(3) p = rill K, S 0.2

The chemostat is a steady-state, continuous cul- I I I I 1 ture system that permits the study of the relationship 10 20 30 40 50 between the growth rate and the nutritional state of Q (vitamin B1 2 .cell - l) microorganisms in an aqueous medium. In balanced growth (Eppley 1981) the rate of nutrient uptake is FIG. 3. Growth rate (g) as a function of cell quota (Q) comparable to the rate of synthesis of new organic of for Pavlova (Monochrysis) lutheri. (From material. Caperon's experimental work with con- Droop 1968.) tinuous cultures heralded the introduction of the chemostat in studies of phytoplankton nitrogenous but for silicic acid and nitrogen the maximum growth nutrition. The origin of the chemostat in microbiolo- rate attainable in chemostat culture (A) was consi- gical studies can be traced to Monod (1942) and derably less than the rt, of Eq. (4). As both Droop Novick and Szilard (1950), and its utility had pre- (1973) and Dugdale (1977) pointed out, the g of the viously been demonstrated for bacteria grown in media Droop expression is not necessarily equivalent to the prepared so that the supply of organic carbon was maximum growth rate attainable in culture. The g growthrate limiting (see Herbert et al. 1956, for parameter is by definition the growth rate at which example). Under steady-state conditions the chemical Q becomes infinite, and as such is never actually composition of both the medium and the organisms, in achieved in culture. As will be seen below, the addition to the physiological state of the population, degree of discrepancy between and 2 is not the can then be related to the growth rate. same for all elements. In many studies of phytoplankton in nutrient- The true maximum growth rate (A) can be limited , chemostat culture, the specific growth rate related to medium nutrient concentration by the of the population, g(T - ') has been shown to be Monod expression hyperbolically related to the cell quota, Q (M or concentration of the limiting nutrient per cell, S (5) g= K where KO is the Q below which there is no population S growth and g is the specific growth rate at which Q is infinite (Droop 1968). where K is the half-saturation constant for growth, and definitions for other terms are as previously used. Q —K g K, is not equal to the K, value in Eq. (1) and (3). (4) Frequent failure to demonstrate empirically the relationship between chemostat population growth In apparent contrast with the results reported rates and nutrient concentration predicted by Eq. above for the relationship between S and v, Caperon (5), or the Monod expression (see, for example, (1967, 1968) observed that for chemostat-grown, Caperon and Meyer 1972; Eppley and Renger 1974; nitrate-limited Isochrysis galbana the growth rate, and Bienfang 1975), led to increasing use of Eq. (4), hence the nutrient uptake rate, was a function of the or the cell quota model. It has been shown, however, concentration of nitrogen within the individual cell that the equations that formulate growth as functions rather than that of the growth media. Equation (4) has of cellular content (Eq. (4) )and media nutrient con- also provided an adequate fit to data from vitamin B 1 2 centration (Eq. (5)) are equally appropriate for de- (Droop 1968), phosphate (Droop 1974; Fuhs 1968; scribing steady-state growth kinetics (Goldman 1977; Goldman 1977; Rhee 1973), and iron (Davies 1970) Dugdale 1977). Burmaster (1979) provides the limited chemostat cultures (see, for example, Fig. 3). algebraic analysis that demonstrates this strict equiva- In general, though, it has been considerably less useful lence. in describing the kinetics of growth on silicic acid The difficulties often reported in efforts to (Paasche 1973; Harrison et al. 1976) and nitrogen obtain data that conform to Eq. (5) apparently stem (Harrison et al. 1976; Goldman and McCarthy 1978). both from earlier expectations that K „ would be In each of these studies, Q was shown to vary with g, equal to K, and from an inability to measure pre-

213 cisely nutrient concentrations that approach K. of nutrient concentration (Eq. (5)) is the assumption For example, Caperon (1968) interpreted his data that the portion of the g vs. S relationship that cannot for invariant residual nitrate concentrations in the be observed with conventional methods for measuring growth chamber of a nitrate-limited chemostat culture nutrients still follows the hyperbolic expression. of I. galbana over a six-fold range in population From considerations of typical elemental com- growth rates as evidence that growth rate is solely position and the different structural and functional regulated by the quantity of nitrogen in an internal roles of various elements, the total cellular Q terms cellular pool. Similarly, Bienfang (1975) observed for major elemental constituents such as carbon, near order of magnitude variability in media nitrate nitrogen, phosphorus, and, in the case of diatoms, and ammonium concentrations for replicate chemostat silicon would be expected to have variable applica- runs with Dunaliella tertiolecta at near maximal bility in population growth models. Over the range growth rates. As 4 is approached, the media con- of possible Q values it is evident that for different centration of the limiting nutrient can increase from elements there are different strategies partitioning undetectable to several gg-at •1_,-1 with a very small the elemental material. From several studies it is incremental increase in p, (Fig. 4), making it very apparent that Q is less variable for N than for P, and difficult to define the precise shape of the p, vs. S it may be invariant for C. Goldman and McCarthy relationship (Caperon and Ziemann 1976; Goldman (1978) proposed that the ratio of minimum to maxi- and McCarthy 1978; Burmaster and Chisholm 1979). mum Q, K0 Q„„ was useful in approximating the Implicit in the kinetic model for growth as a function degree to which the g, in Eq. (3) overestimates

3,0

2.0

0 1:3

1.0

I I I I I I I I I 1 1 I 1 1 II 5 10 15 20

S (du g - at NH:41' -N.1- )

FIG. 4. Steady-state growth rate GO as a function of NH:1* concentration for Thalassiosira pseudonana (3-H). (From Goldman and McCarthy 1978.)

214 or the maximum growth rate, just short of washout The extent to which NO2 is accumulated intracel- in a chemostat culture, lularly following a pulse of N0 is obviously related to the degree of coupling between the processes of Ko uptake and reduction. (6) 1 17 [ RELATIONSHIP BETWEEN UPTAKE AND where Q„, is determined from a population at wash- ASSIMILATION CAPACITIES out, or in batch culture during logarithmic growth. The process of nutrient uptake is usually meas- It was shown that for Dunaliella tertiolecta the 12 ured by the disappearance of substrate from the and Q„, values were identical in batch and chemostat medium or the accumulation of total labeled elemental cultures (Goldman and Peavey 1979). Data from material within the cell. It represents transport of numerous studies were compiled to demonstrate that the substrate across the plasmalemma, and does not the K0/Q, values for vitamin B 12-, iron-, or phospho- permit inference regarding metabolism of the sub- rus-limited cultures are approximately 0.01-0.03, strate. Our understanding of the mechanisms respon- whereas those silicon- or nitrogen-limited cultures sible for transport and metabolism of nitrogen by are 0.20. Expressed another way, cultures for which phytoplankton is , however, insufficiently complete the growth rate is limited by vitamin 1312, iron, or to permit convenient separation of the two processes phosphorus can vary their cellular content of these in most cases. The activities of several enzymes elements over an order of magnitude range that is involved in the metabolism of nitrogen have been greater than can be realized for the cell quota of identified in phytoplankton. Nitrate reductase, nitrite silicon- or nitrogen-limited cultures. Carbon-limited reductase, glutamate dehydrogenase, glutamine growth has been studied only in freshwater phyto- synthetase, and glutamate synthase appear to function plankton, and here data are few, but Q appears to be similarly to enzymes that have long been subjects invariant over a wide range of growth rates; hence the of biochemical investigations in fungi and higher K0 IQ„, for carbon is — unity and, correspondingly, plants (Syrett 1981). In contrast, however, we know g, in Eq. (3) for carbon-limited growth is infinite. much less regarding the process of transport across Some studies have attempted to relate specific the plasmalemma into the cytoplasm. Furthermore, subunits of the total cellular Q term to growth or with the exception of glutamine synthetase in Ske- uptalce potential characteristic of a given steady state, letonema costatwn (Falkowski and Rivkin 1976), but few generalizations can be made. It is known the enzymes thought to regulate nitrogen assimi- that internal ammonium and nitrate can reach meas- lation in phytoplankton have high K„, values relative urable levels (often up to a few percent of total to typical environmental concentrations of the sub- cellular nitrogen) (Eppley and Rogers 1970; Conover strate. In the cases of nitrate reductase, nitrite re- 1975; Bhovichitra and Swift 1977) but for steady- ductase, and glutamate dehydrogenase in marine state population there is no evidence that the con- diatoms, this difference approaches 2-3 orders of centration of either can be related to growth rate magnitude, and it has led to the suggestion that more usefully than the total Q. The model proposed internal pooling of substrate is necessary for the by Greeney et al. (1973) partitions the uptake and enzymes to function efficiently (Eppley and Rogers growth processes as functions of extracellular nitro- 1970). genous nutrient concentration and intracellular amino An examination of recent literature on the subject acid concentrations, respectively. An internal com- of nitrate transport and metabolism in both microalgae partmentalization approach was found to be practical and higher plants clearly points to the problem of for silicon (Davis et al. 1978). The utilization of distinguishing between transport and metabolic silicon within the cell and the rate of population growth processes for the purpose of discussing kinetics. were independently modeled as functions of internal Falkowski (1975) presented evidence for a mem- dissolved silicate pool size and silica content in the brane-bound transferase, a nitrate and chloride- frustule, respectively. activated ATPase, in Skeletonema costatum. He Clearly, subdivision of the Q term has the poten- reported a K„, of 0.9 ,u/14 NO:i, thus indicating that tial of bringing additional insight and precision to the this enzyme could provide an effective transport cell quota model. For example, DeManche (1980) system for nitrate at concentrations that occur natu- noted that for Skeletonetna costatum the amino acid rally in marine waters. There is evidence, however, pools were higher in well-nourished versus nitrogen- that both the transport of nitrate across the plasma- starved populations, but for Thalassiosira aestives lemma and its subsequent reduction to nitrite can be pool sizes for amino acids were not correlated with effected by a single enzyme associated with the nutritional history. His data show, however, the poten- plasmalemma in at least some higher plants and tial for rapid increase in pool size when a pulse of microalgae. The role of such a nitrate reductase in nutrient is added to a nitrogen-starved population. the regulation of nitrate uptake can be inferred from

215 the data of Rao and Rains (1976) for barley and it is sufficient at this point to accept that for significant from that of Nichols et al. (1978) for Chlamydomonas portions of a doubling period they may proceed at reinhardi. Moreover, in both Ana baena (Hattori dramatically different instantaneous rates. and Myers 1967) and Platymonas (Ricketts and Edge 1977) nitrate reductase activity was found to DETERMINATION OF KINETIC PARAMETERS be associated with the particulate fraction of broken Nutrient uptake as a function of substrate con- cells. Experiments with Neurospora crassa have, centration — The idea that the rate of nutrient uptalce however, led to the conclusion that this organism by phytoplanlcton can be expressed as the function has a nitrate transport system independent of the of substrate concentration as described by Eq. ( I) reduction of nitrate (Schloemer and Garrett 1974). (Dugdale 1967; Eppley and Coatsworth 1968) led Butz and Jackson (1977) proposed the existence to multiple efforts to assess the kinetic parameters of a transmembrane enzyme that both transports and of this functionality for both laboratory-grown cul- reduces NO;i . If both transport and reduction are tures and natural assemblages of phytoplankton. The functions of the same enzyme, then similar K„, hypothesis underlying much of this work was that values for the two processes might be expected. for a specific habitat the indigenous species or races This has been found to be the case in some vascular of phytoplankton would have been selected on the plants (Butz and Jackson 1977) and a marine bac- basis of their ability to compete for nutrients at terium (Brown et al. 1975). Likewise, similarity concentrations characteristic of the habitat. It was would be expected in the half-lives for nitrate transport thought that a certain pattern of uptake for a given and reduction activities, but there are few data to species could be taken as a measure of fitness for a either support or refute this hypothesis. For marine nutritional dimension of the habitat, and that this phytoplankton there is, however, a large difference could be a contributing factor in the determination between the K,, values for a nitrate uptake and the of the temporal succession of dominant species in K„, values for nitrate reductase. Numerous studies environments that undergo seasonal changes in nu- with both laboratory cultures and natural assemblages trient availability. Eppley et al. (1969) demonstrated (cf. Eppley et al. 1969; Carpenter and Guillard that the Ks values for NOç and NH; uptake were 1971; MacIsaac and Dugdale 1969) have reported lower for oceanic than for neritic clones, and that Ks values 2 pM , whereas the K„, values for nitrate there was a direct correlation between cell size and reductase in laboratory clones of marine phytoplank- the K,, value. MacIsaac and Dugdale (1969) also ton have ranged from 60 to 110 p,M (Eppley et al. offered evidence for the significance of the K,, value 1969; Eppley and Rogers 1970; Packard 1979), and as a measure of environmental fitness in their demon- those for several natural assemblages in the upwelled stration that average values for natural mixed popu- waters off Cape Blanc, Mauritania, ranged from lation assemblages differed between oligotrophic 40 to 210 ,uM (Packard 1979). Because the nitrate oceanic and eutrophic neritic waters. Carpenter and concentrations in near surface waters of the sea are Guillard (1971) added further support to this notion usually closer to the K,, values, it is reasonable to with their observation that the pattern in K., values expect that a transport mechanism with greater for clones of a single species isolated from oceanic substrate affinity than nitrate reductase is respon- and neritic waters reflects the nutrient regime repre- sible for the translocation of this ion across the sentative of the waters of origin for each clone. plasmalernma of marine phytoplankton. The hypothetical potential for competitive Clearly then, in the case of one well-studied advantage associated with paired K,, and V„, values enzyme system it is at present either difficult or proposed by Dugdale (1967) is shown in Fig. 5. impossible to speak distinctly of transport and meta- The separate lines can be viewed as representative bolic process for the purposes of understanding whole of a species or clone-specific response pattern to cell studies of nutrient uptake. If the rate of leakage a nutrient such as N(:) - . Uptake rate is in dimensions for both the substrate and its products can be either of mass of the element of nutritional significance, estimated or assumed to be negligible during steady- i.e. the nitrogen in NO:7 , talcen up per unit organism state growth, then rates of metabolism for essential mass of the same element per unit time. Uptake is elements must equal rates of uptake. conveniently expressed in these dimensions, which As will be seen below, transient uptake phe- reduce simply to reciprocal time, since it is then nomena may be an important adaptive strategy for equivalent to a growth constant (Dugdale and Goering phytoplankton in nature, and the steady-state con- 1967), as will be seen below. dition that is easily established in laboratory con- It became clear during these early studies of tinuous culture may for this reason be a poor analogue nutrient uptake kinetics that the V„, values obtained of nature. In spite of our incomplete understanding from short-term uptake studies following even a brief of the manner in which nutrient uptake and organism period of nutrient deprivation were elevated. When growth are coupled and regulated (Eppley 1981), expressed in dimensions of reciprocal time, they

216 Vmax = 2.25 K, =2.0

Vmax 1,5 ••■•■••• K s =0,5

j 1 —

171

— uJ

5 10

RELATIVE S (mL-3

FIG. 5. Hypothetical competitive interaction between species with different kinetic constants for nutrient uptake. were much greater than the maximum attainable The most common procedure used in estimating growth constant for the clone under study (Eppley K, values for nutrient uptake requires observing the et al. 1969). As a consequence, estimates of com- change in concentration over time for multiple con- petitive ability for different clones were derived centrations of substrate added to aliquots of a culture from combinations of K, values for nutrient uptake that has been deprived of the nutrient of interest. and maximum growth rates from batch culture. This The period of exposure to nutrient must be long represents a hybridization of Eq. (1) and (5) or, in enough to observe uptake, but at the same time brief other words, it assumes that the half-saturation con- enough to prevent a significant change in the initial stants for nutrient-limited uptake and growth are iden- nutrient concentration. A more sensitive variant of tical. This assumption seemed valid at the time, as this approach is to use isotopically labeled nutrient Eppley and Thomas (1968) had found K, and K and assay the amount of label incorporated by the to be indistinguishable for Chaetoceros gracilis. organisms. Nutrient uptake data collected with both

217 these methods have usually been based on a single incorporation. Analyses of the data reveal that in the or terminal analysis, and they have been assumed incubation flasks for concentrations <4 ttg-at N • 1, -1 to represent a constant rate of uptake over the period substantial depletion of the substrate occurred ( >50%) of incubation. If, however, the rate changes with in the 5-min period. This is a reasonably common time as a consequence of either substrate depletion problem, and if it is not taken into consideration, the or altered physiological response, then the kinetic resultant estimates of K, are erroneously high and the parameters computed from such data cannot be confidence intervals are excessively broad. determined accurately. An example of such a problem An alternative procedure for estimating kinetiè can be seen in a particular example of data for parameters is the disappearance study, in which the ammonium uptake by Thalassiosira pseudonana decrease in nutrient added at a high initial concen- (3H) (Fig. 6). The linear transformation Ma vs. S) tration is observed (Fig. 7). This approach has been of the hyperbola fits the data well from 4 to 16 gg-at used in both nitrogen and phosphorus uptake studies Nfli-N • I.» . Below this range, however, the trans- (cf. Caperon and Meyer 1972; Burmaster and formed values are elevated relative to the extra- Chisholm 1979). When this approach is used for polation from higher concentrations. It is necessary steady-state chemostat populations, inaccurate to remember that the expectation of linearity presumes approximations of K„ values would be expected if that uptake behaves according to saturation kinetics. during the experiment there were significant physio- A K, value derived from uptake data has meaning logical adjustment of the uptake process in response only if this presumption is valid, and numerous data to the elevated nutrient concentration. In other words, sets have been published for which the rectangular the question needs to be asked as to whether a nu- hyperbola provides an adequate approximation. The trient-limited chemostat population maintained at case discussed above for Fig. 6, however, demon- steady state would respond similarly both to the strates the difficulty that can be encountered in such exposure to a low concentration of nutrient and to an effort. This population was growing at 1.55 ' , the exposure to a low concentration immediately or —50% 4, and the uptake determinations were following exposure to a high concentration. Nu- made during 5-min exposure to ' 5 NHI . The quan- merous studies with both batch and continuous cul- tity of culture material used per uptake determination ture have shown that a v at any values of S is enhanced was the minimum required for the analysis of isotope in response to nutrient deprivation (Eppley et al.

10 2

•

•■•••• /0 / 0 -17 • I (r)

xe

I II III II III -1 0 +1 5 10 15

S(,u. g - a t N .L -1 )

FIG. 6. NI-1; uptake (v) as a function of NH1- concentration (S) for Thalassiosira pseudonana (3-H) maintained in steady-state continuous culture at p,= 1.55 d - ' (•), and the Woolf plot linear transformation of S/v vs. S (x). The hyperbolic form (broken line) is extrapolated at low values of S in the region for which the S/v values are higher than expected from the linear transformation. Values for uptake (v) at low values for S that would correspond to S/v values coincident with the linear extrapolative (0) are fit by the hyperbolic form. (From McCarthy 1981.)

218 Se = 0.40

7 Se =0.20 • • • 171 +0 • se =on

o -1:-S e =0.05 17, I ,d L).--cr- S e r0 T FIG. 7. Nutrient concentration (S) disappearance from the media with time (T) by phytoplankton uptake following a nutrient pulse. • 1969; Perry 1972; Rhee 1973; Eppley and Renger 1974; Caperon and Ziemann 1976; Conway and 0.5 , 1.0 Harrison 1977; McCarthy and Goldman 1979) and V`ig -at NH4-N• L '1 added + estimated evidence for this has also been reported for natural assemblages (Glibert and Goldman 1981). For phos- FIG. 8. Calculated effect of erroneously high estimates of phate-limited Pavlova (Monochrysis) lutheri, Bur- ambient nutrient concentration (5„) on rates of nutrient master and Chisholm (1979) did demonstrate linearity uptake (v) determined by isotope techniques. (From Eppley et in uptake rates as a function of time, and they found al. 1977.) no statistically significant differences in K. values high variance in the K, values for constituent popu- determined by the multiple concentration and disap- lations of a natural assemblage (Williams 1973) or pearance methods. Clearly, they were dealing with multiple uptake processes for the same nutrient. populations that maintained constant uptake capacity In reviewing the literature on the kinetics of over the course of the experiment. The controls in phytoplankton nutrient uptake, one frequently finds this study were usually rigorous. it difficult or impossible to ascertain whether or not Another problem that some investigators have appropriate controls were conducted. Unless time encountered in studies of nutrient kinetics for natural course data are available to demonstrate linearity in populations is that of overestimating the available the uptake versus substrate response over the duration nutrient. This can result from either contamination of the experiment, uptake data must remain suspect, or lack of specificity in the nutrient analysis, and it and their utility questionable. has been shown to be a potentially serious problem Nutrient uptake cts a function of cellular com- for both the uptake of phosphorus in freshwater position — The rate of nutrient uptake is a function (Rigler 1966; Brown et al. 1978) and nitrogen both of substrate concentration (Eq. (1) and (3)), and the in freshwater (Liao and Lean 1978) and in the sea rate of population growth is a function of both cellular (Eppley et al. 1977). The effect is to overestimate composition (Eq. (4)) and substrate concentration rates of uptake at the lowest concentrations (Fig. (Eq. (5)). During balanced growth = it, but if the 8). The artifact becomes obvious in data sets that population is nutrient limited, then a potential for V„, include uptake rates for several concentrations, and can materialize that is in excess of the rate of nutrient conversely it can go undetected if uptake rate deter- uptake necessary to provide the needs of the popula- minations are based on a single nutrient concentration. tion. The validity of the hyperbolic fit to nutrient Dugdale (1977), in his analysis of the enhanced kinetic data can be conveniently tested by the use maximum uptake property, suggested that V„', be of linear transformations (Dowd and Riggs 1964). used to denote the maximum "nutrient specific uptalce For systems known to be described by Eq. (1), rate" characteristic of a steady-state population and departures from linearity in transformed data due hypothesized that either to substrate depletion or overestimation of ambient nutrient will be readily apparent. Subtle P„, (7) V„', = — departures can also arise for other reasons, such as Q 219 are upper and lower limits for the range of Q and the lower limit of V„', is that observed when the popu- lation has maximal Q or is at f A. Data for nitrogen uptake by Thalassiosira 7 I- pseudonana oceanic clone 13-1 (Eppley and Renger 1974) and for estuarine clone 3-H (McCarthy and . E Goldman 1979), in addition to phosphorus uptake by Scenedesmus sp. (Rhee 1973) (Fig. I0a, b, c), in- dicate, however, that p„, is not constant and, hence, the term p„; is proposed to denote the maximum rate of uptake per cell for a population in steady state at a specific growth rate. One exception to this Q (M•ce11 -1 ) pattern is the data of Burmaster and Chisholm (1979) Kq Qm for Pavlova (Monochrysis) lutheri (Fig. 10d). In the FIG. 9. Enhanced potential for maximum specific rate of results of Brown and Harris (1978) for Selenastrum , nutrient uptake (V?;) as a function of cell quota (Q) as- p can also be seen to decrease with increasing p. suming that the maximum rate of uptake per cell (p„,) and increasing Q. is invariant with Q. Ke? and Q„, represent minimum and Over roughly similar relative growth rates, as maximum values of Q for steady-state populations main- relationship between and % is similar tained in continuous culture. % A, the p„', A for three of the four studies (Fig. 10, Table 1). In where p„, is the maximum rate of cellular uptake the exception, the Burmaster and Chisholm (1979) ( V/ • cell- ' • T- '). In steady-state populations V„', should study, both the 33 F) uptake and phosphate disap- then be a function of the growth rate and, furthermore, pearance experiments yielded similar data, and the Dugdale proposed that if p„, is constant, then a plot authors suggest that the combined data appear to show of Q vs. V„; should be hyperbolic (Fig. 9). There an apparent maximum at intermediate growth rates.

Scenedesmus sp. a Thalassiosira pseudonana (13-1) \ ( Eppley and Renger 1974 ) (Rhee 1973 ) 0 7 5 5 Q) 71) ------o Ci-

(.0 S.■■•• 0 0

■ E 1■• sE 1 1. cL_ I %A 100 %A 100

) Tholassiosira pseudonana ( 3H) Ld Pavlova lu theri -1) -1 (McCarthy and Goldman 1979) in in ( Burmaster and Chisholm

-m 5 5 1979 ) -1.m -1 11 11 NO • . _----- e- ___0 • l P-ce • a) l N.ce ° o • __- 0 mo 6 mo Ce3.5 0

11■• -I -16 0 (1 10

( Pn Pin

0/04 too %A. 100

FIG. 10. Variability in the maximum rate of nutrient uptake per cell (hence p„', rather than p„,) as a function of relative growth rate (% p.). (a) Thalassiosira pseudonana (13-1) (from Eppley and Renger 1974); (b) Scenedesmus sp. (from Rhee 1973); (c) Thalassiosira pseudonana (3-H) (from McCarthy and Goldman 1979); (d) Pavlova (Monochtysis) Imbed (from Burmaster and Chisholm 1979).

220 TABLE I.

Growth rate region max Range Nutrient (% 4) min p„', in Q Organism Study

N07 or NH71' 19-94 1.5 4 x T. pseudonana (13-1) Eppley and Renger 1974

20-95 1.6 3.8 x T. pseudonana (3-H) Goldman and McCarthy 1978 McCarthy and Goldman 1979

PO4 26-68 1.6 4 x Scenedesmus sp. Rhee 1973

PO, 20-85 4.3 x Pavlova lutheri Burmaster and Chisholm 1979

It might be expected that the relationships between p„', and Q would differ for nitrogen- and phosphorus-limited populations. Goldman and McCarthy (1978) proposed that the ratio of maxi- mum to minimum cell quota (Q /K) over the entire range of growth rates is >30 for phosphorus in phosphate-limited populations and -5 for nitrogen in ammonium-limited populations. It is, however, 1- not just the range, but the shape of the vs. Q relationship that is of importance. The studies cited ,E above plus several others support the generalization that -50% of Q„, for nitrogen is attained at 50% a, under nitrogen-limited growth, whereas only 10-20% of the Q„, for phosphorus is attained at 50% 4 under phosphorus-limited growth. Consequently, the slope of the g vs. Q relation- ia (T-1 ) ship at low growth rates is greater for phosphorus A than for nitrogen. The p„, values for phosphorus- À•Z limited populations may remain uniformly high over FIG. 11. Enhanced potential for maximum specific rate a range of growth rates that extends from low to of nutrient uptake (V„;) as a linear function of growth rate (g) when p„, is assumed constant, and as a curvilinear near maximal values. For the highest growth rate function of bt when p„', varies with .t as seen in Fig. 10. included in the Burmaster and Chisholm (1979) study (-85% g), the average p„, (4.3 x 10- mol values of Q and ,u, is to increase V„', at low growth P • cell- l• min- ' ) was still near the maximum ob- rates relative to that expected from Eq. (8) (Fig. 11). served. However, Q for this population (4 x 10 - '5 The concept of a variable p„', needs to be studied cell - ' ) was less than 25% mol • Q. further to express it adequately as a function of g Of equal ecological significance is the relation- for both nitrogen and phosphorus. Only then can ship between V„', and p, (Fig. 11). If p„, were constant we present an accurate mathematical representation then the expression derived by Dugdale (1977) of V„', as a function of The variable nature of the maximum rate of 1 nutrient uptake (V„',) permits phytoplankton popu- lations to grow at near maximal rates at virtually (8) = pm Q undetectable levels of nutrient in the medium. This concept, which was discussed by Caperon and (substituting iiZ for g„„ the maximum ,u in the Droop Ziemann (1976), allows for K << Ks and it can expression, in order to keep terms equivalent, and be easily demonstrated. Figure 6 represents a typical moving a misplaced parenthesis), would satisfactorily data set for NIit uptake by a marine phytoplankter. relate V„', and g. (It is important to note that the As described above, because of the enhanced po- unattainable p, in the Droop expression rather than tential for uptake, accurate data at low values of 4 , the true maximum growth rate, must be used in S are particularly difficult to obtain. A smooth curve order for V„', to remain finite at the maximum growth drawn through the measured values for uptake (solid rate.) The effect of an enhancement in p„', at low circles) would give the impression that a concen- 221 The family of curves representing the hypo- thetical y vs. S responses for three populations at different steady-state growth rates (Fig. 12) illustrates the control of u by both S and Q. When studies of uptake kinetics with nutrient-limited populations involve alteration of the steady-state nutrient con- centrations, the population will shift from the bal- anced to unbalanced growth. Such a shift may, for practical purposes, be instantaneous, and an inter- pretation of the uptake data must take this into con- sideration. s (mL-3 ) Interactions between Uptake Processes FIG. 12. Proposed relationship between rate of nutrient for Multiple Nutrients uptake (u) and nutrient concentration (S) for steady-state population maintained at three different growth rates (s,, 2 , and 14). As seen in Fig. 11, the population with the g NUTRIENTS THAT SUPPLY THE SAME lowest p. ( .1,,) has the highest potential rate of nutrient CELLULAR ELEMENTAL REQUIREMENT uptake. The half-saturation constant for nutrient uptake (K,,) is assumed to be invariant with it. For elemental needs that can be met with nutrient material in a variety of different forms, there is good reason to investigate the kinetics of interaction for tration of approximately 0.5 ,u,g-at NHT-N • L-1 the uptake and/or assimilation of the various sub- in order to maintain a V equivalent to is necessary strates. In the case of silicon nutrition in diatoms, the measured steady-state growth rate of 1.55 d, the source is believed to be exclusively limited to whereas the observed growth chamber concentration monomeric orthosilicic acid (Si(OH) 1 ). The phos- was lower than expected by at least an order of phorus nutrition of phytoplankton can be met uni- magnitude. If, however, we agree to assume that versally by orthophosphate (P01 - ), and in some the V vs. S relationship is hyperbolic, and then esti- cases by monophosphoesters (MPE). The kinetics mate V values for low S values by extrapolation of interaction between POI - and MPE uptake and from the linear transformation, one can see the pos- assimilation have been the subject of a few inves- sibility of attaining a V of —1.55 d at a concen- tigations with phytoplankton (Perry 1972; Taft et al. tration <0.03 ,ug-at N • I.». The discrepancy that 1977). The general pattern that has been observed appears to exist between data sets like those repre- is that the induction of alkaline phosphatase activity sented in Fig. 4 and 6 can be reconciled in such a associated with the cell surface is correlated with manner if it is remembered that the former is a compo- the depletion of inorganic phosphate in the growth site for multiple steady states and the latter is repre- medium. The phosphatases facilitate the utilization sentative of a single steady state. A single species of organically bound phosphate by a hydrolysis that maintained at different growth rates will be capable releases PO t in the immediate vicinity of the cell of uptake response patterns that are unique for each surface. For nitrogenous nutrition, the nearly uni- growth rate (Fig. 12). Thus far the evidence seems versal suitability of at least three forms of nitrogen for to indicate that, whereas u for a given substrate phytoplankton growth (NO, N07, and NH1 ), in

concentration is a function of p/jî , the IC, for uptake addition to the species or clone-specific potentiality remains constant (Eppley and Renger 1974). It to utilize urea and some amino acids, greatly increases needs to be emphasized, though, that there are few the possibility of competitive interaction between reliable uptake data for very low growth rates. the processes responsible for transport and assimi- Although an uncoupling between uptake and lation of different forms of this element. growth has been well documented for both phos- The importance of 1\11Cq as a source of nitro- phorus- and nitrogen-limited growth, it has not been genous nutrient in upwelling regions, coastal regions seen for silicon. In several studies, the maximum in temperature latitudes, and perhaps deep in the rates of Si(OH) 1 uptake (V„,) are similar to the euphotic zone of oceanic waters is evident from many specific growth rates ( a) (Paasche 1973; Conway studies. The suppression of NO i uptake by NH; et al. 1976; Nelson et al. 1976) regardless of the when both are present in the growth medium has state of nutrient deprivation. This generalization is been the subject of many laboratory and field studies. utilized in the Davis et al. (1978) model that describes The generalization that has emerged from batch cul- Si(OH)., uptake as functionally dependent on media ture and field data is that NH; is preferentially Si(OH), concentration (S), and independent of cel- utilized (Fig. 13), but the actual kinetics of this lular silicon content (Q). interaction have received relatively little attention

222 enzyme is present in NH; -grown cells. The induction process is thought to involve 3 the synthesis of an "activator," which is in direct response to an expo- sure to NO ;i• and an absence of NH; . The evidence for Chlorella is the following: 1) Cyclohexamide will suppress synthesis of ri) the activator when the cells are treated simultaneously with exposure to NO ; (Funkhouser et al. 1980). 2) Deuterium labeling demonstrates that only a small fraction of increased activity can be attributed to de novo synthesis of enzyme following either transfer from NH -r- to NO,i-enriched media or transfer from dark to light (Johnson 1979; Tischner and Hüttermann 1978, respectively). 3) NH; grown cells with almost no active enzyme contain a protein that cross reacts with nitrate reductase antibodies, thus indicating that the inactive T precursor is immunologically related to purified FIG. 13. Typical patterns of NO i and NHI disappear- nitrate reductase (Funkhouser and Ramadoss 1980). ance from culture media resulting from preferential uptake Thacker and Syrett (1972) stated that "indeed of NH1- . it is not yet clear whether the inhibition of NOq reduction (by NH;) results from an inhibition of the (cf. Grant et al. 1967; Eppley et al. 1969; Eppley enzymes of nitrate reduction or from prevention of and Rogers 1970; Packard and Blasco 1974; McCar- NO uptake by the cells." Since that time numerous thy et al. 1977). In still other studies, however, the papers have contributed to our current understanding simultaneous utilization of both NH; and NC-2, has of the effect of NH; on NO utilization, but few been documented (Eppley and Renger 1974; Bienfang have provided clear insight into the actual mechanism 1975; Caperon and Ziemann 1976). The findings by which NH; suppresses either the uptake or the of both preferential or apparently indiscriminant assimilation of nitrate. It has been argued by Chaparro uptake are not contradictory, rather they serve to et al. (1976), for example, that the similarity of the define the degree to which a given concentration effects of both NH; and arsenate in promoting the of NH; will suppress the utilization of NO. reversible inactivation of nitrate reductase lends Again it becomes important to distinguish credence to the hypothesis that the inactivation occurs between results from relatively short-term uptake via uncoupling of photophosphorylation. It remains experiments and those from longer term growth to be seen, however, how the proposed effect on experiments, as well as between nitrogen-limited photophosphorylation can inactivate nitrate reductase and non-nitrogen-limited populations. With short- without having other serious consequences for cel- term experiments the uptake response may represent lular metabolism as well. Ohmori and Hattori (1978) transient forms of behavior in the transport processes, refuted the uncoupling hypothesis by contending whereas in long-term experiments the difference that the suppression of both NOq utilization and N., between supplied and residual nutrient in either a fixation by the presence of NH; in the growth steady-state continuous culture or a batch culture medium resulted from preferential routing of avail- integrated over a period of time equivalent to at able ATP to the glutamine synthesis process, leaving least a few generations presumably reflects the growth insufficient reserves for uptake and assimilation of requirements of the population. nitrogen by other pathways. A similar change in In short-term uptake experiments with nitrogen- ATP levels has been detected in Chlorella (Akimova starved phytoplankton it is possible to observe simul- et al. 1977). However, a change in ATP levels cannot taneously high rates of uptake for both NO and be the only effect of NH .t , because methylammonium, NH; (DeManche et al. 1979). Whether or not the an NH; analogue that is transported into the cell NOq taken up will be reduced or whether or not but not metabolically assimilated, inhibits utilization NO ;;• uptake will continue in the presence of NHt of NO i (Wheeler 1980) and inactivates nitrate depends on whether the nitrate reductase system is reductase as effectively as do NH; and arsenate active, which in turn is related to the sufficiency of (Chaparro et al. 1976). Furthermore, the results the NE11- --N supply. with methylammonium generally support the argu- Results of several studies on the induction and ment that it is NH; rather than a product of NH; repression of nitrate reductase have led to the sug- assimilation, such as glutamine, that inactivates gestion that an inactive precursor protein for this nitrate reductase. It is clear from several studies

223 that even in the presence of NO;7, both NH; and Regardless of the precise nature of the means methylammonium suppress all activities associated by which unicellular algae regulate their nitrate with the nitrate reductase complex NAD(P)H-NR, transport and reduction processes, it is clear from NAD(P)H-cytochrome-c reductase and FMNII2 or many batch culture studies that there is little evidence MVH-NR (see, for example, Diez et al. 1977, for to suggest that NO in the growth medium will be Ankistrodesmus). utilized when the available NH; is sufficient to meet The reduced rate of NO i• utilization in the the nitrogen growth requirement (cf. Syrett 1962; presence of NH; has also been attributed to a re- Morris and Syrett 1963; Eppley et al. 1969; Thacker versible inactivation of nitrate reductase by cyanide and Syrett 1972). In drawing this generalization, (Solomonson 1974; Gewitz et al. 1974). Inactive it is important to distinguish between the results enzyme preparations can be completely reactivated obtained with batch culture, those with continuous by reduction with ferricyanide (Jetschmann et al. culture at steady state, and those with continuous 1972; Solomonson et al. 1973). Hydroxylamine culture in which the steady-state coupling between has a similar effect on nitrate reductase, but higher nutrient uptake and population growth has been concentrations are required than for cyanide (Solo- perturbed by a pulse of nutrient. During exponential monson and Vennesland 1972), and Solomonson growth in batch culture, NH; in the presence of and Spehar (1977) have based a model for the regu- NO may be utilized exclusively as long as the lation of nitrate assimilation on a hydroxylamine and medium concentration is sufficiently high to saturate glyoxylate reaction that yields cyanide. A rather the NH; uptake system. Under these conditions tenuous tenet of this model is that sufficient hydro- nitrate reductase is inactive, and there is no net xylamine accumulates during NH; assimilation to uptake of NOii. Because p approaches V asymp- suppress nitrate reductase activity effectively. totically as a function of S (Fig. 1), the concentration Very few studies of the effect of NH; on NO:i of NH; required to suppress completely NO utili- uptake have used sufficiently controlled experiments zation is difficult to determine with high precision to determine whether the presence of NH; in the from kinetic analyses of uptake data. If for the sake growth media is actually suppressing NO i transport of example one assumes that the K, values for NH; across the plasmalemma. Pistorius et al. (1976) dis- and NO.:i uptake by a certain species are identical, covered that for Chlore/la the degree of inactivation and that the maximum growth rates attainable on of the nitrate reductase in a culture deprived of both forms of nitrogen are also identical, then the nitrogen is not increased when NH; is added. population can remain in an exponential growth Whereas reactivation is stimulated by NQi, an phase at a constant .t as long as the sum of available addition of 1\10 plus NH; yields a level of inacti- NH; and NOà- is in excess of that necessary to satu- vation that is comparable to that attained with NH; rate either uptake system. If there is a lag in NOi alone. In a subsequent series of short-terni time uptake at the beginning of the transition from sole course experiments with Chlore/la, Pistorius et al. dependence on NH; to one of mixed dependence, (1978) observed that NOi. uptake ceases within 5 min possibly resulting from the time required for induction following an addition of NH;, whereas nitrate re- of the necessary enzymes, then the rates of uptake ductase is still highly active up to 60 min later. They and population growth might be unequal for a short interpreted these findings as additional evidence that period. It is likely, however, that compensation for the transport as well as the reduction of NO:; is such a deficit in nitrogen uptake would arise if V„, suppressed by NH; in the growth medium. More- for NH; became enhanced during the transition. over, Tischner and Lorenzen (1979) have demon- The simultaneous utilization of NH; and NO;i- strated with Chlorella that NO :i uptake is totally is well documented. In many cases for which nitro- suppressed within 1 min following an addition of gen-limited chemostat populations have been grown NH; at concentrations as low as 12 ILM, whereas in media containing both NH; and NOiî , the con- an hour later the nitrate reductase activity had only centrations remained at less than the conventional decreased by 50%. limits of analytical detection (Eppley et al. 1971; At this time it is not clear as to the degree to Eppley and Renger 1974; Bienfang 1975; Caperon which the observations and ideas regarding an inac- and Ziemann 1976). By incrementally increasing tivation of nitrate reductase through oxidation, such the dilution rate of the growth chamber to permit as with cyanide, and those regarding an inactive successively faster steady-state population growth constitutive precursor plus an inducible activator rates, the nitrate concentration in the growth chamber are consistent or compatible with each other. In the will rise as the maximum growth rate, is ap- case of the former, no de novo protein synthesis is proached. Unfortunately, data are not available to required to activate the enzyme, whereas in the quantify rigorously the kinetics of the interaction latter a minor component of the enzyme complex between NO;; and NH; uptalce under these con- must be synthesized. ditions. For the sake of simplicity, the patterns of

224 that NH; will be utilized in preference to NO et al. 1970; Brezonik 1972; Con- (cf. Prochazkova J- /NH 4 way 1977; McCarthy et al. 1975, 1977; Eppley et al. 1979), while at least one has shown that this may not always be the case (Conover 1975). In one study with 120 different natural assemblages in the Chesapeake Bay (salinity 2-30%0, temperature 4-29°C) (McCarthy et al. 1975), the sum of N0i- plus NO uptake accounted for <7% of the phyto- NO: plankton nitrogen ration when NH; concentrations exceeded 1 ,uM with only three exceptions (Fig. 15). Media NH4 Concentration The pattern of these data is similar to the hyperbolic form for NO; as a function of NH; concentration (Fig. 14). Clearly there is considerable range in the FIG. 14. Schematic approximation of the simultaneous degree of suppression associated with a particular rates of NIC1 and NH1- uptake as a function of NH; con- NH; concentration, and presumably this is a re- centration. Several assumptions are discussed in the text. flection of the diverse nature of the natural assem- blages studied. For about 10% of the assemblages the NHI uptake and NO uptake over a growth rate degree of suppression was more extreme than the limiting concentration of NH; and a saturating con- generalization given above. centration of NO; can be seen as a pair of compli- Whereas the pattern for NH; suppression of mentary hyperbolas that sum to the same p for any NO utilization is similar for several species and concentration of NH; (Fig. 14). As stated above, natural assemblages, large species-specific differ- however, this is only an accurate representation of ences are apparent in the kinetics of both the inter- the interaction if the uptake of each substrate has action between NO and NO; utilization and in the same Ks value and if each substrate will support those involving urea. Typically, the concentration the same maximum growth rate. A consequence of of NO; in marine water is low. There are exceptions, departure from equality for either term would be an however, such as at the base of the euphotic zone alteration of the paired hyperbolic representation in in oceanic regions of both the Atlantic and the Pacific. Fig. 14. If the maximum growth rate values were Nitrite has received relatively little attention as a to differ such that NH; promoted more rapid growth, source of phytoplankton nitrogen, and to some degree the sum of the v values could be linear with a positive its importance in this regard is not always clear, slope, but if the K, values differed the sum need not as its primary source in the euphotic zone may be a linear function of concentration. be phytoplankton release (Carlucci et al. 1970; One useful approach in investigating the po- Harrison and Davis 1977; Olson et al. 1980). At tential to discriminate between different forms of times, however, it may be the dominant form of nitrogen is to follow the time course of nutrient nitrogen in the nutrition of phytoplankton (McCarthy uptake following a pulse delivery of nutrient. By et al. 1977). observing nutrient concentrations after adding NH; In a few studies designed to examine the inter- to a NO -grown culture of Skeletonema costatum, action between NO and NO; uptake and assimi- Conway (1977) was able to demonstrate a suppression lation by phytoplankton, it has been noted that for of NO uptake at NH; concentration >1 ,u.M. Ditylum brightwellii both forms are utilized simul- Time course data for both nutrient uptalce and cel- taneously at similar rates over wide ranging concen- lular nutrient content are needed in such studies to trations (Eppley and Rogers 1970), but for natural distinguish between utilization that includes assimi- assemblages both were utilized only at low concen- lation and mere transport as DeManche et al. (1979) trations of NO (Harrison and Davis 1977). For observed for NO in the presence of NM- . With another diatom, Thalassiosira pseudonana (66-A) nutrient deprivation, the kinetics of interaction from the central North Pacific gyre, Olson et al. between the two substrates are further complicated (1980) have described the interaction of NO and by the enhancement of V,,. If the regulators of NO; uptake as quasi-competitive. Growth on NO enhancement in the two uptake systems are different, as rapid as that on N0,7, and the K, for was twice then the time courses for changes in activity during uptake of NO was about half of that for NO. These nutrient deprivation and following exposure to nu- differences result in a marked preference for N01, trient may also differ. regardless of substrate concentration, and results of Several field studies with natural assemblages short-term uptake experiments exhibit classical of phytoplanlcton in both marine and freshwater competitive inhibition kinetics. For Chlamydomonas support the generalization from laboratory studies reinhardi, however, the uptake of NO is inhibited

225

100 (till 1

90 10 o o T — 80 - 1 x 0 1 z i 1 a 70 - I ,a) I D 1 01 I N + 60- 0 1 z 1 z % 1 + + 50 - 0 I l in in, t,1 o 0 z z 1 o , in 1 o al 21) z 0I 0 ,,. 30 - 00 °I 4(1 ° I D a) 081 -Ice) 20 - t011 4E,L. D 8 0 o 1 0 1 o 0 o - -0c6o o 0 8o02) o 000 o 0000 edh ° 0 00 0 0 0 0 (4) (2) (2)

1 1 I 1 I I I I 1 1 1 1 1 1 1 1 1 0 2 4 6 8 10 , 12 14 16 18

,ag - at N1-1:1"-N • L -1

FIG. 15. Relationship between the fraction of the total nitrogen utilized by natural phytoplankton assemblages in Chesapeake Bay and the ambient NH; concentration. (From McCarthy et al. 1975.)

completely by the presence of NO 7 (Thacker and et al. (1967) observed that for Cylindrotheca the Syrett 1972). The patterns in NO 7 and NO:7' preference order of nutritional preference proceeded from NH; when both were present in concentrations sufficient to to urea to NO. A similar order of preference can saturate the uptake process are shown schematically be seen in field studies in both Kaneohe Bay (Harvey in Fig. 16. Little attention has been given to the study and Caperon 1976) and Chesapeake Bay (McCarthy of nitrite reductase in phytoplankton. Apparently, it et al. 1977). is similar to nitrite reductase in higher plants in that Not all phytoplankton can utilize urea (Guillard it can be induced by both nitrate and nitrite in the 1963), but for those able to transport and assimilate culture medium, while nitrate, but not ammonium, this substrate the kinetics of uptake can closely suppresses activity in cell-free extracts (Syrett 1981). resemble those for NH; (McCarthy 1972). In one Few data are available for generalization re- recent study it has been shown that a uniform high garding the kinetics of interaction between the uptake potential for urea uptake (V„) is maintained by both of urea and that of other nitrogenous nutrients. Grant Thalassiosira pseudonana and Skeletonema costatum

226 NO NO - 2 unity. Moreover, the RPI values for NO73, NI-1-1 ,

\\s, urea, and NO 7 all converged at values of unity when availability of the more preferred forms was „J, Ditylum brightwellii (Eppley and Rogers 970) insufficient to meet the nutritional needs of the organisms (Fig. 17). Data for Lake Kinneret (McCarthy et al. unpublished manuscript) show an T extremely similar pattern to those of the Chesapeake for NF11- , NO, and urea. The same pattern for NH-; and NO1 is also seen in data for the Southern Ocean, although a high level of NO :i persisted throughout the study (Gilbert et al. unpublished Thalassiosira pseudonana(66-A) manuscript). In a study of the California Current, \ss- (Olson et al. eso) ...... Eppley et al. (1979) found NH; RPI values as low as 0.73, but 75% of their reported values were T >0.95 (max. 13.8). Their NC`,17 RPI values were all less than unity, and, as before, these values con- verged with those for NH1- as the ambient nitrogenous nutrient concentration approached levels known to f Chlamydomonos reinhardi limit the rate of phytoplankton growth. (Thacker and Syrett 1972) For the oceanic regions, which accommodate as much as 80% of the marine phytoplankton pro-

T ductivity, the concentrations of nutrients are too low to undertake similar studies of nutrient preference. FIG. 16. Schematic representation of the three patterns In all likelihood, there is little discrimination between observed for NO ,i and NO disappearance from culture available forms of nitrogen at these low concen- media. Pulse addition of NO 1 equal in concentration to trations, except in the cases of particular species the media NO i• concentration is indicated by an arrow on that cannot utilize certain forms of nitrogen such as the abscissa. urea or some amino acids. while the nutritional state ranges from well nourished NUTRIENTS THAT SUPPLY DIFFERENT on NO ;i• and NO 7 to nitrogen starved (Horrigan and CELLULAR ELEMENTAL REQUIREMENTS McCarthy 1981). This rate of uptake can be —15 times the rate at which NO and N07 are being A question often asked by aquatic ecologists utilized for growth, and the period over which it is is whether particular natural assemblages of phyto- sustained is a function of the nutritional state: minutes plankton are limited by single or multiple nutrients. for the well-nourished population versus several tens Although numerous papers have been written on of minutes for the nitrogen-starved population. field efforts to identify the limiting nutrient in oce- In the Chesapeake Bay study, McCarthy et al. anic, neritic, estuarine, and lake studies, the data (1977) found that a relative preference index (RPI) usually do not lend themselves to unequivocal inter- calculated for each form of nitrogenous nutrient pretation. The purpose of the present discussion is was useful in assessing the competitive interaction not to review this large body of literature, but rather between suitable alternative substrates: to examine the results from some studies of nutrient uptake for which there is a transition from one limiting KV! nutrient to another. In most studies with natural systems, an interest in limiting nutrients has usually Vv RNA , — focused on nitrogen and phosphorus. Rhee (1974, SN, 1978), Goldman et al. (1979), and Terry (1980) Sx examined interactions between uptake capabilities for these two nutrients, and Droop (1974) studied where N, is a particular form of nitrogenous nutrient interactions between phosphate and vitamin B, 2 . N, V is the rate of uptake, and S is the concentration. Although some treatments of this subject have Through this approach it has been demonstrated considered the possibility that population growth that, when NO1- availability is adequate to meet the during transition from one nutrient-limited state to entire nitrogenous demand of the constituent popu- another is regulated in a multiplicative manner that lation, urea may or may not be utilized, while NOi involves both nutrients, recent data indicate that is rejected. For Chesapeake Bay the NH1- RPI values this is not the case for nitrogen and phosphorus were never less than unity, indicating that NH; was (Rhee 1974, 1978; Terry 1980). Clearly, during such never rejected, while those for N0'7 never exceeded a transition there could be a range of growth rates 227 102 • • •

100 • (1. .r• . • • • 0. • .• • 101 • -• gm: • • •. • • • • • • • • • . . .

1 0°

I 1 1 , • 1 Lt11111. 10-, 100 10' 102 EN-nutrient - at N. L -1 10-, 100 10' 102 EN-nutrient 9-t N. L

102 • • # • lo • 1-1 10° e% • • &le • 10 •• • 1-1

tb, • • • J. • • e• • • • • • • 'o- I • 100 •-- •• •

■ • 11 ■ 1 10-, 100 10, 102

eijtttii I I I EN-nutrient p.g- at N. L -1 10-2 I e 1.111 11 10-' too 10' 102 EN-nutrient 1.4 - at N. L -1

• • 100 • • • • • •• 1-1

I h., •• s •

10 — .

10-2 , 10° 10, NH - at N• L -1

Fla. 17. Relationship between relative preference indices (RPI) for nitrogenous nutrients and the concentration of total available nutrient for Chesapeake Bay phytoplankton. (From McCarthy et al. 1977.)

228

for which control is multiplicative, but if it does exist it must represent less than a few percent of the 0.4 7 entire range of growth rates. Redfield (1958) called attention to the similarity c.) o in the elemental N:P content of both well-nourished plankton and the deep oceanic reserves of these cr, elements in the form of dissolved ions. This typical 0.2 elemental composition for well-nourished phyto- 0 plankton is often referred to as the "Redfield ratio." It averages approximately 16, by atoms, and it has o 0. found wide ranging application in models for plankton o productivity that include terms for nutrient limitation. o Goldman et al. (1979) observed that at growth 1 1 1 1 11 • media N:P ratios 15:1 the phytoplankton N:P 0 20 40 60 80 composition reflects the availability of these elements MEDIA N:P ATOMIC RATIO in the media. For an N:P ratio > 50, however, the phytoplankton N:P composition only reached values FIG. 18. Cell quotas for N and P determined at a single as high as the media at the lowest growth rates. steady-state growth rate for Scenedesmus sp. grown with Terry (1980) also observed a lack of correspondence N:P nutrient ratios in the incoming media. (From Rhee between media and cellular elemental composition 1978.) above a media N:P value of 50. Rhee (1978) was able to show a very abrupt transition between nitrogen and phosphorus limitation at a media N:P ratio of about 30 for Scenedesmus (Fig. 18). These data are for a single growth rate, —44% and they clearly demonstrate that for both N and P it is the Q for the nutrient that limits growth that determines g (Fig. 2). Neither g nor "Tu «.) Q are determined by the N:P ratio in the media. Of 2 particular interest is the finding that some poly- phosphate fractions, especially the acid-soluble component, occur in much higher cell concentration 0 during nitrogen limitation than during phosphorus « 8 limitation at the same value for g. Correspondingly, the maximum specific rate of P uptake at a given growth rate is as much as eightfold lower during 0 20 40 60 80 100 120 nitrogen limitation than it is during phosphorus CELLULAR NIP ATOIVIIC RATIO limitation. The maximum potential for NOT, uptake, V,'„, Fi G. 19. Enhanced potential for nutrient uptalce (v,;,) as is an inverse function of the cellular N:P ratio (Fig. a function of cellular N:P ratio for N-limited (N:P < 20) 19). Under N limitation and for a single N:P ratio and P-limited (N:P > 20) populations of Scenedesmus sp. in the medium, V„', is an inverse function of both p. (From Rhee 1978.)

TABLE 2. (From Rhee 1974.)

N/cell P/cell Cell N:P Media N:P ratio % (x 10-1" mol/cell) (X 10-1" mol/cell) (atomic)

1.0 (N-limited) 0.12 455 110 4.1 0.14 539 121 4.5 0.52 863 117 7.4

1900 (P-limited) 0.26 2570 17.9 143 0.32 2223 19.7 111 0.52 2150 30.7 71 0.69 2320 39.3 59 0.83 2350 73.6 32

229 and the N-cell quota (Tablé 2, Fig. 19). Under plankters, but these typically constitute a small P limitation, however, the V„', for NOq at a single portion of the total phytoplankton biomass. media N:P ratio is independent of the N-cell quota The greatest difficulty in extending laboratory although it is still inversely proportional to /L. Where- analysis for nutrient uptake potential to field studies as the results for the N-limited state are like those is that of processing the samples over sufficiently seen before in studies of N-limited growth (McCarthy brief periods to observe a realistic time course re- and Goldman 1979), those for the P-limited state sponse to nutrient pulses. Except in rich coastal or represent another condition for which generalizations estuarine waters, the quantities of biomass are too presented earlier regarding the regulation of V„', for low to permit the use of volumes of seawater that nitrogen do not apply. can be manipulated rapidly. Typically, the time period of interest (minutes) is considerably less than that necessary to terminate the experiment and collect Conclusions the particulate material by gentle filtration. In one coastal study, however, such experiments have shown populations do at times have the potential As can be easily seen in an examination of the that natural - at enhanced rates following even bibliography necessary to document many of the to take up NH1 additions of substrate (Glibert and Goldman findings and concepts discussed in a review chapter trace 1981). By analogy to laboratory culture data, the such as this, much of our current understanding of been used to infer that the nutrient kinetics is based on rather recent work. field observations have populations studied were nitrogen deficient. Laboratory studies with clones isolated from various fuller understanding of the role of nutrients- estuarine, neritic, and oceanic regions have provided A in regulating the primary productivity of the nutrient the substance for increased generalization from the impoverished oceanic waters will also require im- principles of plant physiology to the study of plankton for quantifying the ambient ecology. Moreover, studies with continuous cul- proved methodology As seen above, some phytoplankters can tures maintained at steady state have permitted us nutrient. attain near maximal growth rates in continuous to examine with increasing precision the most steady-state cultures at nutrient concentrations that dynamic aspects of nutrient uptake and assimilation. are below conventional limits of detection. As a One potentially exciting avenue in the pursuit of a consequence, the kinetics of nutrient uptake for such better understanding of the nature of the metabolic species cannot be defined for conditions that range regulation of nutrient transport and assimilation from a moderate to a serious state of nutrient depri- involves the use of analogues and inhibitors (Syrett vation. 1981). The body of information covered in this review In most field studies we are usually far too roots of the investigators who have ignorant of the physiological characteristics of the reflects the it is the work of classically species that constitute the natural assemblage to produced it. Most of but broad-minded physiologists, ecologists, predict with confidence the community response to oriented The future of the field of phy- a change in nutrient availability. To better interpret and oceanographers. ecology lies in the hands of those who field data, we need increased laboratory study with toplankton will be successful both in extending our knowledge the organisms that are typical in natural assemblages. within these subject areas and in closing the gaps This will require more attention to the systematics of the phytoplankton, the nanoplankton in particular, between them. and an increased effort to bring into culture more of the species that are commonly observed in nature References but rarely maintained in our laboratories. At this time there is considerable interest in AKIMOVA, N. I., Z. G. EVSTIGNEEVA, AND V. L. deriving from laboratory experiments specific pat- KRETOVICH. 1977. Regulation of the glutamine terns in cellular elemental composition and physio- metabolism in Chlore/la pyrenoidosa. Regulation of logical potential that have applicability in assessing the glutamine synthetase activity by components of the average nutritional state of the populations that the adenylic system. Biochemistry (USSR) 42: 739— naturally co-occur. The usual problem in drawing 743. and dark inference from elemental composition data collected BitovicutTRA, M., AND E. SWIFT. 1977. Light of nitrate and ammonium by large oceanic is, however, the uncertainty regarding uptake in field studies dinoflagellates: Pyrocystis noctiluca, Pyrocystis the relative contributions attributable to the phyto- fusiformis, and Dissodinium Limnol. Ocea- plankton, other biota, and detritus. Except in bloom nogr. 22: 73-83. conditions the detrital component may be the domi- BIENFANG, P. K. 1975. Steady state analysis of nitrate- nant one. On occasion an abundance of net plankton ammonium assimilation by phytoplankton. Limnol. may permit the selective capture of large phyto- Oceanogr. 20: 402-411.

230 BREZONIK, P. L. 1972. Nitrogen: sources and transfor- of Skeletonema costatum to a single addition of the mation in natural waters, p. 1-47. In H. E. Allen limiting nutrient. Mar. Biol. 35: 187-199. and J. R. Kramer [ed.] Nutrients in natural waters. DAVIES, A. G. 1970. Iron, chelation and the growth of John Wiley, New York, NY. marine phytoplankton. I. Growth kinetics and chlo- BROWN, C. M., D. S. MACDONALD -BROWN, AND S. O. rophyll production in cultures of the euryhaline fla- STANLEY. 1975. Inorganic nitrogen metabolism in gellate Dunaliella tertiolecta under iron-limiting marine bacteria: nitrate uptake and reduction in a conditions. J. Mar. Biol. Assoc. U.K. 50: 65-86. marine pseudomonad. Mar. Biol. 31: 7-13. DAVIS, C. O., N. F. BREITNER, AND P. J. HARRISON. BROWN, E. J., AND R. F. HARRIS. 1978. Kinetics of 1978. Continuous culture of marine diatoms under algal transient phosphate uptake and the cell quota silicon limitation. 3. A model of Si-limited diatom concept. Limnol. Oceanogr. 23: 35-40. growth. Limnol. Oceanogr. 23: 41-51. BROWN, E. J., R. F. HARRIS, AND J. K. KOONCE. DEMANCHE, J. M. 1980. Variations in phytoplankton 1978. Kinetics of phosphate uptake by aquatic micro- physiological parameters during transient nitrogen organisms: derivations from a simple Michaelis- environments. Ph.D. thesis, Oregon State Univ., Menten equation. Limnol. Oceanogr. 23: 26-34. Corvallis, OR. 110 p. BURMASTER, D. E. 1979. The continuous culture of DEMANCHE, J. M., H. C. CURL JR., D. W. LUNDY, phytoplankton: mathematical equivalence among AND P. L. DONAGHAY. 1979. The rapid response three steady-state models. Am. Nat. 113: 123-134. of the marine diatom Skeletonema costatum to changes BURMASTER, D. E., AND S. W. CHISHOLM. 1979. A in external and internal nutrient concentration. Mar. comparison of two methods for measuring phosphate Biol. 53: 323-333. uptake by Monochlysis lutheri grown in continuous DIEZ, J., A. CHAPARRO, J. M. VEGA, AND A. RELIMPIO. culture. J. Exp. Mar. Biol. Ecol. 39: 187-202. 1977. Studies on the regulation of assimilatory nitrate BUTZ, R. G., AND W. A. JACKSON. 1977. A mechanism reductase in Ankistiodesmus braunii . Planta 137: for nitrate transport and reduction. Phytochemistry 231-234. 16: 409-417. DOWD, J. E., AND D. S. RIGGS. 1964. A comparison CAPERON, J. 1967. Population growth in micro-organism of estimates of Michaelis-Menten kinetic constants limited by food supply. Ecology 48: 715-722. from various linear transformations. J. Biol. Chem. 1968. Population growth response of Isochrysis 240: 863-869. galbana to nitrate variation at limiting concentrations. DROOP, M. R. 1968. Vitamin B i, and marine ecology. Ecology 49: 715-721. IV. The kinetics of uptake, growth and inhibition in CAPERON, J., AND J. MEYER. 1972. Nitrogen-limited Monochrysis lutheri . J. Mar. Biol. Assoc. U.K. 48: growth of marine phytoplankton-II. Uptake kinetics 689-733. and their role in nutrient limited growth of phytoplank- DROOP, M. R. 1973. Some thoughts on nutrient limitation ton. Deep-Sea Res. 19: 619-632. in algae. J. Phycol. 9: 264-272. CAPERON, J., AND D. A. ZIEMANN. 1976. Synergistic 1974. The nutrient status of algal cells in con- effects of nitrate and ammonium ion on the growth tinuous culture. J. Mar. Biol. Assoc. U.K. 54: 825- and uptake kinetics of Monochtysis lutheri in con- 855. tinuous culture. Mar. Biol. 36: 73-84. DUGDALE, R. C. 1967. Nutrient limitation in the sea: CARLUCCI, A. F., E. O. HARTWIG, AND P. M. BOWES. dynamics, identification, and significance. Limnol. 1970. Biological production of nitrate in seawater. Oceanogr. 12: 685-695. Mar. Biol. 7: 161-166. 1977. Modeling, p. 789-806. In E. D. Goldberg CARPENTER, E. J., AND R. R. L. GUILLARD. 1971. et al. [ed.] The sea. Vol. 6. Wiley-Interscience, New Intraspecific differences in nitrate half-saturation York, NY. constants for three species of marine phytoplankton. DUGDALE, R. C., AND J. J. GOERING. 1967. Uptalce Ecology 52: 183-185. of new and regenerated forms of nitrogen in primaiy CHAPARRO, A., J. M. MALDONADO, J. DIEZ, A. M. productivity. Limnol. Oceanogr. 12: 196-206. RELIMPIO, AND M. LOSADA. 1976. Nitrate reductase EPPLEY, R. W. 1981. Relations between nutrient assimi- inactivation and reducing power and energy charge lation and growth in phytoplankton with a brief review in Chlorella cells. Plant Sci. Lett. 6: 335-342. of estimates of growth rate in the ocean. Can. Bull. CONOVER, S. A. M. 1975. Nitrogen utilization during Fish. Aquat. Sci. 210: 251-263. spring blooms of marine phytoplankton in Bedford EPPLEY, R. W., AND J. L. COATSWORTH. 1968. Nitrate Basin, Nova Scotia, Canada. Mar. Biol. 32: 247-261. and nitrite uptake by Ditylum brightwellii . Kinetics CONWAY, H. L. 1977. Interactions of inorganic nitrogen and mechanisms. J. Phycol. 4: 151-156. in the uptalce and assimilation by marine phytoplank- EPPLEY, R. W., J. L. COATSWORTH, AND L. SOLOR- ton. Mar. Biol. 39: 221-232. ZANO. 1969. Studies of nitrate reductase in marine CONWAY, H. L., AND P. J. HARRISON. 1977. Marine phytoplankton. Limnol. Oceanogr. 14: 194-205. diatoms grown in chemostats under silicate or ammo- EPPLEY, R. W., AND E. M. RENGER. 1974. Nitrogen nium limitation. IV. Transient response of Chaetoceros assimilation of an oceanic diatom in nitrogen-limited debilis, Skeletonema costatum and Thalassiosira continuous culture. J. Phycol. 10: 15-23. gravida to a single addition of the limiting nutrient. EPPLEY, R. W., E. H. RENGER, W. G. HARRISON, AND Mar. Biol. 43: 33-43. J. J. CULLEN. 1979. Ammonium distribution in CONWAY, H. L., P. J. HARRISON, AND C. O. DAVIS. southem California coastal waters and its role in the 1976. Marine diatoms grown in chemostats under growth of phytoplankton. Limnol. Oceanogr. 24: silicate or ammonium limitation. II. Transient response 495-509.

231 EPPLEY, R. W., AND J. N. ROGERS. 1970. Inorganic GRANT, B. R., J. MADGWICK, AND G. DAL PONT. 1967. nitrogen assimilation of Dityhun brightwellii , a marine Growth of Cylindrotheca closterhon var. Californica plankton diatom. J. Phycol. 6: 344-351. (Mereschk.) Reimann and Lewin on nitrate, ammonia, EPPLEY, R. W., J. N. ROGERS, AND J. J. MCCARTHY. and urea. Aust. J. Mar. Freshw. Res. 18: 129-136. 1969. Half-saturation constants for uptake of nitrate GREENEY, W. J., D. A. BELLA, AND H. C. CURL JR. and ammonium by marine phytoplankton. Limnol. 1973. Effects of intracellular nutrient pools on growth Oceanogr. 14: 912-920. dynamics of phytoplankton. J. Water Pollut. Control EPPLEY, R. W., J. N. ROGERS, J. J. MCCARTHY, AND Fed. 46: 1751-1760. A. SOURNIA. 1971. Light/dark periodicity in nitrogen GUILLARD, R. R. L. 1963. Organic sources of nitrogen assimilation of the marine phytoplankton Skeletonema for marine centric diatoms, p. 93-104. In C. H. costatum and Coccolithus huxleyi in N-limited chem- Oppenheimer [ed.] Symposium on marine micro- ostat culture. J. Phycol. 7: 150-154. biology. Charles C. Thomas, Springfield, IL. EPPLEY, R. W., J. H. SHARP, E. H. RENGER, M. J. HARRISON, P. J., H. L. CONWAY, AND R. C. DUGDALE. PERRY, AND W. G. HARRISON. 1977. Nitrogen 1976. Marine diatoms grown in chemostats under assimilation by phytoplankton and other microorgan- silicate or ammonium limitation. I. Cellular chemical isms in the surface waters of the central North Pacific composition and steady state growth kinetics of Ocean. Mar. Biol. 39: 111-120. Skeletonema costatum. Mar. Biol. 35: 177-186. HARRISON, P. J., AND C. O. DAVIS. 1977. Use of the EPPLEY, R. W., AND W. H. THOMAS , 1968. Comparison of half-saturation constants for growth and nitrate perturbation technique to measure nutrient uptake rates populations. Deep-Sea Res. uptake of a marine phytoplankton. J. Phycol. 5: 375- for natural phytoplankton 379. 24: 247-255. HARVEY, H. W. 1963. The chemistry and fertility of sea FALKOWSKI, P. G. 1975. Nitrate uptake in marine phyto- waters. Cambridge Univ., London. 240 p. plankton (nitrate, chloride) activated adenosine tri- HARVEY, W. A., AND J. CAPERON. 1976. The rate of phosphatase from Skeletonema costatum (Bacillario- utilization of urea, ammonium, and nitrate by natural phyceae). J. Phycol. 11: 323-326. populations of marine phytoplankton in a eutrophic FALKOWSKI, P. G., AND R. B. RivKIN. 1976. The role environment. Pac. Sci. 30: 329-340. of glutamine synthetase in the incorporation of HATTORI, A., AND J. MYERS. 1967. Reduction of nitrate ammonium in Skeletonema costatum (Bacillario- and nitrite by subcellular preparations of Anabaena phyceae). J. Phycol. 12: 448-450. cylindrica. II. Reduction of nitrate to nitrite. Plant FUHS, G. W. 1968. Phosphorus content and rate of growth Cell Physiol. Tokyo 8: 327-337. in the diatom Cyclotalla nana and Thalassiosira HERBERT, D., R. ELSWORTH, AND R. C. TELLING. fluviatilis . J. Phycol. 5: 312-321. 1956. The continuous culture of bacteria; a theoretical FUNKHOUSER, E. A., T. -C. SHEN, AND R. ACKERMANN. and experimental study. J. Gen. Microbiol. 14: 601- 1980. Synthesis of nitrate reductase in Chlorella . I. 622. Evidence for an inactive protein precursor. Plant HORRIGAN, S. G., AND J. J. MCCARTHY. 1981. Urea Physiol. 65: 939-943. uptake by phytoplankton at various stages of nutrient FUNKHOUSER, E. A., AND C. S. RAMADOSS. 1980. depletion. J. Plankton Res. 3: 403-414. Synthesis of nitrate reductase in Chlorella . II. Evidence JETSCHMANN, K., L. D. SOLOMONSON, AND B. VEN- for synthesis in ammonia-grown cells. Plant Physiol. NESLAND. 1972. Activation of nitrate reductase by 65: 944-948. oxidation. Biochem. Biophys. Acta (Amst.) 25: GEWITZ, H. -S., G. H. LORIMER, L. P. SOLOMONSON, 276-278. AND B. VENNESLAND. 1974. Presence of HCN in JOHNSON, C. B. 1979. Activation, synthesis and turnover Chlorella vulgaris and its possible role in controlling of nitrate reductase controlled by nitrate and am- the reduction of nitrate. Nature (London) 249: 79-81. monium in Chlorella vulgaris. Planta 147: 63-68. GLIBERT, P. M., AND J. C. GOLDMAN. 1981. Rapid LIAO, C. F.-H., AND D. R. S. LEAN. 1978. Nitrogen ammonium uptake by marine phytoplankton. Mar. transformations within the trophogenic zone of lakes. Biol. Lett. 2: 25-31. J. Fish. Res. Board Can. 35: 1102-1108. GOLDMAN, J. C. 1977. Steady state growth of phyto- MACISAAC, J. J., AND R. C. DUGDALE. 1969. The plankton in continuous culture: comparison of internal kinetics of nitrate and ammonia uptake by natural and external nutrient equations. J. Phycol. 13: 251- populations of marine phytoplankton. Deep-Sea 258. Res. 16: 45-57. GOLDMAN, J. C., AND J. J. MCCARTHY. 1978. Steady MCCARTHY, J. J. 1972. The uptake of urea by marine state growth and ammonium uptake of a fast-growing phytoplankton. J. Phycol. 8: 216-221. marine diatom. Limnol. Oceanogr. 23: 695-730. 1981. Uptake of major nutrients by estuarine GOLDMAN, J. C., J. J. MCCARTHY, AND D. G. PEAVEY. plants. In B. Neilson and G. Cronin [ed.] Effect of 1979. Growth rate influence on the chemical compo- nutrient enrichment in estuaries. (In press) sition of phytoplankton in oceanic waters. Nature MCCARTHY, J. J., AND J. C. GOLDMAN. 1979. Nitro- (London) 279: 210-215. genous nutrition of marine phytoplankton in nutrient- GOLDMAN, J. L., AND D. G. PEAVEY. 1979. Steady depleted waters. Science (Washington, D.C.) 23: 670- state growth and chemical composition of the marine 672. chlorophyte Dunaliella tertiolecta in nitrogen-limited MCCARTHY, J. J., W. R. TAYLOR, AND J. L. TAFT. continuous cultures. Appl. Environ. Microbiol. 38: 1975. The dynamics of nitrogen and phosphorus 894-901. cycling in the open waters of the Chesapeake Bay,

232 p. 664-681. In T. M. Church [ed.] Marine chemistry uptake, growth rate and polyphosphate in Scenedesinus in the coastal environment. ACS Symp. Ser. 18. sp. J. Phycol. 9: 495-506. 1977. Nitrogenous nutrition of the plankton in the 1974. Phosphate uptake under nitrate limitation Chesapeake Bay. I. Nutrient availability and phyto- by Scenedestnus and polyphosphate in Scenedesmus plankton preferences. Limnol. Oceanogr. 22: 996- sp. J. Phycol. 10: 470-475. 1011. 1978. Effects of N:P atomic ratios and nitrate MONOD, J. 1942. Recherches sur la croissance des cul- limitation on algal growth, cell composition, and tures bactériennes. Hermann and Cie, Paris. 210 p. nitrate uptake. Limnol. Oceanogr. 23: 10-25. MORRIS, I., AND P. J. SYRETT. 1963. The development RICKETTS, T. R., AND P. A. EDGE. 1977. The effect of nitrate reductase in Chlorella and its repression of nitrogen refeeding on starved cells of Platymonas by ammonium. Arch. Mikrobiol. 47: 32-41. striata Butcher. Planta 134: 169-176. RIGANO, C., G. AHOTTA, AND U. VIOLANTE. 1974. NELSON, D. M., J. J. GOERING, S. S. KILHAM, AND Reversible inactivation by ammonia of assimilatory R. R. L. GuILLARD. 1976. Kinetics of silicic acid nitrate reductase in Cyanidiwn Arch. uptake and rates of silica dissolution in the marine caldarium. diatom Thlassiosira pseudonana. J. Phycol. 12: 246- Microbiol. 99: 81-90. RIGLER, F. H. 1966. Radiobiological analysis of inorganic 252. phosphorus in lake water. Int. Ver. Theor. Angew. NICHOLS, G. L., S. M. SHEHATA, AND P. J. SYRETT. Limnol. Verch. 16: 465-470. 1978. Nitrate reductase deficient mutants of Chlamy- SCHLOEMER, R. H., AND R. H. GARRETT. 1974. Nitrate domonas reinhardii. Biochemical characteristics. J. transport system in Neurospora crassa. J. Bacteriol. Gen. Microbiol. 108: 79-88. 118: 259-269. NOVICK, A., AND L. SZILARD. 1950. Experiments with SOLOMONSON, L. P. 1974. Regulation of nitrate reductase the chemostat on spontaneous mutations of bacteria. activity by NADH and cyanide. Biochim. Biophys. Proc. Natl. Acad, Sci. USA 36: 708. Acta (Amst.) 334: 297-308. OHMORI, M., AND A. HATTORI. 1978. Transient change SOLOMONSON, L. P., K. JETSCHMANN, AND B. VEN- in the ATP pool of Anabaena cylindrica associated NESLAND. 1973. Reversible inactivation of the nitrate with ammonia assimilation. Arch. Microbiol. 117: reductase of Chlorella vulgaris Beijerinck. Biochim. 17-20. Biophys. Acta (Amst.) 309: 32-43. OLSON, R. J., J. B. SOOTOO, AND D. A. ICEiFER. 1980. SOLOMONSON, L. P., AND A. M. SPEHAR. 1977. Model Steady-state growth of the marine diatom Thalassiosira for regulation of nitrate assimilation. Nature (London) pseudonana. Uncoupled kinetics of nitrate uptake and 265: 373-375. nitrite production. Plant Physiol. 66: 383-389. SOLOMONSON, L. P., AND B. VENNESLAND. 1972. Nitrate reductase and chlorate toxicity in Chlorella PAASCHE, E. 1973. Silicon and the ecology of marine plankton diatoms. I. Thalassios ira pseudonana vulgaris Beijerinck. Plant Physiol. 50: 421-424. (Cyclotella nana) grown in a chemostat with silicate SYRETT, P. J. 1962. Nitrogen assimilation, p. 171-188. as the limiting nutrient. Mar. Biol. 19: 117-126. In R. A. Lewin [ed.] Physiology and biochemistry of algae. Academic Press, New York & London. PACKARD, T. T. 1979. Half-saturation constants for nitrate reductase and nitrate translocation in marine phyto- 1981. Nitrogen metabolism of microalgae. Can. plankton. Deep-Sea Res. 26A: 321-326. Bull. Fish. Aquat. Sci. 210: 182-210. SYRETT, P. J., AND C. R. HIPKIN. 1973. The appearance PACKARD, T. T., AND BLASCO. D. 1974. Nitrate reductase of nitrate reductase activity in nitrogen-starved cells activity in upwelling regions. 2. Ammonia and light of Ankistrodesmus braunii. Planta (Berl.) 111: 57-64. dependence. Tethys 6: 269-280. TAFT, J. L., M. E. LOFTUS, AND W. R. TAYLOR. 1977. PERRY, M. J. 1972. Alkaline phosphatase activity in Phosphate uptake from phosphomonoesters by phyto- subtropical central North Pacific waters using a sensi- plankton in Chesapeake Bay. Limnol. Oceanogr. 22: tive fluorometric method. Mar. Biol. 15: 113-119. 1012-1021. PISTORIUS, E. K., E. A. FUNKHOUSER, AND H. Voss. TERRY, K. L. 1980. Nitrogen and phosphorus requirements 1978. Effect of ammonium and ferricyanide on nitrate of Pavlova lutheri in continuous culture. Bot. Mar. utilization by Chlorella vulgaris . Planta 141: 279- 23: 757-764. 282. THACKER, A., AND P. J. SYRETT. 1972. The assimilation PISTORIUS, E. K., H.-S. GEWITZ, H. Ross, AND B. of nitrate and ammonium by Chlatnydomonas rein- VENNESLAND. 1976. Reversible inactivation of nitrate hardi. New Phytol. 71: 422-433. reductase in Chlorella vulgaris in vivo. Planta 128: TISCHNER, R., AND A. HUTTERMAN. 1978. Light- 73-80. mediated activation of nitrate reductase in synchronous PROCHAZKOVA, L., B. BLAZKA, AND M. KRALOVA. Chlorella. Plant Physiol. 62: 284-286. 1970. Chemical changes involving nitrogen metab- TISCHNER, R., AND H. LORENZEN. 1979. Nitrate uptake olism in water and particulate matter during primary and nitrate reduction in synchronous Chlorella. Planta production experiments. Limnol. Oceanogr. 15: 146: 287-292. 797-807. WHEELER, P. A. 1980. Use of methylammonium as an RAO, K. P., AND D. W. RAINS. 1976. Nitrate absorption ammonium analogue in nitrogen transport and assimi- by barley. II. Influence of nitrate reductase activity. lation studies with Cyclotella ctyptica (Bacillario- Plant Physiol. 57: 59-62. phyceae). J. Phycol. 16: 328-334. REDFIELD, A. C. 1958. The biological control of chemical WILLIAMS, P. J. L. 1973. The validity of the application factors in the environment. Am. Sci. 46: 205-221. of simple kinetic analysis to heterogenous microbial RHEE, G. 1973. A continuous culture study of phosphate populations. Limnol. Oceanogr. 18: 159-164. 233 Adaptation of Nutrient Assimilation

R. C. DUGDALE, B. H. JONES JR., AND J. J. MACISAAC Department of Biological Science, Allan Hancock Foundation, University of Southern California, Los Angeles, CA 90007, USA

AND J. J. GOERING Institute for Marine Science, University of Alaska, College, AK 99708, USA

Introduction and nitrate uptake using ' 5 N as a tracer (MacIsaac and Dugdale 1969) and for silicic acid using "Si (Goering Nutrient uptake clearly is a key process in the et al. 1973) and "s Ge (Azam and Chisholm 1976) growth of marine phytoplankton, and the ability as tracers. Perry (1976) observed a trend toward of a species to integrate changes in the nutrient hyperbolic uptalce using 33 P as a tracer. In the labo- environment into its own synthetic process with ratory, Michaelis-Menten type kinetics have been minimal disruption may determine its success in obtained with various species of marine algae for relation to other species and groups of species. The inorganic nitrogen uptake (e.g. Eppley et al. 1969), objective of this paper is to assess current knowledge for silicic acid uptake in diatoms (Paasche I973a; of the changes in nutrient uptake and assimilation Azam 1974; Nelson et al. 1976), for inorganic that occur in response to changes in nutrient con- phosphate uptake (Rhee 1973; Perry 1976), and for centrations, irradiance, and temperature. Adaptation carbon (Goldman et al. 1974; Caperon and Smith is used here in the sense that the organism responds, 1978). passively or actively, in some manner to circumvent Natural populations of marine phytoplankton or /ameliorate the stress imposed by environmental have been shown to obey Michaelis-Menten kinetics changes linked to nutrient uptake and utilization. in relation to light intensity for uptake of am- Adaptation, then, is the end result of a chain of also monia and nitrate (MacIsaac and Dugdale 1972) events that occur with a range of time scales. Cog- and of silicic acid (Goering et al. 1973; Azam and nizance of these time scales and their relationship Chisholm 1976). Thé same relationship has been to the spectra of variability in marine environments shown in the laboratory for nitrogen uptake (Mac- is essential to an understanding of the significance Isaac et al. 1979). From these results and from our of adaptation phenomena in marine phytoplankton, published results on nitrogen uptake it appears that and further is often the key to effective communi- both light- and substrate-limited uptake of nitrate cation between laboratory and field scientists, and and silicic acid follow virtually identical patterns. even between groups of the latter. To the extent Lehninger (1971) points out that active transport possible, these relevant time scales will be indicated through cell membranes often shows kinetics that throughout this communication. are similar to enzyme kinetics, an observation amply the studies of phytoplankton nutrient Scientific Background borne out by uptake reported above. Moreover, active transport of two major Dugdale (1967) suggested the use of the systems are considered to be composed specific carrier and the energy- Michaelis-Menten expression to describe the uptake components, the Guerrero (1979) of a limiting nutrient by marine phytoplankton. The transferring component. Losada and reduction in relation expression is: reviewed the literature on nitrate to photosynthesis. These authors concluded that the V = enzyme system was simpler than that involved with Vm " K, + S carbon reduction with only two enzymes, nitrate where V„, is the nutrient-specific uptake rate, V, reductase and nitrite reductase, using ferredoxin at infinite substrate concentration; S is the concen- or pyridine nucleotide as the electron donor. Further, tration of limiting autrient; K, is the half-saturation the enzymatic activity appears to be tightly bound constant, the concentration of limiting nutrient at to pigment-containing particles, with the evidence which V = Vim', /2. for a direct link much better for nitrite reduction Evidence for Michaelis-Menten type kinetics than for nitrate reduction. From energetic consider- has been obtained from experiments with natural ations, Losada and Guerrero pointed out that nitrate populations of marine phytoplankton for ammonia assimilation can proceed at the energy level of fer- 234 redoxin or pyridine nucleotide in contrast with carbon Laboratory Observations reduction, which requires energy at a higher level , i.e. ATP. Regulation of nitrate assimilation appears Studies of Skeletonema costaturn (Greve.) Cleve to be targeted to nitrate reductase with ammonia grown under silicic acid and ammonia limitation were as a key element. Most conclusions reached by Losada made in our laboratory and the research has been and Guerrero (1979) appear to be at least consistent reported by Davis (1973), Harrison (1974), and with the laboratory and field results reported below. Conway (1974). It has been standard practice in our Our view of nutrient interactions is essentially experiments to measure all the primary nutrients, the same as that of Droop's (1974) threshold concept, in contrast with most other studies of phytoplankton i.e. only one nutrient is limiting at a given time. in continuous culture where only the limiting nutrient The uptake rate of the limiting nutrient is set externally has been measured (Fuhs 1969; Carpenter 1970; by the concentration of that nutrient and the uptake Caperon and Meyer 1972; Paasche 1973b; Eppley rates of nonlimiting nutrients are controlled internally and Renger 1974; Goldman and McCarthy 1978). by the cell to correspond with the externally set, Consequently, , we have obtained some insight into limiting-nutrient uptake rate. To distinguish the mode the interactions between inorganic phosphorus, silicic of control of uptake of a cellular component it is acid, and inorganic nitrogen uptake under chemostat convenient to designate the externally controlled culture conditions (Harrison et al. 1976; Conway et specific uptake rate as V, and internally controlled al. 1976; Dugdale 1977). specific uptake rate as I/1 . Because light appears nearly identical with a limiting nutrient in its effect on LIMITING AND NONLIMITING NUTRIENT nutrient uptake, the threshold concept may be applied UPTAKE INTERACTIONS to it also (Dugdale and MacIsaac 1971). The basic theory and our supporting experimental results con- Conway et al. (1976) were able to distinguish cerning nutrient interactions and phytoplankton between uptake rates associated with external and growth processes are presented below. A glossary of internal contiol, as limitation shifted from one varia- terminology and symbols is on p. 248. ble to another, and to observe the effect of this shift on

0.2

A 1 A • 1 B • 2A lb. 0 2B

4à • 6 0 -c 1 •

o jr • A •1A 0111 • Ulu • 1140 • • :A 111 0) t) 0 delle 19, 00 • 0•Le. A2p8P A.‘e•0 Aeie p 0 MO 0 a e • • • A • °

aglitehmsele" 0 • .) II • • 0 6 12 T (h) FIG. 1. Silicic acid uptake rate versus time from perturbation in four experiments. (From Breitner unpublished ms.)

235 uptake rates of nonlimiting nutrients.. Their evidence centration at which the transition occurred is called was obtained from perturbation experiments in which Sr for convenience, and the resulting hyperbola a large amount of the limiting nutrient was added to a referred to as truncated. These relationships are shown chemostat reactor, the subsequent decline with time in Fig. 3b. observed, and the specific uptake rates calculated The perturbation experiments yielded values of from the time series. The time course of uptalce in four V; for the original limiting nutrient that were dif- replicate experiments with S. costatum and silicic ferent from the values of V, for the nonlimiting acid limitation (N. Breitner unpublished ms) is shown nutrients. Specifically, in Table 1, values of V, in Fig. 1. A high rate or "spike" of uptake occurs for limiting silicic acid were about twice the dilution initially, followed by a period of reduced, but con- rate. The values of V I for the nonlimiting nutrients, stant, uptake rate, V i . Finally, nutrient-limited however, remained approximately equal to the uptake, 1/,, , occurs when the substrate concentration initial dilution rate. The agreement between dilution has fallen to a sufficiently low level. The same uptake rate and V, for nonlimiting nutrient uptake indicates data are plotted against substrate concentration in Fig. strong cellular control of these rates. 2, where the three phases of uptake can be distin- Perry (1976), using chemostat cultures of Tha- guished clearly with the transition between Ve and lassiosira pseudonana Hasle and Helmdal (formerly V, occurring at a silicic acid concentration of about Cyclotella nana) clone 66-H, found that saturated 2 btg-at • L- ' . phosphate uptake rates, on a per cell basis, were From these results , and from ''N uptake experi- 10 times higher under phosphate limitation than ments made on ammonia-limited cultures of S. cos- under nitrogen limitation. Her observations and tatum, the limiting-nutrient uptake hyperbolas were those of Rhee (1974) indicate that maximal phos- shown to be truncated at the point where a transition phate uptake is controlled under limiting nitrogen between V,. and V I for the limiting nutrient occurred, in similar fashion to that shown above for limiting i.e. at the point where a new variable (probably silicic acid control of maximal phosphate and nitrate light) became limiting (Fig. 3a). The substrate con- uptalce.

•••■

1 I I I I I I 6 12 S (pg-at • 1: 1 ) FIG. 2. Silicic acid uptake rate versus reactor silicic acid concentration in four experiments with overlaid weighted regression curve. (From Breitner unpublished ms.)

236

A TABLE 2. Kinetic parameters obtained from perturbations

0.08 conducted on the 30 and 100% light populations. and K values are given ± 1 standard deviation.

0.06 1 from 7.; z 0.04 C onway et al.1976 Light Vmax K, level (ly- min- ') (h- ') 83 0.02 ps-at Si -L- ' (h- ')

0.14 0.101 ± 0.02 1.81 ± 0.89 0.074 0 1.0 2.0 3.0 4.0 5.0 0.04 Silicate Concentration 0.055 (pg-at.L-1 ) 0.14 0.127 ± 0.01 1.1 ± 0.2 0.088 0.04 0.051

Rhee (1973, 1974) suggested a physiological basis for the control of phosphate uptake in Scene- V1 for limiting nutrient desmus sp. Under nitrate or phosphate limitation, •c phosphate uptake apparently is controlled by the intra- cellular concentration of inorganic, acid-soluble operating VI for nonlimiting nutrients V or D polyphosphate; noncompetitive inhibition kinetics I j I I are followed. An abrupt change occurred in the slope I I of the maximal phosphate uptake versus cellular P/N

oper4 ST+ ratio for Scenedesmus sp. at a value of P/N of 0.033 ating conc. where P limitation changed to N limitation and vice versa. In a later paper, Rhee (1978) suggested that the internal concentrations of free amino acids control FIG. 3. The distinction between V,. and VI shown by nitrate uptake under phosphate limitation. On the chemostat perturbation experiments (A) and schematically basis of his studies of Scenedesmus sp. Rhee considers (B). The data points above the truncation point, Sr, give an estimate of VI for the nutrient limiting the chemostat the threshold concept the most appropriate to describe at the operating dilution rate, and those below are values the interaction between limiting nutrient and non- of V,. From steady-state chemostat theory and the concept limiting nutrient uptake. of balanced growth, the dilution rate, D, is equivalent to growth, it, and to V; and in the Conway et al. (1976) INTERACTIONS OF LIGHT AND experiment, D 0.04 h- ' and lies in the V, range. Only LIMITING SILICIC ACID UPTAKE the values of G are useful for estimating the K, and Vma0 values of the full hyperbola. The strong cellular control of nonlimiting- nutrient uptake to levels that correspond to the exist- ing limiting-nutrient uptake has been demonstrated above. Davis (1973, 1976) was able to show further that the limiting-nutrient (silicic acid) concentration at which truncation occurred, ST, was determined by TABLE 1. Mean uptake rates of phosphate, nitrate, and light intensity and silicic acid during two silicic acid limited perturbation suggested that the resulting values experiments. Number of observations in parentheses. (From of V; for silicic acid approximated light-limited maxi- Conway et al. 1976.) mal growth rates. The silicic acid uptake kinetics obtained by Davis (1976) are summarized in Table 2. The V, values give a saturation Dilution Vino 111,„ curve when plotted against light rate as shown in Fig. 4a. There was good agreement (h- ') (h- ') (11-1 ) at all but the highest light intensity between these values of V1 and those for growth rate as a function of light intensity (Fig. 4b) as measured by McAllister et 0.040 0.040 0.040 0.070 al. (1964). (15) (15) The value of 1/1 measured for limiting silicic 0.041 0.038 0.040 0.060 acid in Davis' (1976) perturbation experiments was (15) (15) 0.081 , set by the light intensity at which the culture was grown (Table 3). The values of V; for

237

A in this case). On the same time scale the V, for nitrate appeared to be strongly dependent on light intensity and the V I for phosphate showed a similar 0.075 relationship.

m0.050 ADAPTATIONS TO NUTRIENT LIMITATION AND EFFECTS ON NUTRIENT UPTAKE KINETICS

0.025 A primary adaptation of marine phytoplankton species to nutrient limitation occurs through the reduction of growth rate, it, and of cell quota, Q 0.05 010 0.14 (cellular concentration of limiting nutrient). The ly • minl variation in the cell quota of the limiting nutrient is related under some circumstances to the dilution (growth) rate in the chemostat at steady state by the expression (Droop 1968):

0.10 (2) = (1

0.05 where Kg is the subsistence cell quota of the limiting nutrient at zero growth rate, and p,,'„ is an abstract maximum growth rate, different from the real ,LL max of the cell. The maximum realizable growth rate is 0.10 0.20 0 30 0.40 always less than p,„1„ as pointed out by Goldman 4•min -1 and McCarthy (1978). reduction of cell quota with nutrient limi- FIG. 4. (A) Maximal silicic acid uptake as a function of The light intensity. Symbols denote different chemostat culture tation has been documented in chemostats for vitamin experiments and time periods within experiments. (B) B12 (Droop 1968), for nitrate (Caperon and Meyer Growth rate as a function of light intensity, with selected 1972), for ammonia (Conway 1974), for phosphorus data from (A) plotted with the curve for growth from (Fuhs 1969; Rhee 1973), and for silicic acid (Paasche McAllister et al. (1964). All data for S. costaturn. The 1973b; Harrison 1974). One of the consequences lowest values were obtained under non-steady-state con- of this response is a possible increase in the maximum ditions by measuring washout rates. (From Davis 1976.) specific uptake rate, V,„„, or V, if truncation has occurred, of the limiting nutrient (Dugdale 1977). TABLE 3. Nutrient uptake rates and growth rates from This result can be seen from the relationship between 9 h batch uptake experiments. Rates are per hour; samples V and the cellular uptake rate, p: taken from chemostats at steady state at D = 0.04 h' and light intensity indicated. Q

Source Incub. Similarly, ,

light light Vro VN0a Vsli , V111110 = - Q 100 100 0.043 0.081 0.049 0.011 0.060 30 0.027 0.077 0.030 0.010 0.040 where p,„„, is the maximum capacity of the cell to 15 0.032 0.078 0.022 0.019 0.041 take up nutrient. The effect of decreasing Q is to 1 0.025 0.073 0.015 0.017 0.032 increase the initial slope of the specific uptake hyper- bola with decreasing nutrient concentrations, and nonlimiting phosphate and nitrate for the full light thereby to hold up the specific uptake rate and growth intensity reactor were 0.043 h - ' and 0.049 h- ' , respec- rate at low concentrations. The effect may result tively, demonstrating again that the V, values for in virtually undetectable limiting nutrient in the the nonlimiting nutrients were controlled internally chemostat at all moderate dilution rates. at the initial dilution rate, 0.04 h - ' . An additional Using inorganic nitrogen as the example, if feature of the data in Table 3 is the relative inde- Pmx remains constant rises hyperbolically with pendence of the V I for silicic acid on light intensity decreasing Q v . This effect was shown by Dugdale when the latter was reduced for short periods (9 h (1977) using data of Eppley and Renger (1974)

238 0.6 MULTIPLE GROWTH PATTERNS

• 4.NO3 Another adaptive response to nutrient limitation 0 v ni.NH4 observed in chemostats is that a species may show 0.4 more than one pattern of growth. These pattems X V COMPUTED 7 \\\\ are readily distinguishable by the differing slopes, p.„„ and intercepts, KO of their ,u,Q vs. Q plots - E > (the linearization of the Droop hyperbola, Eq. 2). 0.2 Cells showing this response are described by Droop (1974) as "slow adapted" when exhibiting the lower x L1 values of and KO , and "fast adapted" with the

higher p,„, and K0 . This change in growth patterns o 0 0.1 0.2 was observed in our laboratory for S. costatutn grown under silicic acid limitation (Harrison 1974) (pg -at N•ce11 -1 ) and under nitrogen limitation (Conway 1974; Harri- FIG. 5. Relationship between Q and V„,„ with constant son et al. 1976), and for Monochrysis lutheri under p„,;,,, data from Eppley and Renger (1974) as plotted with phosphorus and vitamin B1.2 limitation by Droop a computed curve of the relationship in Dugdale (1977). (1974). The p,Q vs. Q plots for the two species V„; is the same as are shown in Fig. 7 where it appears that a similar phenomenon was being observed. for Thalassiosira pseudonana (13-1) under nitrogen limitation (Fig. 5). Using ' 5 1\1 as a tracer in a more detailed study, McCarthy and Goldman (1979) • obtained the same result, also with T. pseudonana 0.32 (3 H). In both studies V„,„, tended to rise with de- creasing Q,v , , and the measured curves rose a little 0.24 more steeply than that calculated with constant p„,„,. 7 0.16 Brown and Harris (1978) found, in Nostoc sp. and Selenastrum capricornutum grown in batch 'CO 0.08 culture, that phosphate uptake per cell (pp) varied inversely with the cell quota, Q,., rising quite rapidly in the vicinity of K( . Rhee (1974) found in Scene- 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 desmus sp. that an abrupt change in maximal phos- Q (pg-at N x10 -7 ) phate uptake rate, p„,„„„ occurred with the onset of phosphate limitation; under these conditions of low cellular P/N ratio, ,0 11111%r increased to 8 times over the value characteristic of nitrogen-limited growth. Under silicic acid limitation, S. costatum showed a different pattern, Fig. 6, with pis, de- creasing with decreasing Qs1 ; however, the net effect is still an increase in Vis, with decreasing Qsi .

o

014

0.08 7'

004

o O 05 10 FIG. 7. Multiple growth patterns. (A) For S. costatum. Qs, pg -at x10-'•CELL' Circles represent dilution rates <0.05 h-1 , crosses represent FIG. 6. The relationship of 1/1 and pi for silicic acid dilution rates >0.08 h- ' . (From Harrison et al. 1976) (B) For uptake in S. costatum to cell quota, Q, for silicic acid. M. lutheri. D = p, under steady-state chemostat condi- (Data from Harrison 1974, table 13.) tions. (From Droop 1974)

239 A phenomenon closely related or identical with of their experimental technique, and thus the char- that described above has been reported for bacteria acterization of T. pseudonana as a species not (Herbert et al. 1956) and for yeast (Mian et al. 1969; showing multiple states of adaptation might be pre- Button et al. 1973). All these workers observed mature. higher growth rates for these organisms in chemostat The important feature exhibited by slow-adapted culture than the maximum growth rates found in batch cells in the context of Dugdale (1977) is the large culture. The terminology used to describe this mode increase in maximal specific uptake rate that occurs of adaptation has varied. For example, Droop (1974) and the resulting ability to grow relatively rapidly used "fast-adapted," Button et al. (1973) used at very low nutrient concentrations; the disadvantage "II., -adapted," and we (Harrison et al. 1976) used incurred is a reduced 1-Lim, • For example, in the "shifted-up" to refer to cultures exhibiting this en- S. costatum data of Dugdale (1977, Fig. 8) the hanced bt,„„, state. In the case of a marine yeast, slow-adapted cells have a of 0.05111 - ' and the D. Baton (personal communication) found that fast-adapted cells have a ,u„,„, of 0.098 h-1 . These -adapted cells resume their original slower values of maximum realizable growth rate were growing state when the chemostat dilution rate is calculated from an equation developed by solving lowered appropriately. This evidence suggests that the Droop hyperbolic equation (Eq. 7 in Dugdale in his experiments the adaptation was a physiological 1977) for Q, and substituting the resultant expression response and not selection occurring within the with Eq. 13 in Dugdale; ,u,„„, is taken as the value chemostat. of ,U, when silicic acid concentration is infinitely The growth response characteristics of M. large. The equation is: lutheri cells growing in the two modes were described KQ 1 by Droop (1974) and are paraphrased in the following: (3) IL - r c:i The state of adaptation will remain unchanged so Pmax [Si]m' long as the appropriate dilution rate is maintained. Ks; [Sil If the population has been maintained in one mode The effect on the nutrient concentration versus growth for a long time, it cannot adapt to a sudden large rate curve calculated from the above equation for change in dilution rate across the threshold value these two sets of parameters is shown in Fig. 8. that distinguishes the two states. On the other hand, The slow-adapted populations apparently can main- change in if the population has undergone recent tain respectable growth rates at concentrations of the state of adaptation, the population can adjust silicic acid of 0.05 itg-at • L -1 and less, and these sudden change its mode almost immediately to a growth rates are about twice those of the fast-adapted in dilution rate across the threshold value. It possibly cells in this low nutrient range. (No allowance has is significant that the threshold dilution rate is similar been made in the calculations or Fig. 8 for an S o , to p,, for the slow-adapted mode (a correspondence the silicic acid concentration observed in some observed as well in our work with S. costatum). experiments below, in which no uptake or growth Some adaptations to low nutrient conditions takes place.) known to occur in microorganisms are discussed Doyle (1975), in developing the case for selec- by Tempest and Neijssel (1976). Among these are tion of cells or cell lines under varying nutrient (1) synthesis of alternate high-affinity pathways for conditions, predicted that a given genetic line would uptake and assimilation of the growth-limiting have either low V., and low K, for low-nutrient nutrient, (2) modulation of nonlimiting nutrient environments, or high V., and high Ks for high- uptake, and (3) modulation of macromolecular syn- nutrient environments. For selection to be effective thesis to allow growth at submaximal rates. The then it is necessary that the growth (or uptake) rate- Droop model has an empirical basis and has the nutrient concentration curves cross each other. Thus, serious drawback that physiological adaptations one type of organism is selected for under a low- cannot be interpreted easily in terms of cell quota nutrient regime and the other under a high-nutrient and vice versa. Nevertheless, it seems virtually regime. The evidence shown above for S. costatum certain that some or all adaptations mentioned above suggests that this organism maintains both options would be expressed as changes in the variables of within a single clone (Fig. 8). There is the suggestion the cell quota model. also that, in our experiments, selection in the chem- Goldman and McCarthy (1978) found no evi- ostat may have been responsible for the appearance dence for such response in their data obtained with of populations exhibiting different growth states. Thalassiosira pseudonana (3 H) grown under nitro- gen limitation. It is not clear in the Goldman and PHOTOSYNTHESIS, NUTRIENT UPTAKE, McCarthy work that the slow change of dilution AND NUTRIENT LIMITATION rate necessary for adjustment from one mode to the An understanding of the relationship between other (Button et al. 1973; Droop 1974) was a part the photosynthetic rate, nutrient limitation, and 240 0.10,

FAST-ADAPTED

• r) 0.05 SLOW-ADAPTED

0 1.0 2.0 3.0 4.0

Si - L Si (OH)4 - »g at • 1 FIG. 8. Theoretical effect of different states of adaptation on the nutrient concentration versus growth rate curve. nutrient uptake is essential to an understanding of and phosphate limitation and found a single hyper- the interaction between nutrient and light limitation. bolic curve to describe the relationship between Many more data are available on the effect of light assimilation number and dilution rate. Chlorophyll reduction under nutrient-limited growth on chloro- per cell increased with dilution rate under each phyll concentration and on maximum photosynthetic nutrient limitation. rates than for maximum nutrient uptake rates. Three generalizations about the photosynthetic Therefore, the discussion of laboratory results will response to nutrient limitation emerge from the be limited largely to this area. Some necessary above data: (1) the chlorophyll a /carbon ratio in- definitions are (1) PI , the assimilation number is creases linearly with the growth rate, shown for the rate of photosynthesis at saturating light intensities nitrogen by Caperon and Meyer (1972) and Laws (per unit chlorophyll, g Cg chl a' • tr" ') and (2) and Wong (1978), and for phosphorus by Rhee a, the initial slope of the photosynthesis per (1978); (2) the chlorophyll a /cell ratio increases unit chlorophyll versus I curve. more or less linearly with growth rate, shown for Several investigators have explored the relation- nitrogen by Caperon and Meyer (1972) and Thomas ship between assimilation number and nutrient- and Dodson (1972), and for phosphorus by Rhee limited growth but the published results do not show (1978); S. costatum under silicic acid limitation a clear pattern. Thomas and Dodson (1972) have showed no particular trend of chlorophyll a /cell shown that for Chaetocerus gracilis the assimilation with growth rate (Harrison 1974); (3) with nitrogen number increases with increasing growth rate under limitation, the assimilation number generally remains nitrogen-limited growth. Senft (1978) has shown constant (Eppley and Renger 1974) or increases that the assimilation number increases with increasing (Thomas and Dodson 1972) with increasing growth cell quota for phosphorus in Anabaena wisconsinense rate. The results of the last two generalizations indi- and Chlorella pyrenoidosa. On the other hand, cate that the maximum photosynthetic rate per cell Eppley and Renger (1974) saw no relationship increases with increasing growth rate under nutrient between the growth rate and the assimilation number limitation. for Thalassios ira pseudonana grown in a chemostat The results of studies of the assimilation number under nitrogen limitation. Laws and Wong (1978) as a function of the light intensity at which the phyto- did not observe an increase in the productivity index plankton are grown have been more consistent than with increasing growth rate for Monochrysis lutheri studies of the assimilation number as a function and Dunaliella tertiolecta , but observed a decrease of nutrient-limited growth rate. Yentsch and Lee for Thalassiosira allenii, also grown in nitrogen- (1966) demonstrated with Nannochloris atomus in limited chemostats. Laws and Bannister (1980) semicontinuous culture that the assimilation number grew Thalassiosirafluviatilis under nitrate, ammonia, increased with increasing mean light intensity. Using

241 Phaeodactylum tricornutum in a turbidostat culture, creased. Such an increase in limiting nutrient per Beardall and Morris (1976) also showed that the cell results in a specific uptalce hyperbola with a assimilation number increased with increasing mean lower initial slope as well as a lower V., , with light intensity. In their experiments, the increased the result that with constant limiting nutrient con- assimilation number resulted from both an increase centration, the growth rate would be reduced. In in maximum rate of photosynthesis per cell and a the chemostat where growth rate is set by the dilution decrease in chlorophyll a per cell with increased rate, the end result is an increase in the limiting intensity. Laws and Bannister (1980) also showed nutrient concentration in the reactor. Harrison (1974) with T. fluviatilis that assimilation number increased observed the sanie sequence of events also with and chlorophyll per cell decreased with increasing S. costatum grown under silicate limitation when he light intensity under nutrient-saturated conditions. reduced the temperature from 18 to 12°C. However, the curve of growth rate versus assimilation number had a very different shape from that obtained THE TRUNCATION MODEL by the saine authors for the nutrient-limited condition. As a result, the assimilation number of natural Lhniting light — The possible interrelationships populations cannot be used as an index of growth between light intensity, , limiting nutrient concen- rate unless the growth-limiting factor is known. tration, and nonlimiting nutrient concentration are In the experiments by Davis (1976) to study illustrated for the condition of limiting silicic acid and the effect of reduced light intensity on limiting suboptimal light intensity in Fig. 9. The maximum silicic acid uptake using chemostat cultures of S. growth rate, it., ,- may be described by a saturation costatum reduction of light intensity from 0.14 ly • curve, as indicated in Fig. 9a, where the average light min to 0.042 ly • min" resulted in a doubling of intensity is seen to set the specific net carbon uptake silicic acid/ cell and a reduction of V, adap- rate: tation took place slowly over a period of days. In one experiment, chlorophyll per cell increased by

about 50%; in the other, where rapid adaptation (4) V mn . occurred, an initially high chlorophyll/cell de- ° = + I

7

0

4 ST [S 1]

d

ma x

<}.1 max 1■•■■ 0

1 INST INST K Q si

FIG. 9. The truncation model with figures arranged to illustrate interactions. See text for explanations.

242 The value of V,. corresponding to T, shown by the Light and limiting silicic acid — Dependence arrow in Fig. 9a, is an internally controlled rate, of limiting silicic acid uptake on light intensity is Vi „ in the sense that further increases in light in a weak or nonexistent as shown in Fig. 9d, and the short-term experiment would not result in an in- values of Vs3 obtained during silicic acid saturated crease in photosynthetic rate (a definition of perturbations at all light levels approximate Vis, = Because respiration would reduce the carbon available a. • for growth, the vs. I curve would fall somewhat Light and nonlimiting nutrient —Light de- below the V,. vs. / curve. (This area is complex and pendence of nitrate uptake is hyperbolic, but truncated care must be taken in making definitions and meas- at Viso , as shown in Fig. 9e. The same pattern urements of the carbon variables. For the present it is expected to hold for phosphate uptake. will suffice to note that a net V,, should be equivalent If Droop kinetics are followed, there is a hyper- to the maximum realizable growth rate, and ,unini. bolic relation at steady state between the growth that it should be possible to calculate V,„ net, by rate and the cell quota, Fig. 9f, according to the subtracting respiration from the gross maximal photo- expression synthetic rate.)

Limiting silicic acid — The point marked ST , (Qs1 — Kam) in Fig. 9b is the truncation point for the uptake hyper- (7) = bola, designating the boundary between external ksi control expected in this case by average light intensity = Vi,„„ = V, = Ves, and internal control. At the silicic acid concentration indicated by the arrow, Under silicic acid limitation, the arrow in Fig. 9f indicates the operating point for Qs1 . Because cell VIWIN çf [Si] quota decreases with decreasing p and (5) VS/ = Ves ' K,,,« «+ [Si] P max gi Vma 5si , the shape of the limiting-nutrient the specific growth rate, hyperbola on a specific basis changes, rather than since in this example the concentrations of nitrate staying constant as shown in the diagrammatic and phosphate (latter not illustrated) are set to be representation of Fig. 9b. The growth rate is calcu- nonlimiting. If the concentration of limiting nutrient, lated from Eq. 3 rather than from Eq. 5. S i , is greater than ST, then Light and fixed limiting-nutrient concentration — The effect of light intensity on growth rate at Vs Vi = Vic. = 'max fixed limiting-nutrient concentration is expressed in Eq. 3 through the value of ft,'„. From Eq. 3 an ex- Nonlimiting nutrient — The arrow on the ab- pression for Ilia>, can be obtained by letting the scissa in Fig. 9c designates the ambient concentration silicic acid concentration become infinitely large: of nitrate in a silicic acid limited system. Under such conditions, (K0 1 P'111ZIS Pm. 14, 1 1/3 ,„„,v0 „ [ NO3 ] (6) > vos , K50,, + [NO3 ] and because we have equated p,,,„, with VI ,, and VV03 — V,,0, — V,50, K0 1 (8) Vic, = = — + = 1-13;111x i.e. the V,v0., vs. [NO3] curve is truncated at V„.,. P 1015 gin The same will be true for phosphorus. For balanced There is insufficient information available to say growth with silicic acid limitation, whether light intensity and temperature influence ,a,'„ directly as suggested by Dugdale (1977) and = Vos, = V,„, = whether K0 and pn,„, are affected directly by changes in these variables. However, Eq. 8 can be solved and the nutrient-saturated uptake rates of nonlimiting for p,,'„ to give: nutrients, VIN „, or can be used as measures of Iles, . The nitrate uptake kinetic constants are 1 influenced by ammonia concentration (e.g. Conway (9) 1977). For simplicity, the concentration of ammonia 1 K0si is assumed here to be O. Pmaxs,

243 •

A reduction in / results in a decrease in ,u,,'„ and a ANTON BRUUN CRUISE 15 - 1966 change in shape of the p, vs. Qs; curve, Fig. 9f, and DROGUE EXPERIMENT a new operating point at a higher Q would be estab- NEAR PTA. SAN NICOLAS, PERU lished if K0s, and p,„„,,, remain unchanged. In any event, it appears from laboratory experi- ments that the shape of the ,u, vs. Q hyperbola is 0.08 affected by changes in both temperature and light intensity, a reduction in either at fixed growth rate 7 .c resulting in increased cell quota of the limiting nu- - 0.06 trient (Harrison 1974; Davis 1973). Field Observations +CI 0.04 Nitrogen often is postulated to b'e the limiting O nutrient in various parts of the marine environment (e.g. Ryther and Dunstan 1971; Eppley et al. 1972; co 0.02 Thomas et al. 1974; Perry 1976; Goldman 1976). However, limitation by other primary nutrients such as phosphorus (Perry 1976), silicic acid (Thomas and Dodson 1975; Azam and Chisholm 1976), or by light (Huntsman and Barber 1977) may occur 0 5 10 15 20 and also can generate signals in the nitrogen uptake Si(OH)4 - Si (,ug -at • t.:1 ) data, which may be interpreted in the framework of the truncation model described above. The fol- • 250 lowing discussion of results from our work in the 7 Peru upwelling system demonstrates sequences of 2001— 7 events influencing the uptake of nitrate and assimi- co 574-100% 150 577-100% lation number by changes in the stability of the water 580-50% column and reductions in the ambient concentrations cr) 1001586-100% of silicic acid. • 586-25% INORGANIC NITROGEN UPTAKE 501— A drogue was placed near the Peru coast at about 5 10 15 20 15°S on Anton Baum 15 and followed for 5 d in March 1966. Although drogues cannot track a water Si(OH)4 - Si (pg -at • L:1 ) mass faithfully in an upwelling area, they provide FIG. 10. Inorganic nitrogen and carbon uptake as functions sampling targets related to the mean drift at the of silicic acid concentration. Carbon, chlorophyll a, and drogue depth. Ryther et al. (1970) described the silicic acid data were taken from Ryther (1966). evolution of the diatom bloom that occurred during this drogue experiment. The three most abundant Two drogue experiments were made in the same species were Chaetoceros debilis, C. lorenzianus, region during the JOINT II expedition to Peru on and C. socialis. Specific inorganic nitrogen uptake the RV Wecoma in March 1977 and the results rates were measured using ''N on water taken from described by Brink et al. (1980). The first of these the 50% light-penetration depth. The ambient nitrate was initiated near the end of a calm period when concentrations were nonlimiting, such that the uptake diatom populations (primarily Detomda puinilla rates measured may be designated VI . When plotted according to D. Blasco personal communication) against nitrate, phosphate, and silicic acid concen- and uptake rates were high. The values of 1/10. trations, the specific nitrate uptake rates (I/;, ) for samples from the 50% light-penetration depth are showed no likely relation to the first two, but gave plotted together with those from Anton Brutal 15 a hyperbolalike curve with silicic acid (Dugdale and in Fig. 11. The agreement between the two data Goering 1970). The summed V,, 0 . (am- sets is good, and these results suggest control of monia concentrations were brought to nonlimiting nonlimiting nitrate uptake by silicic acid concen- levels by additions of NH,C1) showed a positive trations of up to at least 10 p. g-at • L- ' . The conditions correlation with silicic acid concentration, Fig. 10a for observing this interaction between silicic acid (Dugdale 1972). For the same data set, the carbon concentration and nitrate uptake during the Wecoma assimilation number (n) also was correlated posi- drogue 1 experiment were especially favorable: (1) tively with silicic acid (Fig. 10b). winds remained low, contributing to stability in the 244 concentration resulting in increased nitrate uptake rate or growth on the side of the bottle, the major "bottle effect" appeared to be simply the stabilization of the population at natural saturating light intensity. The characteristically low uptake rates observed off northwest Africa during the JOINT I cruise would increase significantly during a 24-h holdover when the samples were collected from a deeply wind-mixed water column. The Wecoma drogue 1 experiments showed a decrease in V; „, during the 24-h holdover (Fig. 13). Nutrient concentrations measured at the begin- ning and end of the 24-h period showed silicic acid to decline and the inorganic nitrogen concentrations to vary but little. The silicic acid concentrations 10 20 decreased in the bottles just as occurred in nature. The data in Fig. 13 show that the 24-h holdover - at • L Si (OH) 4 Pg values predict accurately the direction of change FIG. 11. Nitrate and silicic acid uptake in the Pentvian in the value of V; ,„ in the regular, nonholdover upwelling system as functions of silicic acid concentration. samples. Presumably, the nitrate uptake rates in Open circles are silicic acid uptake from JOINT II; solid the holdover bottles and along the drogue track circles are nitrate uptake from JOINT II; X's are nitrate were being set by limiting silicic acid uptake. uptake from Anton Bruun 15. A second drogue experiment was initiated at the same point immediately following termination of the water column and leading to high-average light first. Strong upwelling occurred just prior to the conditions for near-surface populations, (2) ammonia experiment (Brink et al. 1980) flooding the area concentrations were below levels where a significant with nutrients; the diatom populations were replaced suppression of nitrate uptake would be expected, by small unidentified flagellates. Low values of and (3) the nitrate/silicic acid ratio has been shown Vi„.„ were obtained throughout the experiment and to be high in the upper region of the pycnocline in no relationship between them and silicic acid con- Peru (Dugdale 1972), so that upwelling from this centration could be seen. Further, the 24-h holdover part of the pycnocline can be expected to result in experiments showed enhancement in V,,,„, similar surface waters characterized by excess nitrate in re- to that seen in the JOINT I experiments, reflecting lation to silicic acid. The section at 15°S from the the observed physical instability of the system during PISCO cruise (RV Thomas G. Thompson , March the drogue 2 experiment. 1969), Fig. 12a, clearly shows vertical and horizontal gradients in the nitrate/silicic acid ratio; where SILICIC ACID UPTAKE upwelling is strongest, near stations 14 and 15, the Goering et al. (1973) obtained the first evidence ratio is about 1.6-1.8. The nitrate section, Fig. 12b, for Michaelis-Menten kinetics for silicic acid uptake shows surface concentrations to be 15-20 g-at in natural populations using "Si as a tracer. (Two with active upwelling; the silicate section, Fig. 12c, stable isotopes of silicon, "Si , and "Si, are available shows surface concentrations in the nearshore region for use in tracer experiments; Goering has used both to be about the same as for nitrate. at different times and now uses :105j) At station 51 We routinely performed holdover experiments of the PISCO cruise to Peru, they found V„,„, = on shipboard to aid in the interpretation of temporal 0.075 11 - ' and K„ = 2.93 tr, g-at • L-1 . During the changes observed in nitrate uptake. Water from the JOINT II expedition, the results of a "Si Ks meas- 50% light-penetration depth was obtained on the urement at Wecoma station 124 gave a value for 0800 productivity station and placed in two bottles K, of approximately 3 p, g-at • L' silicic acid and of with 50% light screens. One bottle was inoculated V„,„,, approximately 0.08 . Azam and Chisholm immediately with ''N-labeled nitrate or ammonia, (1976) obtained values of 1.59 and 2.53 iu,g-at • incubated for 6 h in a deck incubator, and filtered silicic acid for K„ in two natural populations in the (MacIsaac and Dugdale 1972). The other bottle was Gulf of California. Harrison (1974) summarized held in the deck incubator until the next morning. silicic acid kinetic observations from culture and At that time ''N was added and the bottle returned reported a range of values for K„ of 0.7-3.37 gg-at to the incubator for 6 h before filtering. Although a • I,- ' . Azam (1974) found a K„ of 6.8 p, g-at • number of changes may occur within the bottle during silicic acid for the nonpelagic diatom Nitzschia alba the 24-h preincubation, e.g. decrease in ammonia in culture. The range of values for both laboratory

245 A Stations Numbers Stations Numbers 18 17 16 15 14 18 17 16 15 14 o 5.0 10.0 15.0 50 50 20.0

25.0 100 -100

1-150 -150 • a26.2) h-26.0

- 1.0 .200 .200

28.0 .250 ei 1-2501 <1.0 o 0.8 30.0

.300 300

i-350 350

> 1.0 1-400 400

NO3 SiO4

'" 450 1 15 450 105 90 75 60 45 30)5 15 105 so 6 0 45 30 1 Distance ( rom shore Ikmi Distance tram shore lkml Stations Numbers 18 17 16 15 14 ,

50

-100

-150

.200

.250 o

1-300

-350

-400

0 450 105 75 60 45 30 15 Distance from shore Ikrni

FIG. 12. Nitrate and silicic acid in the Peruvian upwelling system during the PISCO cruise. (A) Nitrate/silicic acid ratio; (B) nitrate; (C) silicic acid. (From Dugdale 1972.) and field populations is strikingly small, and both The uptake of ""Si was measured along with the values and the range are similar to those reported the ' 5 N experiments during the Wecoma drogue 1 for nitrogen and tabulated by Dugdale (1976). experiment. The silicic acid enrichments used in

246 0.06- 0041 • - \ NO3 (Ste. 19)

_ 0.04- •------/x/ Si(OH)4 (Stu 19) 0.03-

0.02 H . 1 JOINT IE WECOMA DROGUE I EXP

Tc 0.02- . I 0 1 1 1 1 1 15 16 17 19 24

STATION NO. • .--o Si(OH)4 (Sto.15) FIG. 13. Nitrate uptake as measured at sampling time • (closed circles) and 24 h later (open circles), in samples /c. collected at the same time from the Wecoma drogue 1 001 - experiment in Peru. Stations were -24 h apart. / I the experiments were high enough to be nonlimiting, although even the highest ambient concentrations (12 tg-at •L" ) appeared to be limiting. Values of 0 ei)'T V11, for the 50% light-penetration depth populations o 50 100 are against ambient silicic acid concentration % SURFACE LIGHT in Fig. 11. Two possible explanations for the observed increase in Vis, with decreasing silicic acid con- FIG. 14. Silicic acid (open circles, Sta. 15; solid circles, centration are (1) changes in nonnutrient environ- Sta. 19) and nitrate (X's, Sta. 19) uptake as a function of mental variables leading to enhanced biological light, from two stations during the Wecoma drogue 1 ex- rates, and (2) decreasing silicon per cell (cell quota). periment. The latter effect is consistent with decreasing growth rate, as implied by decreasing ViN01 , with decreasing this section is 14.3 - 26.4 ,ug-at • L- ' ; from the labo- silicic acid concentration (Fig. 11). ratory, the computed range is 4.32-52.7 ,ug-at • L" Two curves of light versus silicic acid uptake, (ignoring the value for N. alba). These calculations Fig. 14, also were obtained in the course of the give results that are consistent with the field observa- Wecoma drogue 1 experiment. The large increase tions reported above that suggest observable effects in dark uptake of silicic acid between the two on phytoplankton uptake processes with silicic acid experiments coincided with a decrease in ambient concentrations up to at least 10 gg-at • L-1 . There is silicic acid concentration. This observation of light the possibility that uptake may cease at some nonzero independence of limiting silicic acid uptake is con- value of silicic acid concentration, So (Paasche sistent with the chemostat results of Davis (1976) 1973a, b), and in that case the values of K I ,„, and Nelson et al. (1976). Nitrate uptake as a function calculated above should be increased by S o (no of light also was measured with the second of the greater than 2 ,ug-at • L - ' in Paasche's data); Nelson et light versus silicic acid experiments (station 19), and al. (1976) did not observe an So . the results are shown in Fig. 14. Two features are Although the Klim value should provide a good of interest and indicate that nitrogen was nonlimiting: indicator of expected changes in the dynamics of (1) dark uptake was low, and (2) reciprocal plot nutrient uptake processes in phytoplankton popu- analysis of the kinetic parameters showed the curve lations as silicic acid concentrations fall, effects on to be truncated between 30 and 50% surface light. growth rate may not be realized until lower con- A parameter K 1 11 , the substrate concentration centrations are reached. For example, the K„ for at which V/V„,„, = 0.9, has been used (Nelson silicic acid limited growth reported by Guillard et al. et al. 1976) to estimate the ambient silicic acid (1973), 0.19 and 0.98 ,ug-at • L-1 for two clones of concentrations below which significant and ob- T. pseudonana, compare with K, values of 0.8 and servable reductions in uptake may be expected to 1.5 ,ug-at • L for uptake in the same clones (Nelson occur. On this basis, the range of K I ,,„ predicted et al. 1976). Paasche (1973b) found approximately from the field values of K, given at the beginning of the sanie relationship, and it appears that the ranges

247

for K I ,„, for uptake given above may be divided by half-saturation constant, substrate about 1.5-4.0 to obtain an estimate of K i ,„, for growth concentration at which V = V„„,,/2 under silicic acid limitation. (gg-at • L ) K I ,„, the limiting substrate concentration at which = 0.9 Acknowledgments K, the light intensity at which the = (ly • min ' or This paper is based on Technical Report No. 51 microeinsteins ni s ') (Dugdale et al. 1979) of the Coastal Upwelling subsistence quota for limiting nutrient at zero

Ecosystem Analysis program prepared under Grant growth rate ( p.g - at • cell ') No. OCE-7727006 from the National Science Found- absolute substrate transport rate, usually on ation IDOE program. The assistance of D. Boisseau a per cell basis (gg-at•cell ' • h ') and G. Grunseich in conducting experiments and absolute substrate transport rate at infinite analyzing the samples for, respectively, the Wecorme substrate concentration (gg-at • cell ' • h ') Si and 'N work is deeply appreciated. Support for p, constant, absolute substrate uptake rate the present effort was provided also by the National following initial high uptake in a perturbation Science Foundation under Grant No. OCE-8015705 experiment (gg-at•cell ' • h ') from the IDOE program and under Grant No. OCE- phytoplankton carbon concentration 7910112. (either gg C•cell ' or gg C L ) Pi` the photosynthetic rate per unit phytoplankton biomass (usually g Cg chl a -1 • h- ') Symbols PP, the assimilation number at optimum light (usually g C• g chl a • h ') ✓ specific uptake rate (h ') P„„,, photosynthetic rate at saturating light I (IS (g C• cell '•11 ') V -= — — which, during balanced growth, S a the initial slope of the P" vs. / curve - ' 1 dx (g C• g chl a -1 h' microeinsteins.m equals •s- ')

P,, the particulate silicon concentration where S is the phytoplankton particulate (gg-at Si • L or gg-at Si •cell ') substrate concentration and .1- is the cell concentration (no. cells/unit volume) ✓ ,, specific uptake rate at infinite substrate References concentration (h ') V. the externally controlled specific uptake A/AM, F. 1974. Silicic-acid uptake in diatoms studied rate (h ') with [68GE] germanic acid as tracer. Planta (Berl.) 11, the internally controlled specific uptake 121: 203-212. rate (h ') A/AM, F., AND S. W. CHISHOLM , 1976. Silicic acid substrate concentration (gg-at • L ') uptake and incorporation by natural marine phyto- Si substrate concentration at which control plankton populations. Limnol. Oceanogr. 21: 427— of nutrient uptake switches from external 435. to internal (gg-at • L ') BEARDALL, J., AND I. MORRIS. 1976. The concept of • the dilution rate of the culture, equal to the light intensity adaptation in marine phytoplankton: sonie experiments with Phaeodadylinn it-km.111nm. flow rate of medium into and out of the Mar. Biot. 37: 377-387. chemostat divided by the culture volume (h ) BRINK, K. H., B. H. JONES JR., J. C. VAN LEER, C. N. K. • the mean light intensity in the culture MOOERS, D. STUART, M. STEVENSON, R. C. (ly • min ' or microeinsteins • m '• s ') DUGDALE, AND G. HEBURN. 1980. Physical and • cell quota, the concentration of limiting biological structure and variability in an upwelling nutrient per cell (gg-at • cell ) center off Peru near 15° during March 1977. In F. growth rate (h ') Richards [ed.] Coastal upwelling-1980. American Geophysical Union. ILw maximum apparent growth rate obtained from plots of the linearized Droop hyperbola; BROWN, E. J., AND R. F. HARRIS. 1978. Kinetics of an abstract variable, greater than g„„,, (h ') algal transient phosphate uptake and the cell quota concept. Limnol. Oceanogr. 23: 35-40. gm- the maximum (nutrient- and light-saturated) BuTroN, D.K., S. S. DUNKER, AND M. L. MORSE. growth rate realizable under the given 1973. Continuous culture of Rhodotorula rubra: culture conditions (h ') kinetics of phosphate-arsenate uptake, inhibition, and /1 111,1\ the maximum (nutrient-saturated) growth phosphate limited growth. J. Bacteriol. 113: 559— rate realizable under light limitation (h ') 611.

248 CAPERON, J., AND J. MEYER. 1972. Nitrogen-limited DUGDALE, R. C., AND J. J. MACISAAC. 1971. A com- growth of marine phytoplankton: changes in population putation model for the uptake of nitrate in the Peru characteristics with steady-state growth rate. Deep- upwelling region. Invest. Pesq. 35: 299-308. Sea Roi. 19: 601-618. EPPLEY, R. W., Er AL. 1972. Evidence for eutrophication CAPERoN, J., AND D. F. SmiTH . 1978. Photosynthetic in the sea near Southe rn California coastal sewage rates of marine algae as a function of inorganic carbon outfalls - July, 1970. Calif. Coop. Oceanic Fish. concentration. Limnol. Oceanogr. 23: 704-708. Invest. Rep. 16: 74-83. CARPEN I ER, E. J. 1970. Phosphorus requirements of EPPLEY, R. W., AND E. H. RENGER. 1974. Nitrogen two planktonic diatoms in steady state culture. J. assimilation of an oceanic diatom in nitrogen-limited Phycol. 6: 28-30. continuous culture. J. Phycol. 10: 15-23. CONWAY, H. L. 1974. The uptake and assimilation of EPPLEY, R. W., J. H. ROGERS, AND J. J. MCCAW! HY. inorganic nitrogen by Skeletonema cosialum (Greve) 1969. Half-saturation constants for uptake of nitrate Cleve. Ph.D. thesis, Univ. Washington, Seattle. and ammonium by marine phytoplankton. Limnol. 125 p. Oceanogr. 14: 912-20. 1977. Interactions of inorganic nitrogen in the Fuus, W. G. 1969. Phosphorus content and rate of growth uptake and assimilation of nitrate by marine phyto- in the diatoms Cyclotella nana and Tha/assiosira plankton. Mar. Biol. 39: 221-232. . J. Phycol. 5: 312-321. CONWAY, H. L., P. J. HARRISON, AND C. O. DAVIS. GOERING, J. J., D. M. NELSON, AND J. A. CARTER. 1976. Marine diatoms grown in chemostats under 1973. Silicic acid uptake by natural populations of silicate or ammonium limitation. II. Transient response marine phytoplankton. Deep-Sea Res. 20: 777-789. of Skeleumema cvslanun to a single addition of the GOLDMAN, J. C. 1976. Identification of nitrogen as a limiting nutrient. Mar. Biol. 35: 187-199. growth-limiting nutrient in wastewaters and coastal DAVIS, C. 0. 1973. Effects of changes in light intensity marine waters through continuous culture algal assays. and photoperiod on the silicate-limited continuous Water Res. 10: 97-104. culture of the marine diatom Skeleionema cos/an/In GOLDMAN, J. C., AND J. J. McCARTHY. 1978. Steady (Greve) Cleve. Ph.D. thesis. Univ. Washington, state growth and ammonium uptake of a fast-growing Seattle. 122 p. marine diatom. Limnol. Oceanogr. 23: 695-703. 1976. Continuous culture of marine diatoms under GOLDMAN, J. C., W. J. OSWALD, AND D. JENKINS. silicate limitation. II: Effect of light intensity on growth 1974. The kinetics of inorganic carbon-limited algal and nutrient uptake of Skeleionema costaium. J. growth. J. Water Pollut. Control Fed. 46: 554-574. Phycol. 12: 291-300. GuiLLARD, R. L., P. KILHAM, AND T. A. JACKSON. DOYLE, R. W. 1975. Upwelling, clone selection, and the 1973. Kinetics of silicon-limited growth in the marine characteristic shape of nutrient uptake curves. Limnol. diatom Thalassiosira pseudonana Hasle and Heimdal Oceanogr. 20: 487-489. (= Cyclotella nana Hustedt). J. Phycol. 9: 233-237. DRooP, M. R. 1968. Vitamin B-12 and marine ecology. HARRISON, P. J. 1974. Continuous culture of the marine IV: The kinetics of uptake, growth and inhibition in diatom Skeletonema costalum (Greve). Ph.D. thesis, Monochruis futile/i. J. Mar. Biol. Assoc. U.K. 48: Univ. Washington, Seattle. 140 p. 689-73i. HARRISON, P. J., H. L. CONWAY, AND R. C. DUGDALE. 1974. The nutrient status of algal cells in con- 1976. Marine diatoms grown in chemostats under tinuous culture. J. Mar. Biol. Assoc. U.K. 54: 825- silicate or ammonium limitation. I: cellular compo- 855. sition and steady-state growth kinetics of Skelewnema DUGDALE, R. C. 1967. Nutrient limitation in the sea: eostaium. Mar. Biol. 35: 177-186. dynamics, identification, and significance. Limnol. HERBERT, D., R. ELSWORIFI, AND R. C. TELLING. Oceanogr. 12: 685-695. 1956. The continuous culture of bacteria: a theoretical 1972. Chemical oceanography and primary and experimental study. J. Gen. Microbiol. 14: 601- productivity in upwelling regions. Geoforum 11: 622. 47-61. Hurg FSMAN, S. A., AND R. T. BARBER. 1977. Primary 1976. Nutrient cycles, p. 141-172. In D. H. production off northwest Africa: the relationship to Cushing and J. J. Walsh [cd.] The ecology of the sea. wind and nutrient conditions. Deep-Sea Res. 24: Blackwell, London. 25-33. 1977. Modeling, p. 789-806. In E. D. Goldberg LAWS, E. A., AND T. T. BANNISTER. 1980. Nutrient- et al. [cd.] The sea. Vol. 6. Ideas and observations on and light-limited growth of Thalassiosim fi uv/ai/li.r progress in the study of the seas. John Wiley & Sons, in continuous culture, with implications for phyto- Inc., New York, NY. plankton growth in the ocean. Limnol. Oceanogr. 25: DUGDALE, R. C., AND J. J. GoEitiNG. 1970. Nutrient 457-473. limitation and the path of nitrogen in the Peru current E. A., production. Ailloli Bruun Rep. 4: 3-8. Texas A & M LAws, AND D. C. L. WONG. 1978. Studies of Press. carbon and nitrogen metabolism by three species of DUGDALE, R. C., B. H. JONES JR., J. J. MACISAAC, AND marine phytoplankton in nitrate-limited continuous culture. J. Phycol. 14: 406-416. J. J. GOERING. 1979. Interactions of primary nutrient and carbon uptake in the Peru current: a review and LEHNINGER, A. L. 1971. Bioenergetics. Benjamin/Cum- synthesis of laboratory and field results. Coastal Up- mings. 245 p. welling Ecosystems Analysis (CUEA) Program Tech. LOSADA , M., AND M. G. GUERRERO. 1979. The photo- Rep. 51. 40 p. synthetic reduction of nitrate and its regulation, p.

249 365-408. In J. Barber led.] Photosynthesis in relation sp. J. Phycol. 9: 495-506. to model systems. Elsevier, Amsterdam. 1974. Phosphate uptake under nitrate limitation MAcIsAAc, J. J., AND R. C. DUGDALE. 1969. The kinetics by Scenedesmus sp. J. Phycol. 10: 470-475. of nitrate and ammonia uptake by natural populations 1978. Effects of N:P atomic ratios and nitrate of marine phytoplankton. Deep-Sea Res. 16: 45-57. limitation on algal growth, cell composition, and 1972. Interactions of light and inorganic nitrogen nitrate uptake. Limnol. Oceanogr. 23: 10-25. uptake in the sea. Deep-Sea Res. 19: 209-232. RY HIER, J. H. 1966. Cruise report R/V Anion Bruun , MAcIsAAc, J. J., G. S. GituNsEic0 , H. E. GLOVER, AND cruise 15. Texas A & M Univ. Mar. Lab. Spec. C. M. VE.N .isci t. 1979. Light and nutrient limitation Rep. 5. 53 p. nitrogen and carbon tracer in Gonyaulax exearata: RY I HER, J. H., AND W. M. DUNal AN. 1971. Nitrogen, results, p. 107-110. In D. Taylor and H. Seliger led.] phosphorus and eutrophication in the coastal marine Proceedings second conference on toxic dinotlagellate environment. Science (Washington, D.C.) 171: 1008- blooms. Elsevier, Amsterdam. 1013. MCALLISTER, C. D., N. SHAH, AND J. D. H. STRICK- RY1 wilt, J. H., D. W. MEN/EL, E. M. HULBEIVI . , C. J. LAND. 1964. Marine phytoplankton photosynthesis LoREN/EN, AND N. CORWIN. 1970. Production and as a function of light intensity: a comparison of meth- utilization of organic matter in the Peru coastal current. ods. J. Fish. Res. Board Can. 21: 159-181. Texas A & M Univ. Amon Braun Rep. 5: 3-12. McCAit-11tv, J. J., AND J. C. GOLDMAN. 1979. Nitro- Skisi , W. H. 1978. Dependence of light-saturated rates genous nutrition of marine phytoplankton in nutrient- of algal photosynthesis on intracellular concentrations depleted waters. Science (Washington, D.C.) 203: of phosphorus. Limnol. Oceanogr. 23: 709-718. 670-672. TEMPEST, D. W., AND O. M. NELISSEL. 1976. Micro- MIAN, F. H., Z. FENCL, AND A. PROKOP. 1969. Growth biological adaptation to low-nutrient environments, rate and enzyme activity in yeast (Candida utilis), p. 283-296. In A. C. R. Dean et al. [ed.] Continuous p. 105-115. In E. O. Powell et al. [ed.] Continuous and new fields. Ellis Harwood culture of micro-organisms. culture 6: applications Ltd., Chichester. NELSON, D. M., J. J. GOERING, S. S. KILHAM, AND R. L. GUILLARD. 1976. Kinetics of silicic acid THOMAS, W. H., AND A. N. DODSON. 1972. On nitrogen uptake and rates of silica dissolution in the marine deficiency in tropical Pacific oceanic phytoplankton. diatom Thalassiosira psewlonana. J. Phycol. 12: Il. Photosynthetic and cellular characteristics of a 246-252. chemostat-grown diatom. Limnol. Oceanogr. 17: PAASCHE, E. I973a. Silicon and the ecology of marine 515-523. plankton diatoms. I. Thalassiosira pseudonana 1975. On silicic acid limitation of diatoms in (Cyclotella nana) grown in a chemostat with silicate near-surface waters of the eastern tropical Pacific as limiting nutrient. Mar. Biol. 19: 117-126. Ocean. Deep-Sea Res. 22: 671-677. I973b. Silicon and the ecology of marine plankton TIIONIAS, W. H., D. L. R. SEIREI2T, AND A. N. DODSON. diatoms. II. Silicate uptake dynamics in five diatom 1974. Phytoplankton enrichment experiments and species. Mar. Biol. 19: 262-269. bioassays in natural coastal seawater and in sewage PERRY, M. J. 1976. Phosphate utilization by an oceanic outfall receiving waters off Southern California. Est. diatom in phosphorus-limited chemostat culture and Coast. Mar. Sci. 2: 191-206. in the oligotrophic waters of the central north Pacific. YENTSCH, C. S., AND R. W. LEE. 1966. A study of Limnol. Oceanogr. 21: 88-107. photosynthetic'light reactions, and a new interpretation RHEE, G. 1973. A continuous culture study of phosphate of sun and shade phytoplankton. J. Mar. Res. 24: uptake, growth rate and polyphosphate in Scenedesinus 319-337.

250 Relations between Nutrient Assimilation and Growth in Phytoplankton with a Brief Review of Estimates of Growth Rate in the Ocean

R. W. EPPLEY

institute of Marine Resources, University of California, San Diego, La Jolla, CA 92093, USA

Nutrient Assimilation and Growth mental conditions. Nevertheless, there is a special case in which growth in a cyclostat or in natural THE CONCEPT OF BALANCED GROWTH populations may be expected to be nearly balanced. That is when average rates of assimilation of the "Growth is balanced vvhen the amounts of all various nutritional elements are considered over the cellular components increase exponentially at the cell cycle imposed by the environment (Shuter 1979). same rate. Under these conditions, cellular com- The natural day–night cycle is a principal environ- position remains fixed" (Shuter 1979). The assimi- mental forcing function that imposes periodicity on lation of the various nutritional elements will then phytoplankton growth and elemental assimilation take place in the same proportion as they occur in rates. Exponential growth of natural populations may the cellular composition. Measurement of the rate be balanced if rates are averaged over this 24-h light constant for any one of these elements will then be cycle, unless temporal variations in nutrient supply valid for all, including cell division (Eppley and rates are overriding or unless short-term variations Strickland 1968). in irradiance or temperature within the 24-h cycle The concept of balanced growth is fundamental are sufficiently large. to the use of nutrient assimilation rates as measures of the rate constants of phytoplankton growth. The concept is not usually stated explicitly in texts and reviews on algal physiology although it is implied GROWTH IN THE OCEANS throughout. It is not likely to be a strange or new Discontinuous supply — Sunlight is obviously idea. Nevertheless, the conditions under which discontinuous and periodic on both a daily and a balanced growth can be expected are not very com- seasonal basis. Entrainment of phytoplankton photo- mon, even in studies with laboratory cultures. It is synthesis to the daily cycle results in circadian perio- doubtful that balanced growth of phytoplankton ever dicity in photosynthesis (Doty and Oguri 1957) with takes place in natural waters except under certain the ratio between daily maxima and minima a function restricted definitions of balanced growth (to be dis- of latitude (Doty 1959) or, more precisely, of day cussed later). length (Lorenzen 1963). Soeder (1965) concluded Shuter (1979) pointed out that balanced growth that microalgae tend toward synchrony in their is found only in totally asynchronous growing popu- activities under light–dark cycles whenever expo- lations. Single cells do not display balanced growth nential growth is possible. Sournia (1974) reviewed "since the synthesis of individual components occurs the literature on such periodicity in several phyto- at discrete intervals over the cell cycle." Balanced plankton processes. We can expect discontinuous growth can only be a property of a population of or periodic assimilation whenever illumination is cells growing asynchronously. Asynchronous popu- discontinuous or periodic. Diel changes in dissolved lation growth is to be expected only under strictly carbohydrates in seawater have been observed in constant and uniform supply rates of all materials both coastal (Walsh 1965; Sellner 1980) and open and energy sources required for growth. Algal cul- ocean locations (Burney et al. 1979). Diel changes tures growing exponentially with excess nutrients in oxygen (Tijssen 1979) and particulate organic (batch cultures) or with constant nutrient supply rate carbon (Postma and Rommets 1979) also imply the (continuous cultures) and with constant irradiance 24-h cycle is important in phytoplankton dynamics. and temperature approach the ideal of balanced Nutrient substances in the euphotic zone of the growth. oceans are also subject to temporal and spatial vari- Neither cyclostats (Chisholm and Nobbs 1976), ation in ambient concentrations and rate of supply. i.e. continuous cultures growing under periodic The source of the supply may be mixing with, or illumination, nor natural systems with periodicity replacement by deeper ocean water, lateral advection or irregularity in light, temperature, or nutrient supply of surface water from elsewhere, or from rivers or can be expected to show balanced growth; rather, the atmosphere. Coastal upwelling generates spec- periodicity in growth is imposed by the environ- tacular differences in nutrient supply rate to the

251 euphotic zone episodically. Winter–summer dif- are required if specific nutrient assimilation rate ferences in nutrient concentration, related to seasonal , bc \ ju l ., etc.) is to be equivalent to ic for cell differences in mixing/stratification, may also be numbers and, therefore, the null hypothesis rejected. extreme, especially in the coastal regions of temperate The cell cycle may usually be assumed to be linked and boreal waters of the North Atlantic and North to the 24-h illumination cycle, however. In that Pacific. case average assimilation rates over 24 h may be The kinds of variation in nutrient supply most eqttivalent to g if the chemical composition and important for the estimation of growth rate from cell size remain constant between days. rates of nutrient assimilation, however, are those Certain cell constituents vary greatly in mag- which take place within the generation time of the nitude over the 24-h day–night cycle. For example, phytoplankton. Variation in nutrient supply on a time the water-soluble carbohydrate content of the par- scale of hours, where there is assurance the same ticulate matter in Mikawa Bay varied almost an order parcel of water was being sampled, is not well docu- of magnitude whereas chlorophyll remained essen- mented. tially constant (Handa 1975). This carbohydrate was Ryther et al. (1961) and Beers and Kelley (1965) produced in the light and respired by the phytoplankton measured day–night differences in the ambient con- at night. Its specific rate of increase in the light would centrations of ammonium and other nutrients in the exceed lc, severalfold. ATP levels in a Cm/du/nit/ma Sargasso Sea. Müller-Haeckel (1965) found a daily culture showed similar light–dark variations, with the periodicity in silicic acid in a saline creek. The daytime ATP increase an order of magnitude greater excretion of nutrients into the euphotic zone waters than for cell division (Weiler and Karl 1979). The by zooplankton and fish would be expected to be cellular nitrate content of phytoplankton in shipboard greater by night than day if a significant fraction of cultures also showed large day/night differences the excretion was due to vertically migrating ani- (Collos and Slawyk 1976). mals. Discontinuous supply may be a much more Microscopic observations have revealed phased common and important phenomenon than we have cell division (i.e. cell division restricted to a portion considered in the past (Goldman et al. 1979; McCar- of the 24-h day) in a number of phytoplankters in thy and Goldman 1979). natural samples. The large-celled dinoflagellates are Considering the spatial scale, Oviatt et al. (1972) best known in this respect. Recent observations found marked increases in the ammonium concen- include Pyrocystis spp. (Swift and Durbin 1972; Swift tration in the wake of a menhaden school. Calcu- et al. 1976) and Ceratittm spp. in coastal waters (El- lations suggest the ammonium concentration is brüchter 1973; Weiler and Chisholm 1976) and in probably doubled over ambient levels in southern the central North Pacific (Weiler 1980). Phased cell California waters in the wake of an anchovy school division was also noted in the large-celled diatom (P. E. Smith and R. W. Eppley unpublished data). Dityluin biiglitiielIli (Smayda 1975). Such patches of elevated nutrient content would be The diel periodicity in photosynthesis of phyto- expected to have a lifetime of minutes to hours, plankton in natural samples has been examined ex- depending on ambient phytoplankton concentrations tensively. Malone (1971) noted differences in nano- and assuming dispersal by eddy diffusion with a plankton versus netplankton in this respect in samples typical horizontal" mixing length" of about 1 km d '. of oligotrophic surface water. The nanoplankton Nutrient patches brought about by schools of showed a greater productivity index (milligrams C per fish and individual large animals, such as sharks, milligram chlorophyll a per hour) in the morning than seals, and whales, are likely important to phyto- the afternoon whereas afternoon values for netplank- plankton in low nutrient waters (J. J. McCarthy ton were the same or higher than morning values. personal communication). McCarthy and Goldman Morning and afternoon values were similar for both (1979) proposed that the nutrient patelles in the wake size fractions in the more eutrophie California Current of individual macrozooplankton may also be sig- waters. Paerl and MacKenzie (1979) made similar nificant for phytoplankton nutrient uptake. Very short observations in freshwater. Statistical analysis of lifetimes are calculated for patelles of such small many photosynthetic light curves indicated the diel size, however (Jackson 1980). periodicity was in the maximum photosynthetic rate Discontinuous assiutilatiou and cell division — (P,;;„,); cliel variation in the initial slope (a) of the In what follows I will argue for the validity of the curves was not statistically significant although a hypothesis that measurement of nutrient assimilation and P,;,'„, were correlated (MacCaull and Platt 1976). rate per unit biomass does not provide an accurate Truly endogenous circadian rhythms are well- measure of phytoplankton specific growth rate (it). known for certain marine phytoplankters. What The evidence available suggests that periodicity in remains unclear is the extent of their modification elemental uptake and assimilation and in cell division by environmental factors such as ambient nutrient is the rule. Thus, measurements over the cell cycle levels (Stross et al. 1973).

252 Diel periodicity in nutrient uptake by natural of the phytoplankton). Departures from balanced populations of phytoplankton is also well-known. growth during the experimental incubations may Goering et al. (1964) found diel variations in am- contribute appreciably to these discrepancies (G. A. monium and nitrate uptake in Sargasso Sea samples Jackson and R. W. Eppley unpublished data). with maximum rates in the early morning. In richer waters off Peru the peak was near noon; the obser- CULTURE STUDIES: EXAMPLES OF vations there were extended also to silicic acid uptake UNBALANCED GROWTH (Goering et al. 1973). In 200-L shipboard cultures Experiments with laboratory cultures of phyto- off California diel periodicity was noted for nitrate, plankton provide many examples of discontinuous ammonium, urea, and phosphate uptake. Peak rates or phased assimilation of nutrient elements, including were near noon as for phytosynthesis. Diatoms in photosynthetic carbon assimilation. These further divided in the afternoon and early evening the cultures examples of unbalanced growth lend credence to the (Eppley et al. 1971a). null hypothesis that rate measurements of nutrient Although photosynthesis is by definition re- elements will not usually provide accurate estimates stricteci to daytime, several nutrient substances are also of specific growth rate. Cellular periodicity and taken up, often at reduced rate, in darkness. Silicic either nutrient deprivation or discontinuous supply acid uptake and incorporation in natural samples is not result in most of these observations. completely light dependent (Azam and Chisholm 1976; Goering et al. 1973). Neither is ammonium Cellular periodicity — It used to be common uptake, as noted above. MacIsaac and Dugdale (1972) practice to maintain algal cultures in the north win- prepared uptake rate versus irradiance curves indi- dow. Here they were exposed to diffuse daylight cating low half-saturating irradiances for nitrate up- and its natural light–dark cycle. The algae in such take and even lower ones for ammonium uptake. Dark cultures often divided at night (Braarud 1945). In uptake rates of nitrate were much lower than those for the late 1940s and early 50s, serious consideration ammonium in Kaneohe Bay, Hawaii (Harvey and was given to the mass culture of algae, greatly Caperon 1976). Phosphate uptake took place at the stimulating activity and research on the physiology same rate day and night in the oligotrophic central of algae. The notion of synchronously dividing algal North Pacific (Perry 1976) yet showed die periodicity populations has its roots in those times (see the in coastal waters (Harrison et al. 1977). Reshkin and review of Tamiya 1966). In some cases it could be Knauer (1979) provided phosphate uptake rate versus shown that the rhythmic cell division was endogenous irradiance curves for natural samples of coastal plank- and would persist, once established during growth ton. on light–dark cycles, in continuous dim light (Sweeney and Hastings 1958). Falkowski and Stone (1975) made the interesting Recently Nelson and Brand (1979) examined observation that adding 3 /LA/ nitrate or ammonium the periodicity in cell division in 26 clones (13 to coastal British Columbia water samples resulted species) of marine phytoplankton. Phased cell divi- in a lowered maximum rate of photosynthesis (P1, ) sion resulting simply from growth in light–dark that persisted several hours. Presumably the additions cycles was found in 22 of the 26 strains. Cell division brought about a temporary change in the C/N assimi- began in the light period in 10 clones of diatoms. lation ratio of the phytoplankton (although N-assimi- Other taxa showed maximum cell division at night. lation rates were not measured). These also showed slight decreases in cell number These and many other observations indicate in the light period, a phenomenon not seen in the that uptake of various nutrient substances may be diatoms. Such spontaneous cell lysis has not been either in phase or somewhat out of phase with one systematically observed in previous culture studies. another and with photosynthetic carbon assimilation. The lysis resulted in negative rate constants Cell division activity may peak at a totally different for growth based on cell counts about –0.01 h ' for time, such as at dawn in Cercuitun spp. (cf. Weiler periods of 4-8 h. Paasche (1968) showed that some and Chisholm 1976), an unlikely time for peak diatoms display maximum growth rates of photosynthesis and nutrient assimilation. rates on light- dark cycles. For example, Dityluni brightwellii grew These observations draw attention to the balanced faster with 16-h photoperiods than in continuous growth concept and to the idea that only average light. The significance and role of biological clocks values over the 24-h day could possibly give rates in phytoplankton growth are reviewed by Chisholm related to the rate of cell division and net growth. (1981). It is particularly worrisome to those making uptake rate measurements of several nutrients in Cyclostais — The cyclostat is a continuous the same samples that the assimilation ratios, e.g. culture operated with a light–dark illumination cycle C uptake/N uptake, are not the same as the compo- (Chisholm and Nobbs 1976). Cyclostat cultures sition ratios of the particulate matter (and presumably typically show diel periodicity in cell division, photo-

253 synthetic rate, nutrient uptake rate (at high dilution indicate rapid growth, yet N assimilation would be rates approaching maximum growth rates), and in zero. Under this transient condition nutrient assimi- chemical composition. Thus, growth in cyclostats lation measurements cannot provide general measures begins to approach the complexity we assume for of growth, which is rapidly approaching zero by the phytoplankton of the ocean. The cultures are any definition, and rate constants may even take homogeneous, however, and display steady states negative values for cell constituents such as chlo- when rates and concentrations are averaged over rophyll. Some earlier studies were reviewed with the duration of the light–dark illumination cycle, respect to growth models a decade ago (Eppley usually 24 h. Examples of cyclostat results with and Strickland 1968, p. 44-45). marine phytoplankton can be found in Davis et al. Phytoplankton respond to environmental change (1973), Eppley et al. (197 lb), Eppley and Renger by adjusting their chemical composition. " . . phy- (1974), Laws and Wong (1978), Laws and Bannister toplankton have mechanisms for regulating uptake (1980), and in Malone et al. (1975). of each element and . . . these mechanisms are used These studies have made comparisons based to maintain composition and achieve balanced on samples talcen at a given time each day or on rates growth" (Laws and Bannister 1980). The direction averaged over 12 or 24 h. To date, nutrient uptake of these adjustments tends to maximize growth rate rates for several elements have not been examined under the new conditions (Shuter 1979). Shuter sum- in detail as a function of time in the cell cycle. At marized the responses to increases in light, nutrient low growth rates compared to ,u„„ the growth-limiting supply, and temperature and reviewed published nutrient (N) remained undetectable in the medium examples showing changes in chlorophyll, RNA, over the light–dark cycle (Malone et al. 1975; Eppley carbon reserves, and growth rate. Droop's model et al. 197 lb) and its specific uptalce rate varied with (Droop 1970) clarifies these changes in cellular com- cell number and N/cell. (In continuous light the position as functions of nutrient supply rate (McCarthy limiting nutrient in chemostats, either N or P, re- 1981). When nutrients are suddenly added to depleted mained undetectable in the medium for < 0.85 ,u„, cells the response is often a rapid uptake of the (Goldman et al. 1979)). Assimilation of C and N depleted nutrient. For example, when nitrogen was were clearly out of phase in the cyclostat cultures added to N-limited diatom cultures the nitrogen was as N uptake was continuous while C assimilation taken up rapidly. Carbon assimilation was depressed took place only in the light. (as Falkowski and Stone (1975) found with natural Most physiological processes of phytoplankton samples) such that the ratio of assimilation rate, become entrained to some degree during growth on V,. /l/N , was as low as 1.0 by atoms (Collos and light–dark cycles. This has provided an experimental Slawyk 1979). In balanced growth the assimilation tool for probing many aspects of the timing of events ratio would be the same as the composition ratio: in the cell cycle and of cellular biochemistry well five or more. beyond the scope of this review (see Sournia 1974; Conway et al. (1976) and Conway and Harrison Mitchison 1973). Studies of rhythmic phenomena (1977) measured specific uptake rates of N, P, and in unicellular algae began in the laboratory. Studies Si at the same time using three diatom species growing relative to understanding primary production and in N-limited chemostat cultures. Silicic acid limited phytoplankton growth in the sea are more limited cultures were also studied as were cultures starved in history and scope. The diel periodicity in photo- of nutrient for 72 h, and nutrient-sufficient batch synthesis, nutrient assimilation, and cell division cultures. Nutrient uptake rates were calculated from in natural samples discussed earlier has also been the decline in external concentrations measured observed in cultures growing on light–dark cycles, chemically. The cultures were kept in continuous as have several more subtle phenomena of interest light and constant temperature. Departures from to planktologists. Respiration is one of these; it varied balanced growth in the exponentially growing, over the cell cycle in Chlorella (Sorokin and Myers N-limited chemostat cultures reflected responses to 1957). pulse, 7-8 p,M additions of ammonium. Not only Nutrient deprivation or discontinuous supply — were subsequent uptake rates of the ammonium-N A striking departure from balanced growth is provided higher than the steady-state growth rate, but phos- when a batch culture approaches the stationary phase phate and silicic acid uptake rates were depressed because of the depletion of a necessary nutrient. It after the ammonium addition (Table 1). is often observed that two or more cell divisions This result is also reminiscent of the Falkowski and considerable photosynthetic carbon assimilation and Stone (1975) observation and takes us back to may continue following the depletion of nitrogen earlier observations where adding ammonium (Syrett from the medium. The N content of each cell then 1958) or silicic acid (Coombs et al. 1967) to N- or declines, being "diluted" by the subsequent cell Si-depleted cells, respectively, resulted in a transient division. Carbon assimilation measurement would but spectacular decline in cell ATP content. Holm- 254

TABLE 1. Departures from balanced growth due to addition of 7 44 ammonium to an ammonium-limited chemostat culture of Skeletonema costatum. Flow was stopped upon the NF11- addition. The dilution rate had been 0.040 18°C, irradiance 0.14 ly • min', continuous. (From Conway et al. 1976.) Uptake rates are per weight of element in the algae.

Ammonium uptake Phosphate uptake Silicic acid uptake Surge Duration Steady Lag Subsequent During Subsequent rate of surge rate rate lag rate -1 ) Expt. (h-1) (h) (h- ') (h) (11- ') (11-1 ) (h

BMP-1 0.218 1.8 0.098 0.6 0.059 0.004 0.087

2 0.235 1.2 0.103 0.6 0.058 0.020 0.067

3 0.188 0.9 0.130 0.6 0.075 0.020 0.110

Hansen (1970) and Sakshaug and Holm-Hansen that ambient nutrient concentration at the surface, (1977) supply many dramatic examples of changing the depths at which measurable concentrations are cellular composition when nutrients are resupplied first encountered, and the nutrient flux into the to depleted cells. These results together suggest that euphotic zone provide similar (qualitative) measures if there is small-scale spatial and temporal patchiness of primary production and the nutritional status of in nutrients in the ocean, and if the phytoplankton the phytoplankton. Sverdrup (1955) produced a rough are at all nutrient depleted, then the cells may indeed map of global production based on general knowledge channel their available energy resources into the of such physical processes as vertical convection, uptake and assimilation of the depleted nutrient. upwelling, and turbulent diffusion which bring nu- The time scale for recovery from depletion, as as- trients to the surface waters. Modern maps of primary sessed by the time required for uptake rates to recover, production based on ''C measurements are very was several hours (Conway and Harrison 1977). similar to Sverdrup's map (Koblentz-Mishke et al. Whatever the time required it is clear that balanced 1970). The provinces with the most depauperate growth was not taking place and th4t specific uptake euphotic zones are the anticyclonic central gyres rates would not be accurate measures of growth in of the North and South Atlantic and Pacific oceans. that period. Antarctic waters, temperate and boreal coastal waters Myers (1951) concluded that cells suddenly in winter, and offshore waters to the north and south exposed to high irradiance, saturating for photosyn- of about 40°N and 40°S, respectively, have meas- thesis, show an abnormally high C/N assimilation urable surface nutrients the year around. Nutrient- ratio. On return to lower irradiances, below saturation limited growth of phytoplankton seems unlikely for photosynthesis, the rate of N assimilation will in those locations. Temporal and spatial variations be unusually high. Thus, changes in irradiance can in nutrient concentrations there may not be important mimic N limitation and pulsed N additions in their for balanced growth if concentrations are sufficient effects on C/N assimilation ratios. to saturate the nutrient uptake mechanisms. The above lead us to inquire: what is the nutri- Nitrogen is the nutrient first depleted by phyto- tional status of phytoplankton in the ocean? Is it plankton growth in most locations examined (Ryther apt to be uniform over the 24-h illumination cycle and Dunstan 1971). Silicic acid may be depleted such that balanced growth can be expected as in before nitrate in the eastern South Pacific, however cyclostat cultures, or is it forever fluctuating on a time (Zentara and Kamykowski 1977), and phosphate scale of minutes to hours such that balanced growth may be the chief limiting nutrient in the Mediterranean can rarely be expected even over the 24-h cycle? (D. Bonin and S. Maestrini personal communication). Various methods are in use for assessing the NUTRITIONAL STATUS OF PHYTOPLANKTON physiological state of the phytoplankton with respect IN THE OCEAN to nutrients. Compositional ratios of the elements Geographical differences between biotic prov- in the seston are a particularly powerful tool, espe-

inces — The major biotic provinces of the oceans cially in the present context of assessing balanced are determined by the ocean boundaries and major growth. Ratios of C/P have been determined, for current systems. For the most part, we can assume example, in the central gyre of the North Pacific 255 (Perry 1976; Perry and Eppley 1980) and in lake 0: PO4 limited waters (Peterson et al. 1974). The C/N ratio provides 0 N114 limited a relative measure of growth rate (Donaghay et al. CI NO3 limited 1978) and, thus, may provide a guide to growth A light limited; reduction due to nutrient lack. The ratio of protein/ NO3, PO4 saturated carbohydrate in the seston is also a measure of physio- logical state of diatom populations (Myklestad and 0 Haug 1972; Sakshaug and Mykelstad 1973). The o acid-soluble fraction of carbohydrate containing the — /3-1, 3-glucan is the more likely carbohydrate fraction T to be associated with living cells (Myklestad 1977). . 5 These observations from cultures and Trondheims- ' al fjord have been extended to the cape upwelling area of South Africa where the protein/carbohydrate ratio of cn a large crop of diatoms fell to <1.0 as nutrients were • 4 depleted (Barlow 1980). Most indicators of physiological state in phyto- Lu 3 plankton suggest only marginal deprivation of N or o P even in the most oligotrophic ocean waters. For example, the productivity index (PI) was examined for its potential value as such an indicator by Curl t= 2 and Small (1965); low values, 1-3h - ' were indicative of regions of very low nutrient supply. Thomas (1970) found values about 5 where nitrate was measurable 01 and about 3 where it was absent in the eastem tropical O Pacific. Values of 3-5 were found in regions "border- ing on nutrient sufficiency" (Curl and Small 1965). The effect of added ammonium on dark ''C uptake also varies with the nitrogen status of the 0.5 1.0 phytoplankton (Morris et al. 1971); Florida Strait DILUTION RATE d -1 ) plankton were N sufficient by this assay. Nitrogen FIG. 1. Photosynthetic carbon assimilation rate per weight deficiency was indicated, however, in Gulf of Maine of chlorophyll a in Thalassiosira ihiviatilis. Data (from phytoplankton at the end of the spring bloom Laws and Bannister 1980) are for nutrient- and light- (Yentsch et al. 1977). limited continuous cultures grown on a 12: 12 light-dark PI values take on added significance as indicators cycle. Curves were calculated from their growth model. of nutritional status since the report of Laws and Bannister (1980). They developed a mathematical quantities of ['5 N]ammonium must be added at con- relation for PI as a function of nutrient-limited growth centrations equal to or exceeding ambient levels in rate common to phosphate, nitrate, or ammonium oligotrophic waters. The measured N-uptake rates limitation (Fig. 1). The PI values are insensitive to would then be high and not representative of previous ,t,t, at growth rates greater than about 0.25 d'. The N-uptake rates before 15 N addition (McCarthy 1981). PI ranged from >1 to —5 for it, values 0-0.25 ' in Considerable agonizing over this problem resulted the model and experimental data, from a cyclostat in tentative corrections for central North Pacific culture of the diatom Thalassiosira fhiviatilis. In samples (Eppley et al. 1973, 1977). Even with the reviewing the literature values of PI they concluded corrections, however, specific assimilation rates that phytoplankton in oligotrophic central ocean calculated for N exceeded those of C (Sharp et al. waters must be growing at rates <0.25 d - . Such 1980). PI values for those samples were only a little reduced rates must reflect nutrient limitation of greater than 1.0, on average, but varied directly growth. If so, nutrient assimilation ratios (C/N, with the N/C ratio of the seston (Sharp et al. 1980). N/P) would be affected and would be expected to Glibert and Goldman (1980) examined short- depart from the Redfield ratio of 106C: 16N: 1P term (5 min-2 h) [ 15 N]ammonium uptake of Vinyard (Redfield et al. 1963). Addition of nutrients to such Sound, MA, plankton. Specific uptake rates were waters would be expected to cause transient changes low, compared with expected values of p. for coastal in C/N and N/P assimilation ratios and cellular plankton. Perhaps this is due to unusually high levels composition ratios as described earlier. of detrital-N. Nevertheless, the rates declined over Unfortunately, in the ' 5 N stable isotope tracer time, when saturating concentrations of ammonium methodology for measuring ammonium uptake, were added, as if the plankton in situ were N limited.

256 Temporal changes in nutrient status — McCar- day, as if the cells became relatively P depleted thy and Goldman (1979) have proposed that micro- over the light period and recovered at night. scale patchiness of nutrients may be important for The potential importance of temporal variation phytoplankton growth in nutrient-depleted regions in nutrient availability is brought forth in a report of the oceans. Such patches could result from indi- by Turpin and Harrison (1979). They operated am- vidual zooplankton and from organic particles colo- monium-limited chemostats which differed only in nized by bacteria. They calculated, for example, the time variation of ammonium input. One culture that small oceanic copepods may release enough had a constant ammonium input, one had eight pulse ammonium in 5 s to bring the ammonium concen- additions per day, and a third culture had one am- tration up to 5 i..tM in a volume of water equal to monium addition per day. The total daily additions the displacement volume of the animal. Their own were the same, only the frequency of addition dif- N-limited chemostat culture data, and other data cited fered for the three cultures. Uptake rates of am- earlier, show quite clearly that the phytoplankton, monium in the three cultures were quite different; if sufficiently N depleted, have the capacity to take rates varied directly with the time between ammonium up ammonium as much as 30 times faster than the additions. We can infer that uptake of C and P were steady-state growth rate when the cells are exposed uncoupled from N uptake in these experiments briefly (5 min) to saturating ammonium concen- although 24-h averages would reflect balanced trations (i.e. >1 itA4). It is easy to calculate that growth. about twenty 5-min exposures to saturating ammo- The coupling between carbon, nitrogen, and nium concentrations over 24 h would support a phosphate assimilation by phytoplankton may be growth rate of 0.5 doublings McCarthy and altered, leading to departure from balanced growth, Goldman (1979) indicated that ammonium concentra- in yet another way. The phytoplankton cells, behaving tions may be <0.03 itM in oligotrophic ocean waters, as passive scalers, are carried over several metres undetectable with present chemical methods. The in the vertical direction by the semidiurnal internal postulated microscale patches, perhaps of only a few tide on the continental shelf. Depending on the nanolitres volume, would also be undetectable by phase relationship of the semidiurnal internal tide present methods. to the daily insolation, the phytoplankton may ex- Exposure of nutrient-deficient phytoplankton perience irradiance levels over the course of the day intermittently to discrete patches of ammonium or quite different than if the cells remained fixed at phosphate and the resulting episodes of nutrient one depth in the euphotic zone (Kamykowski 1974). uptake and assimilation would effectively uncouple Differences in saturating irradiance for uptake of the nutrient uptake from photosynthesis and lead to different nutrients, C, N, P, Si, and the rates of unbalanced growth. The frequency of encountering dark uptake would determine the precise departures the patches, as visualized, appears high, relative to from balanced growth in any particular case. Higher the frequency of cell division, however, and the frequency internal waves would cause qualitatively extent of departure from balanced growth might similar problems for balanced growth. I know of no be relatively small. Storage of nutrient in inorganic detailed analysis of either phenomenon nor their form within the cells (DeManche et al. 1979) could likely importance for balanced growth. result in fairly uniform protein synthesis rates over time, depending on the cells' storage capacity for Estimates of Growth Rate the nutrient, the growth rate, and the frequency of of Ocean Phytoplankton nutrient encounters. Jackson (1980) carried out an analysis of the It was easy to review the available estimates effects of diffusion on such microscale patches and of phytoplankton specific growth rate in the ocean concluded the individual nutrient pulses from zoo- in 1972 as only a few were available (Eppley 1972). plankters decayed too quickly to serve as the main The early estimates from oligotrophic waters were nutrient source for phytoplankton. Nevertheless, less than 0.5 doublings per day while estimates for nutrient patches due to large animals and fish schools nutrient-rich waters were higher, about 1.0 doublings will persist longer than the postulated micropatches per day when averaged over the depth of the euphotic and surely must influence nutrient uptake by phyto- zone. Recently Goldman et al. (1979) compiled plankton. additional estimates (Table 2). They also proposed Perry (1972) used the cell-surface alkaline phos- that rates in oligotrophic waters are probably high, phate activity of the plankton of the central North approaching maximum expected rates, i.e. it, —> Pacific as an indication of cellular P status. Activity The brief review of Koblentz-Mishke and Vedernikov developed over time when natural samples were (1976) also indicated high growth rates in oligotrophic incubated. The shortest times for appearance of central ocean waters of 6.6 doublings per day. Values activity, zero in some cases, were found late in the were lower in mesotrophic and eutrophic waters, 2.3

257 TABLE 2. Summary of available phytoplankton growth rate data in doublings/day for natural marine waters compiled by Goldman et al. (1979).

Growth rate Location (doublings / day)" Reference

Sargasso Sea 0.26" } Florida Strait 0.45" Riley et al. (1949) Carolina Coast 0.37" Montauk Pt., LI 0.35" S. California Coast 0.25-0.4" Eppley et al. (1972) S. California Coast 0.7" Eppley et al. (1970) N.W. Atlantic 0.2-1.7 Sutcliffe et al. (1970) N. Pacific 0.2-0.4 Eppley et al. (1973) Northem N. Pacific 0.36-0.89 Saino and Hattori (1977) North Sea 0.67-1.33 Cushing (1971) Sargasso Sea 0.05-0.14 Swift and Durbin (1972) Tyrrhenian Sea 0.07-0.25 Baja, California Coast 0.2-1.4 Smayda (1975) Peru Current 0.67" Strickland et al. (1969) Peru Current 0.73" Beers et al. (1971) S.W. Africa Coast 1.0" Hobson (1971) W. Arabian Sea >1.0" Ryther and Menzel (1965) Narragansett Bay, RI 0.4-1.94 Durbin et al. (1975) Narragansett Bay, RI <0.1-3.8 Smayda (1973) Santa Monica Bay, CA 0.3-0.7 Weiler and Chisholm (1976) Oligotrophic waters 6.6 Mesotrophic waters 2.3 Koblentz-Mishke and Vedernikov (1976) Eutrophic waters 0.14 .

"Doublings/day x 0.693 = specific growth rate pt. "Obtained from Table 2 in Eppley (1972). and 0.14 doublings per day, respectively. Nearly all waters and the Scripps estimates of 0. 1-0.2 doublings these estimates have been based upon "C photosyn- per day (averaged over the upper euphotic zone). thetic rate and the phytoplankton biomass as carbon. Our mean phytoplankton carbon estimate is Actually the only growth rate estimate of Table 2 about 10 mg C• m-", still 20-fold higher than their that appears high to me is the 6.6 divisions per day value corrected for nanoplankton loss. The SIO for oligotrophic, subtropical waters from Koblentz- values are based upon cell counts, POC, ATP, and Mishke and Vedernikov (1976). This estimate was chlorophyll a. The cell counts would be subject to based on photosynthesis rates of many surface samples. nanoplankton losses also. On the other hand, the Those rates ranged from <0.110 10 mg C • m-'• d- ' , possible inclusion of heterotrophic flagellates would with an average value of 1.0 mg C • m -'• d'. They are lead to overestimation of phytoplankton carbon essentially identical in range and average with those (Beers et al. 1975). measured by my colleagues and I in the central gyre Clearly the phytoplankton biomass estimation of the North Pacific (Eppley et al. 1973; Perry and is as difficult as the physiological rate measurements. Eppley 1980; Sharp et al. 1980). The difference in At present both are required for estimating the growth the two data sets is in the estimate of phytoplankton rate of the entire phytoplankton assemblage. biomass. It is always fashionable to criticize existing Koblentz-Mishke and Vedernikov (1976) ack- views, especially if the standing outlook can be nowledge that their carbon biomass estimates, derived termed dogma. For example, "it is now essential from cell counts, are probably low by at least 50-fold, to have unequivocal evidence as to what is being resulting from the loss of nanoplankton in the pre- measured by the various techniques which can be served samples. Their values ranged from 0.001 to employed in marine productivity work" was written

0.05 with a mean value of 0.01 mg C • 111-3 . Increasing by McAllister et al. (1964). Nevertheless, the these 50-fold to correct for nanoplankton would raise evidence is still equivocal in spite of much careful the average value from 0.01 to 0.5 mg C • m-3 . This work in the intervening years. In many cases the would reduce the growth rate estimate to 1.6 dou- problems seem not to lie so much in the quality of blings d- ' , a rate still higher than their eutrophic the experimental data as in the conceptions and inter- 258 pretations of phytoplankton growth in the sea which Challenges for the Future guide the measurements. Originally, it was my intent to review critically The paradigm that substantially complete food the available information on growth-rate measure- webs may operate within the volume of seawater ment. I have been unable to do this constructively in a productivity bottle, due to overlapping size cate- and suggest instead some criteria for experimental gories of the planktonic primary producers, her- work and some comparisons that need to be made bivores, and decomposers, stresses the need for new before drawing generalizations. approaches in planktonic rate measurements (Sieburth 1) Published growth rates include values either 1977; Peterson 1980). Phytoplankton growth rate for near-surface samples from a single depth, or is best measured, and incubations in bottles avoided, alternatively, rates averaged over the depth of the by assessing the cell division rate in situ. To date euphotic zone. The former are usually much greater this has been successful only with a few large-celled than the latter but the distinction is not always taxa. Primary production (P) would best be estimated made clear. as the sum of the product of the growth rate and 2) Balanced growth and the 24-h light—dark biomass carbon (B) of each taxon of photoautotrophs cycle should be considered. Although it may not be feasible to measure rates over 24 h, nevertheless, P (14 B i) both daytime and nighttime rates must be considered I along with the expected periodicity in rate over the light hours of the day. Rates based upon brief incu- Such an ideal scheme would also provide the popu- bations near midday can be expected to give very lation dynamics of each taxon, information ideally high values, not representative of balanced growth suited for examination of species succession and its or of the actual growth of the phytoplankton assem- causes. blage in the sea. Given a less perfect ability than the ideal above, phytoplankton dynamicists must use other approaches 3) Calculation of growth rate from nutrient in estimating growth rates under all but the most assimilation rate also requires assessment of the ele- favorable circumstances. We need not remain content mental content of that nutrient in the phytoplankton. with carrying out incubations over finite time periods The phytoplankton biomass is just as important as in containers of convenient volume, but we seem the uptake rate for the growth rate calculation and to be stuck with them for most should be given equal attention. routine work. If C, N, P, Si uptake and assimilation rate measurements 4) Some existing data sets are not internally prove unsuitable for growth rate estimation, even consistent. Growth rates reported in one part of a when average values over the 24-h light—dark cycle publication may be inconsistent with, for example, are used along with acceptable biomass measure- productivity indices reported in another section. ments, then the only recourses for estimating p. for Internal consistency and corroborating evidence lead oceanic phytoplankton are (1) biochemical meas- to reader confidence. urements which determine rates of processes asso- 5) Geographical variation in the productivity ciated with cell division (DNA, RNA, protein syn- index can take place along gradients which so far thesis), (2) analysis of the cell cycle of individual defy experimental definition in terms of ambient taxa or size categories, or (3) to find compositional nutrient concentration, chlorophyll content, species ratios that reflect absolute rather than relative values composition, etc. For example, Bienfang and Gunder- of p.. son (1977) found maximum PI values in the euphotic The present state of the art is to make the rate zone of about 3 mg C (mg ChIc) 11 -1 25 km off measurements with several methods at once, for Oahu in the central North Pacific, but only 1.3 and example, with "C, ' 5 N, and 33 P isotopes, and to 1.1 at stations about 100 and 200 km offshore, evaluate temporal changes over the 24-h cycle. respectively. Given equal carbon biomass, these PI State of the art biomass estimation includes, values imply differences in p. at the three stations, along with chemical measurements, microscopic which in other characteristics were nearly identical. measurement of cell dimensions and examination of Relatively small differences in growth rate could fresh samples with the fluorescence microscope in bring about such differences (Laws and Bannister order to recognize and differentiate between cells 1980). Thus, it may be superficial to draw generalities with chlorophyll (presumed to be photoautotrophs) for large portions of the oceans. It seems equally and those without (presumed to be heterotrophs). unwise to insist on growth rate values to be expected At the same time, the physiological state of the in the various biotic provinces until their horizontal, phytoplankton assemblage can be judged as an indi- depth, and temporal variation is defined more fully cation of whether growth rates should be high or in terms of phytoplankton dynamics. low relative to the expected pi,„, .

259 Greater understanding is needed also of the BEERS, J. R., F. M. H. REID, AND G. L. STEWART. dynamics of water motions that bring nutrients into 1975. Microplankton of the North Pacific Central the euphotic zone, at least for those times and places Gyre Population structure and abundance, June 1973. Int. Rev. Ges'amten Hydrobiol. 60: 607-638. where nutrient input rate is a strong regulator of BIENFANG, P., AND K. GUNDERSEN. 1977. Light effects phytoplankton production. The CUEA upwelling on nutrient-limited, oceanic primary production. Mar. program was successful in combining physical and Biol. 43: 187-199. biological information in the study of upwelling BRAARUD, T. 1945. Experimental studies on marine regions. This merger of interest has not yet developed plankton diatoms. I. Mat. Naturv. Klasse No, 10. for studies of the most oligotrophic oceanic regions, 15 p. Avh. norske VidenskAkad, Oslo. however. BURNEY, C. M., K. M. JOHNSON, D. M. LAVOIE, Chemical information can also be helpful for AND J. M. SIEBURTH. 1979. Dissolved carbohydrate probing phytoplankton dynamics. Measurement of and microbial ATP in the North Atlantic: concentration ambient nutrient levels and cell constituents is already and interactions. Deep-Sea Res. 26: 1267-1290. S. W. 1981. Temporal patterns of cell division periodicity in dissolved CHISHOLM, a necessary aspect. Diel in unicellular algae. Can. Bull. Fish. Aquat. Sci. 210: materials released or consumed by organisms can 150-181. provide boundaries on daytime production and night- CHISHOLM, S. W., AND P. A. NOBBS. 1976. Light/dark- time respiration. Particle interceptor traps (PITS) phased cell division in Englena gracilis (Z) (Eugleno- can provide, at least in the ideal, the loss rates of phyceae) on PO 1 -limited continuous culture. J. particulate materials to depth. The downward loss Phycol. 11: 367-373. rate of particulate carbon and nitrogen from the CoLLos, Y., AND G. SLAWYK. 1976. Significance of populations of marine euphotic zone provides a lower limit of primary cellular nitrate content in natural phytoplankton growing in shipboard cultures. Mar. production in the overlying euphotic zone. What Biol. 34: 27-32. few flux data exist were reviewed earlier (Eppley 1979. "3C and uptake by marine phyto- 1980). plankton. I. Influence of nitrogen source and concen- Probably the best approach for the interim is tration in laboratory cultures of diatoms. J. Phycol. to pursue multiple lines of investigation that bear 15: 186-190. on the magnitude of plant production, biomass, and CONWAY, H. L., AND P. J. HARRISON. 1977. Marine growth rate and not to focus too intently on any diatoms grown in chemostats under silicate or am- one of the existing methods of estimating these. monium limitation. IV. Transient responses of and We have reason to mistrust all the published growth Chaetoceros debilis, Skeletonema costatum, vida to a single addition of the limiting those Thalassiosira gra rate data for the oceans, except, perhaps, for nutrient. Mar. Biol. 43: 33-43. few data based on microscopic observations of cell CONWAY, H. L., P. J. HARRISON, AND C. O. DAVIS. division. Actually, the data for coastal waters are 1976. Marine diatoms grown in chemostats under fairly harmonious; data are most Conflicting for silicate or ammonium limitation. III. Transient re- the oligotrophic central oceans. sponse of Skeletonema costatum to a single addition of the limiting nutrient. Mar. Biol. 35: 187-199. COOMBS, J., P. J. HALICKI, O. HOLM - HANSEN, AND Acknowledgments B. E. VOLCANI. 1967. Studies on the biochemistry and fine structure of silica shell formation in diatoms. II. Changes in concentration of nucleoside triphos- This paper was prepared during a stay at, and phates in silicon-starvation synchrony of Navicula with the financial support of, the Friday Harbor pelliculosa (Breb.) HiIse. Exp. Cell Res. 47: 315- Laboratories, University of Washington, Friday 328. Harbor, WA. I thank Doris Osborn for typing the CURL, H. JR., AND L. F. SMALL. 1965. Variations in manuscript. photosynthetic assimilation ratios in natural marine phytoplankton communities. Llinnol. Oceanogr. 10 (Suppl.): 67-73. DAVIS, C. O., P. J. HARRISON, AND R. C. DUGDALE. References 1973. Continuous culture of marine diatoms under silicate limitation. I. Synchronized life cycle of Skele- AZAM, F., AND S. W. CHISHOLM. 1976. Silicic acid tonema costatum. J. Phycol. 9: 175-180. uptake and incorporation by natural marine phyto- DEMANCHE, J. M., H. C. CURL JR., D. W. LUNDY, plankton populations. Limnol. Oceanogr. 21: 427- AND P. L. DONAGHAY. 1979. The rapid response 435. of the marine diatom Skeletonema costatum to changes BARLOW, R. G. 1980. The biochemical composition of in external and internal nutrient concentration. Mar. phytoplankton in an upwelling region off South Africa. Biol. 53: 323-333. J. Exp. Mar. Biol. Ecol. 45: 83-93. DONAGHAY, P. L., J. M. DEMANCHE, AND L. F. BEERS, J. R., AND A. C. KELLEY. 1965. Short-term SMALL. 1978. On predicting phytoplankton growth variation of ammonia in the Sargasso Sea off Bermuda. rates from carbon:nitrogen ratios. Limnol. Oceanogr. Deep-Sea Res. 12: 21-25. 23: 359-362.

260 DOTY, M. S. 1959. Phytoplankton photosynthetic perio- HANDA, N. 1975. The diurnal variation of organic con- dicity as a function of latitude. J. Mar. Biol. Assoc. stituents of particulate matter in coastal water, p. India 1: 66-68. 125-132. In S. Moni and G. Yamamoto [ed.] JIBP DOTY , M. S., AND M. OGURI. 1957. Evidence for a photo- synthesis. Univ. Tokyo Press 12. synthetic daily periodicity. Limnol. Oceanogr. 2: 37 HARRISON, W. G., F. AZAM, E. H. RENGER, AND -40. R. W. EPPLEY. 1977. Some experiments on phosphate DROOP, M. R. 1970. Vitamin BI2 and marine ecology. assimilation by coastal marine plankton. Mar. Biol. V. Continuous culture as an approach to nutritional 40: 9-18. kinetics. Helgoland. Wiss. Meeresunters. 20: 629- HARVEY, W. A., AND J. CAPERON. 1976. The rate of 636. utilization of urea, ammonium, and nitrate by natural ELBRACHTER, M. 1973. Population dynamics of Ceratium populations of marine phytoplankton in a eutrophic in coastal waters of Kiel Bay. Oikos (Suppl.) 15: environment. Pac. Sci. 30: 329-340.

43-48. HOLM -HANSEN, 0. 1970. ATP levels in algal cells as EPPLEY, R. W. 1972. Temperature and phytoplankton influenced by environmental conditions. Plant Cell growth in the sea. Fish. Bull. (US) 70: 1063-1085. Physiol. 11: 689-700. 1980. Estimating phytoplankton growth rates in JACKSON, G. A. 1980. Phytoplankton growth and zoo- the central oligotrophic oceans. Brookhaven Symp. plankton grazing in oligotrophic oceans. Nature Biol. 31: 231-242. (London) 284: 439-441. EPPLEY, R. W., A. F. CARLUCCI, O. HOLM-HANSEN, KAMYKOWSKI , D. 1974. Possible interactions between D. KIEFER, E. VENRICK, AND P. M. WILLIAMS. phytoplankton and semi-diurnal internal tides. J. 1971a. Phytoplankton growth and composition in Mar. Res. 34: 499-509.

shipboard cultures supplied with nitrate, ammonium, KOBLENTZ - MISHKE, O. J., AND V. I. VEDERNIKOV. or urea and the nitrogen source. Limnol. Oceanogr. 1976. A tentative comparison of primary production 16: 741-751. and phytoplankton quantities at the ocean surface. Mar. EPPLEY, R. W., W. G. HARRISON, S. W. CHISHOLM, Sci. Commun. 2: 357-374.

AND E. STEWART. 1977. Particulate organic matter KOBLENTZ - MISHKE, O. J., V. V. VOLKOVINSKY, AND in surface waters off Southern California and its relation J. G. KABANOVA. 1970. Plankton primary production to phytoplankton. J. Mar. Res. 35: 671-696. of the world ocean, p. 183-193. In W. Wooster [cd.] EPPLEY, R. W., AND E. H. RENGER. 1974. Nitrogen exploration of the South Pacific. NAS, Scientific assimilation of an oceanic diatom in nitrogen-limited Washington, DC. continuous culture. J. Phycol. 10: 15-23. LAWS, E. A., AND T. T. BANNISTER. 1980. Nutrient- EPPLEY, R. W., E. H. RENGER, E. L. VENRICK, AND and light-limited growth of Thalassiosira fluviatilis M. M. MULLIN. 1973. A study of plankton dynamics in continuous culture, with implications for phyto- and nutrient cycling in the central gyre of the North plankton growth in the ocean. Limnol. Oceanogr. Pacific Ocean. Limnol. Oceanogr. 18: 534-551. 25: 457-473. EPPLEY, R. W., J. N. ROGERS, J. J. MCCARTHY, AND LAWS, E. A., AND D. C. L. WONG. 1978. Studies of A. SOURNIA. 1971b. Light/dark periodicity in ni- carbon and nitrogen metabolism by three marine trogen assimilation of the marine phytoplankters phytoplankton species in nitrate-limited continuous Skeletonema costatutn and Coecolithus huxleyi in N- culture. J. Phycol. 14: 406-416. limited chemostat culture. J. Phycol. 7: 150-154. LORENZEN, C. J. 1963. Diurnal variation in phytosynthetic EPPLEY, R. W., AND J. D. H. STRICKLAND. 1968. activity of natural phytoplankton populations. Limnol. Kinetics of marine phytoplankton growth, p. 23-62. Oceanogr. 8: 56-62. In M. Droop and E. J. Ferguson Wood [ed.] Advances MCALLISTER, C. D., N. SHAH, AND J. D. H. STRICK- I. Academic Press, in microbiology of the sea. Vol. LAND. 1964. Marine phytoplankton photosynthesis as London. a function of light intensity; a comparison of methods. FALKOWSKI, P. G., AND D. P. STONE. 1975. Nitrate J. Fish. Res. Board Canada 21: 159-181. uptake in marine phytoplankton: energy sources and MCCARTHY, J. J. 1981. The kinetics of nutrient utilization. the interaction with carbon fixation. Mar. Biol. 32: Can. Bull. Fish. Aquat. Sci. 210: 211-233. 77-84. MCCARTHY, J. J., AND J. C. GOLDMAN. 1979. Nitro- GLIBERT, P. M., AND J. C. GOLDMAN. 1981. Rapid genous nutrition of marine phytoplankton in nutrient- ammonium uptake by natural marine phytoplankton. depleted waters. Science (Washington, D.C.) 203: Mar. Biol. Lett. 2: 25-31. 670-672. GOERING, J. J., R. C. DUGDALE, AND D. W. MENZEL. MACCAULL, W. A., AND T. PLATT. 1976. Diel variations 1964. Cyclic diurnal variations in the uptake of am- in the photosynthetic parameters of coastal marine monia and nitrate by photosynthetic organisms in the phytoplankton. Limnol. Oceanogr. 22: 723-731. Sargasso Sea. Limnol. Oceanogr. 9: 448-451. MACISAAC, J. J. , AND R. C. DUGDALE. 1972. Interactions GOERING, J. J., D. M. NELSON, AND J. A. CARTER. of light and inorganic nitrogen in controlling nitrate 1973. Silicic acid uptake by natural populations of uptake in the sea. Deep-Sea Res. 19: 209-232. marine phytoplankton. Deep-Sea Res. 20: 777-790. MALONE, T. C. 1971. Diurnal rhythms in netplankton and GOLDMAN, J. C., J. J. MCCARTHY, AND D. G. PEAVEY. nannoplankton assimilation ratios. Mar. Biol. 10: 1979. Growth rate influence on the chemical compo- 285-289. sition of phytoplankton in oceanic waters. Nature MALONE, T. C., C. GARSIDE, K. C. HAINES, AND O. A. (London) 279: 210-215. ROELS. 1975. Nitrate uptake and growth of Chaeto-

261 ceros sp. in large outdoor continuous cultures. Limnol. composition of seawater, p. 26-77. 10 M. N. Hill [ed.]. Oceanogr. 20: 9-19. The sea. Vol. 2. Interscience Publishers, New York, MITCHISON, J. M. 1973. The biology of the cell cycle. NY. Cambridge Univ. Press, London. 312 p. RESHKIN, S. J., AND G. A. KNAUER. 1979. Light stimu- MORRIS, I., C. M. YENTSCH, AND C. S. YENTSCH. lation of phosphate uptake in natural assemblages of 1971. The physiological state with respect to nitrogen phytoplankton. Limnol. Oceanogr. 24: 1121-1124. of phytoplankton from low-nutrient subtropical water RILEY, G. A., H. STOMMEL, AND D. F. BUMPUS. as measured by the effect of ammonium ion on dark 1949. Quantitative ecology of the plankton of the carbon dioxide fixation. Limnol. Oceanogr. 16: 859- western North Atlantic. Bull. Bingham Oceanogr. 868. Coll. 12: 1-169. MULLER - HAECKEL, A. 1965. Tagesperiodik des Silicium- RYTHER, J. H., AND W. M. DUNSTAN. 1971. Nitrogen, gehaltes in einem Fliesswasser. Oikos 16: 232-233. phosphorus and eutrophication in the coastal marine MYKLESTAD, S. 1977. Production of carbohydrates by environment. Science (Washington, D.C.) 171: 1008- marine planktonic diatoms. II. Influence of the N/P 1013. ratio in the growth medium on the assimilation ratio, RYTHER, J. H., D. W. MENZEL, AND R. F. VACCARO. growth rate, and production of cellular and extra- 1961. Diurnal variations in some chemical and bio- cellular carbohydrates by Chaetoceros affinis var. logical properties of the Sargasso Sea. Limnol. Willei (Gran) Hustedt and Skeletenema costatum Oceanogr. 6: 149-153. (Grey.) Cleve. J. Exp. Mar. Biol. Ecol. 29: 161- SAKSHAUG, E., AND O. HOLM - HANSEN. 1977. Chemical 179. composition of Skeletonema costatum (Grey.) Cleve MYKLESTAD, S., AND A. HAUG. 1972. Production of and Pavlova (Monochrysis) iutheri (Droop) Green carbohydrates by the marine diatom Chaetoceros as a function of nitrate-, phosphate-, and iron-limited «flints var. willei (Gran) Hustedt. I. Effect of con- growth. J. Exp. Mar. Biol. Ecol. 29: 1-34. centration of nutrients in the culture medium. J. Exp. SAKSHAUG, E., AND S. MYKLESTAD. 1973. Studies on Mar. Biol. Ecol. 9: 125-136. the phytoplankton ecology of the Trondheimsfjord. MYERS, J. 1951. Physiology of the algae. Annu. Rev. III. Dynamics of phytoplankton blooms in relation to Microbiol. 5: 157-180. environmental factors; bioassay experiments and para- NELSON, D. M., AND L. E. BRAND. 1979. Cell division meters for the physiological state of the populations. periodicity in 13 species of marine phytoplankton on J. Exp. Mar. Biol. Ecol. 11: 157-188. a light:dark cycle. J. Phycol. 15: 67-75 , SELLNER, K. G. 1981. Primary productivity and the flux OVIATT, C. A., A. L. GALL, AND S. W. NIXON. 1972. of dissolved organic matter in several marine environ- Environmental effects of Atlantic menhaden on sur- ments. Mar. Biol. (In press) rounding waters. Chesapeake Sci. 13: 321-323. SHARP, J. H., M. J. PERRY, E. H. RENGER, AND R. W. PAASCHE, E. 1968. Marine plankton algae grown with EPPLEY. 1980. Phytoplankton rate processes in the light:dark cycles. II. Ditylum brightwellii and oligotrophic waters of the central North Pacific Ocean. Nitzschia turgidula. Physiol. Plant. 21: 66-77. J. Plankton Res. 2: 335-353. PAERL, H. W., AND L. A. MACKENZIE. 1979. A Com- SHUTER, B. 1979. A model of physiological adaptation parative study of the diurnal carbon fixation patterns in unicellular algae. J. Theor. Biol. 78: 519-552. of nannoplankton and net plankton. Limnol. Oceanogr. SIEBURTH, J. M. 1977. International Helgoland Sympo- 22: 732-738. sium: Convenor's report on the informal session on PERRY, M. J. 1972. Alkaline phosphatase activity in sub- biomass and productivity of microorganisms in plank- tropical central North Pacific waters using a sensitive tonic ecosystems. Helgolander Wiss. Meeresunters. fluorometric method. Mar. Biol. 15: 113-119. 30: 697-704. 1976. Phosphate utilization by an oceanic diatom SMAYDA, T. J. 1975. Phased cell division in natural popu- in phosphorus-limited chemostat culture and in the lations of the marine diatom Ditylum brightwellii oligotrophic waters of the central North Pacific. and the potential significance of die phytoplankton Limnol. Oceanogr. 21: 88-107. behavior in the sea. Deep-Sea Res. 22: 151-165. PERRY, M. J., AND R. W. EPPLEY. 1980. Phosphate SOEDER, C. J. 1965. Some aspects of phytoplankton growth uptalce by phytoplankton in the central North Pacific and activity, p. 47-59. In C. R. Goldman [ed.] Ocean. Deep-Sea Res. 27: 39-49. Primary productivity in aquatic environments. Mem. PETERSON, B. J. 1980. Aquatic primary productivity and 1st. Idrobiol. 18 (Suppl.). the "C-0O5 method: a history of the productivity SOROKIN, C., AND J. MYERS. 1957. The course of respi- problem. Annu. Rev. Ecol. Syst. 11: 359-385. ration in the life cycle of Chlorella cells. J. Gen. PETERSON, B. J., J. P. BARLOW, AND A. E. SAVAGE. Physiol. 40: 579-592. 1974. The physiological state with respect to phos- SOURNIA, A. 1974. Circadian periodicities in natural phorus of Cayuga Lake phytoplankton. Limnol. populations of marine phytoplankton. Adv. Mar. Oceanogr. 19: 396-408. Biol. 12: 325-389. POSTMA, H., AND J. W. ROMMETS. 1979. Dissolved STROSS, R. G., S. W. CHISHOLM, AND T. A. DOWNING. and particulate organic carbon in the North Equatorial 1973. Cause of daily rhythms in photosynthetic rates Current of the Atlantic Ocean. Nether. J. Sea Res, of phytoplankton. Biol. Bull. 145: 200-209. 13: 85-98. SVERDRUP, H. V. 1955. The place of physical oceanog- REDFIELD, A. C., B. H. KETCHUM, AND F. A. RICH- raphy in oceanographic research. J. Mar. Res. 14: ARDS. 1963. The influence of organisms on the 287-294.

262 SWEENEY, B. M., AND J. W. HASTINGS. 1958. Rhythmic WALsx , G. E. 1965. Studies on dissolved carbohydrate cell division in populations of Gonyaulax polyedra. in Cape Cod waters. II. Diurnal fluctuations in Oyster J. Protozool. 5: 217-224. Pond. Limnol. Oceanogr. 10: 577-582. SWIFT, E., AND E. G. DURBIN. 1972. The phased division WEILER, C. S. 1980. Population structure and in situ and cytological characteristics of Pyrocystis spp. division rates of Ceratiutn in oligotrophic waters of can be used to estimate doubling times of the popu- the North Pacific Central Gyre. Limnol. Oceanogr. lation in the sea. Deep-Sea Res. 19: 189-198. 25: 610-619. SWIFT, E., M. STUART, AND V. MEUNIER. 1976. The WEILER, C. S., AND S. W. CHISHOLM. 1976. Phasexl in situ growth rates of some deep-living oceanic dino- cell division in natural populations of marine dino- flagellates: Pyrocystis fusiformis and Pyrocystis flagellates from shipboard cultures. J. Exp. Mar. Biol. Limnol. Oceanogr. 21: 418-426. noctiluca Ecol. 25: 239-247. SYRETT, P. J. 1958. Respiration rate and intemal adenosine triphosphate concentration in Chlorella. Arch. Bio- WEILER, C. S., AND D. M. KARL. 1979. Diel changes chem. Biophys. 75: 117-124. in phased-dividing cultures of Ceratium furca (Dino- TAMIYA, H. 1966. Synchronous cultures of algae. Annu. phyceae): nucleotide triphosphates, adenylate energy Rev. Plant Physiol. 17: 1-26. charge, cell carbon and patterns of vertical migration. THOMAS, W. H. 1970. On nitrogen deficiency in tropical J. Phycol. 15: 384-391. Pacific oceanic phytoplankton: photosynthetic par- YENTSCH, C. M., C. S. YENTSCH, AND L. R. STRUBE. ameters in poor and rich water. Limnol. Oceanogr. 1977. Variations in ammonium enhancement, an 15: 380-385. indication of nitrogen deficiency in New England TUSSEN, S. B. 1979. Diurnal oxygen rhythm and primary coastal phytoplankton populations. J. Mar. Res. 35: production in the mixed layer of the Atlantic Ocean 539-555. at 20 °N. Nether. J. Sea Res. 13: 79-84. ZENTARA, S.-J., AND D. KAMYKOWSKI. 1977. Latitu- TURPIN, D. H., AND P. J. HARRISON. 1979. Limiting dinal relationships among temperature and selected nutrient patchiness and its role in phytoplankton plant nutrients along the west coast of North and South ecology. J. Exp. Mar. Biol. Ecol. 39: 151-166. America. J. Mar. Res. 35: 321-337.

263 Competition Among Phytoplankton Based on Inorganic Macronutrients

SERGE Y. MAESTRINI AND DANIEL J. BONIN Station marine d'Endowne Chemin de la Batterie des Lions F 13007 Marseille, France

Introduction little effect on the species competition and succession, which is then thought to be controlled by other Theories and tentative models of competition processes, such as selective grazing, whose impor- established to give a scientific basis to the "struggle tance has previously been underestimated. for life" were stimulated first by observations made Trying to appreciate the arguments of both sides on animal species (Lotka 1920, 1934; Volterra 1926; by criticizing the main contributions seems to us Volterra and d' Ancona 1935). Since Gause (1932a, a sufficient goal for this review; an exhaustive listing b), theories have been developed to describe the of published papers is certainly useless here and respective behavior of two or more species inhabiting most of the foremost knowledge on algal nutrition the same niche and competing for the same food. mechanisms is summarized in other chapters by They apply very well to microorganisms sharing R. Dugdale, R. Eppley, and J. McCarthy. the same pool of nutrients. All such models have been based on the assumption that, as a population becomes more dense, the conditions of life become Resource-Based Competition more severe and increase the rate of mortality or Between Algae decrease the rate of growth, or both. The particular an individual inhibits the growth mechanism by which THE PHYSIOLOGICAL BASIS of another of the same or of a different competing species is usually left vague, or interpreted in terms Monod (1942) presented a mathematical model of resources. The rationale is that rates of growth expressing the fact that the growth rate of a micro- decrease because the available food decreases as the organism increases with increasing concentrations population increases. of the nutrient limiting growth, but that there is a Because all algae require the same major nu- limit to this increase. The growth model is described trients (except silicon used in great amounts only by a rectangular hyperbola which is virtually identical by diatoms) which they take up from a common with the one of Michaelis and Menten (1913) for water resource pool, and because Liebig (1840) stated representation of enzyme kinetics. Therefore, the that plants are limited by the one nutrient that is constants used to represent the specific growth and least available relative to the plants requirements uptake features are homologous. The so-called for growth, the models established from data on maximum specific growth rates and the specific animals have implicitly been considered to be valid half-saturation constants have come into common also for plant competition. It is demonstrative, for use in microbiology, including also the study of instance, that all the basic background cited by aquatic microheterotrophs since Parsons and Strick- Hutchinson (1961) in his classic paper includes land (1962). only such references. Thus, algal competition was Efforts to extend the hyperbolic model to phyto- also perceived to operate on the basis that the one plankton were attempted first by Droop (1961) who species best able to obtain and assimilate the actual studied the uptake of vitamin B 12 by the rock pool limiting nutrient should take the upper hand and flagellate Pavlova lutheri (formerly Monochrysis outgrow the other competing species. Thus, by lutheri). The kinetics he first recorded did not follow measuring the relevant uptalce or/and growth para- the model, but this was a result of inadequate technical meters and comparing them to the nutrient contents approach (batch culture), as he demonstrated later of waters, one could hope to predict the species using continuous culturing. On the contrary, Price dominance and its succession in time and space. and Care11 (1964) obtained data from batch culture However, recently a few authors claimed to have with Euglena gracilis that fit a hyperbola, but they evidence that natural oceanic phytoplankton is not did not discuss this feature. Then, Caperon (1967) nutrient limited. According to this new concept, discussed the applicability of the hyperbolic model nutrient competition should be considered as having to the growth of a nitrogen-limited flagellate, /so-

264 chrysis galbana, and demonstrated it fits quite well. specifically distinguish between intracellular storage But emphasis and detailed discussion with respect to issued from luxury uptake and elements that have natural populations were the contributions of Dugdale been incorporated into the biochemically active cell (1967), whose article has become classic. The im- material. That was done through the introduction of portant point of his concept is that the concentration a three-compartment model by Grenney et al. (1973) of a limiting nutrient is set by the magnitude of loss who separated the extracellular pool of nutrients, rate and by the Michaelis—Menten parameters of the the intracellular reservoir of inorganic nutrients, the natural phytoplankton. Thus, the system is considered intermediate organic compounds, and the constitutive as regulating itself through the interaction between molecules of the biomass. algae and nutrients and these interactions are dy- At that time, supporters of the cell quota models namic. Dugdale demonstrated the importance of the ( = two-compartment model) claimed that it is not growth kinetic characteristics in the study of species necessary to consider the biochemical details of the competition and succession. He postulated that cell processes as internal nutrient pools represent species characteristic of nutrient-poor waters are only a few percent of the total cell nutrients, and expected to show low values for maximum uptake challenged supporters of the internal-pool model rate and half-saturation constant, and species of ( = three-compartment model). The contention has nutrient-rich waters opposite values; this was fully produced abundant literature of comments and demonstrated a little later. replies, from which it seems to us that the cell-quota Soon after Dugdale's stimulating theory, other models are fitted quite well to steady-state systems authors demonstrated that the model was obeyed both while the internal-pool models apply better to natural by laboratory-cultured algae (Caperon 1968; Gates dynamic conditions. Hence, for predicting the be- and Marlar 1968; Eppley and Coatsworth 1969) and havior of two or more species competing for the same by natural populations (MacIsaac and Dugdale 1969). nutrients, the former are probably more useful in The models based on a linear effect of nutrient laboratory experiments and the latter in field obser- concentration on growth (Steele 1959; Riley 1963) vations. Yet, one should not forget that the nutrient were definitively displaced. However, further con- reservoir, if any, may have been supplied by the tributions beginning with Caperon (1968) and Droop parent cells of those studied, and the true competition (1968) and reviewed by Dugdale (1977) pointed may have taken place long before the observed out that substitution of growth for nutrient uptake situation. is valid only in the steady state of a continuous culture Thus, the physiological processes of algal limited by organic carbon. That was the case in nutrition now appear to be considerably more com- Monod's (1942) carbon-limited bacterial culture, but plicated than one would have expected two decades it does not generally occur in natural waters; with ago. Yet, Droop (1974) suggested that the whole other limiting nutrients, growth and uptake are un- algal machinery might adapt to external conditions coupled. Price and Care11 (1964) and Price and by shifting from "fast-adapted cells" to "slow-adapted Quigley (1966) were the first to suggest the existence cells," and vice versa. Therefore, the same species of an internal pool of nutrients, namely iron and might show two sets of quite different kinetic con- zinc, in their study on Euglena gracilis , but their stants. The fast-adapted cells, which appear only statement was not related to a general concept of in nutrient-rich medium, have a high subsistence growth and uptake. That was done by Caperon (1968) quota (Droop's k g ); nevertheless, they divide very and Droop (1968) who hypothesized that growth fast. The slow-adapted cells are the common cells; rate is most likely related to the intracellular nutrient their k0 is low, and their division rate also. Further reservoir. In this case, the nutrient history of the researches reported by Harrison et al. (1976) and phytoplankton becomes important. Ability of a cell Conway et al. (1976) confirmed this hypothesis by to store nutrients is a considerable advantage, because demonstrating the diatom Skeletonema costatum a cell with high rate of uptake that spends only a can have two such physiological states. This is not brief time in a nutrient-rich deep water would have unique among microorganisms, however, as an a very high growth potential when returning to lighted homologous mechanism has been already described surface waters (Caperon 1968). Dinoflagellates for bacteria (Schaechter 1968; Bremer and Dennis that migrate below the thermocline during the nights 1975), under the name of "shift-up" and "shift-down" and return to the surface during the days (Dugdale process; other algal species will probably show such and Goering 1967; Holmes et al. 1967) are a fasci- a pattern when studied. nating example of this mechanism. Otherwise, Law and Button (1977) criticized The introduction of a second compartment had the rationale of expressing species-specific constants, already improved the concept of algal growth as a because the K values are dependent on Um „, and function of nutrient processes in natural waters. because their measurements lead to imprecise data However, the two-compartment model does not ( ± 100%). They proposed a new parameter: the

265 "affinity" of an organism to a substrate, which waters should be inhabited mostly by fast-growing pertains to microalgae as well as bacteria (Button species, because, as nutrients are more abundant, 1978). They concluded that nutrients taken up and ability to take them up at low concentrations is not as stored in the internal nutrient pool may leak back important, and natural selection will tend to favor through the cell envelope because of excess accumu- organisms capable of dividing rapidly. Although other lation, but that they are also transported to the cell adaptative processes like predator defense, ability to deficient in macromolecules and assimilated. To remain suspended in the photic layer, and growth summarize, affinity is "either a measure of effective responses to temperature and illumination variations collision between substrate and transport site, or a contribute to the success of a species, Dugdale's measure of the absolute substrate flux at low con- hypothesis was complete in itself and stimulated centration normalized against substrate concentration research for almost a decade. and cell population." Button (1978) stressed the The simplest application of the model, i.e. the importance of affinity in evaluating the competitive comparison of the K and Um „, values of species com- ability of an organism in a nutrient-limited system, peting for the same limiting nutrient, was done soon but available data on this parameter are rare. after Dugdale's paper. Eppley and Coatsworth and Thomas (1969), and Eppley NUTRIENT-BASED ALGAL COMPETITION: (1968), Eppley al. (1969) hypothesized that species involved THEORIES, EXPERIMENTAL STUDIES et sequentially in a seasonal succession of dominant During the past two decades, the nutrition of species related to declining nutrient concentrations individual species of algae has stimulated much should be ordered by lower and lower K values. research in the laboratory, which has led to several Then they measured the kinetic constant values of models, as summarized above. Natural populations 16 cultured species of phytoplankton and made cal- have also been studied and modeled using the same culations by combining the K and U,„„, values for approaches. Thence, it is surprising that little has nitrogen with irradiance and temperature. Results been done to observe the concomitant behaviors of obtained showed, as expected, that the coastal dia- two or more species sharing the same nutrient pool tom, Asterionella japonica, has a higher K value for in the same ecological niche or experimental vessel. nitrate uptake (1.3 bt,g-at • L - ' ) than an isolate of It was also a surprise to us that we could not find a Chaetoceros gracilis from the open sea (0.3 cg-at • review fully and specifically dedicated to nutrient- L- ' ). In general terms, they remarked that algal based algal competition, and that the few articles K (NO„) decreasing values paralleled a succession of dealing with this topic are introduced typically by aquatic natural niches, namely freshwater ponds, such a sentence: "Classical ecological competition estuarines, and marine coastal waters and oceanic concept" without any citation, or by referring to one waters. This obviously reflects an adaptation to of the- models established for single species growth. decreasing nitrogen richness of waters. They sug- That concepts established for other organisms, gested that an alga like the coccolitophorid Cocco- especially bacteria, have been implicitly considered lithus Intxleyi is a good competitor when nitrogen as also valid for algae is obvious. Nothing clear-cut concentration and irradiance are low. On the contrary, was available until Titman (1976), who first asso- laboratory results predicted that the diatoms Ditylunt ciated the words "algae" and "resource-based com- brightivellii and Skeletonetna costatunt would grow petition" for a tentative prediction of a two-species faster than C. huxleyi at high irradiance and nitrogen experimental competition, but several authors had available at a concentration higher than 1.5 iug-at- already included this idea in models established to N • L- '. Hence, in the California coastal waters stud- describe the nutrient limitation of algal growth. ied, C. lualeyi would predominate over diatoms un- Besides such particular contributions, several authors less upwelling increased the nitrate concentration in have referred to nutrient competition as a whole to surface water; then diatoms would fare better. This explain some specific dominance in natural or experi- was actually observed and confirmed later by Estrada mental situations (e.g. Elbrâchter 1977). and Blasco (1979). Carpenter and Guillard (1971) Use of single species nutrient-kinetic constants and Guillard et al. (1973) also compared the kinetic — Dugdale (1967) emphasized the ecological ad- constant values of nitrogen- and silicon-limited vantage given to an alga by the set of physiological growth of estuarine and oceanic strains of the algae adaptations represented by a low half-saturation Bellorochia sp., Fragilaria pinnata , and Thalas- constant (K) and a high maximum growth rate siosira pseudonana . Data obtained demonstrated He discussed the question of whether the species that clones isolated from oceanic low-nutrient area,

with lower ,a Ill a \ values also have lower K that would i.e. Sargasso Sea, have K (NO„) values ranging enable them to compete with species with high spe- from 0.25 cg-at • , while those of the same species cific uptake rate (U,„„,) values, when living in oligo- taken from estuarine or eutrophicated water have K trophic waters. On the other hand, eutrophicated values from 1.6 cg-at • L- ' to 6.8 bt,g-at • L-1.

266 Isolates from continental shelf areas have intermediate corded the kinetic constant values of the cyanophyta, or low K values. These differences between clones Oscillatoria agardhii , and remarked the iu,,„„, he are paralleled by those observed in silica uptake obtained was consistent with the turn over times of kinetics, which are 5 times higher for K and 1.6 carbon in lake waters where this organism is strongly times higher for U„,„,. These authors concluded dominant. Then, by comparing these values with that physiological races of phytoplankton are adapted those of the green alga, Selenastrum capricornutuin , to high- or low-nutrient levels. Another similar sup- he predicted the cyanophyta would out-compete the porting contribution is Qasim et al. (1973), who green alga when nutrient reserve decreases below determined the phosphate and nitrate requirements 0.2 ,u,g-at-N • L- ' and 0.03 itg-at-P • L- '. Brown of the diatom, Biddulphia sinensi s , and the dino- and Harris (1978) and Brown and Button (1979) flagellate, Ceratium furca, and showed that the dia- examined in detail the ability of S. capricornutum tom, which is a large-celled organism, has higher K to compete for orthophosphate when P limits algal values than the dinoflagellate for both elements. When growth (frequently the case in fresh waters), and comparing the seasonal variations in abundance and concluded that it will not be growth-competitive the averaged monthly concentrations of the two ions with some other common aquatic autotrophs such in the Cochin Backwater area, they remarked that as Nostoc sp. They also reported some peculiar at low nutrient concentration C. furca appears to be kinetic features that will be discussed. at an advantage. At slightly higher concentrations Thus, it seems to be well established that specific of the two nutrients, B. sinensis begins to predomi- kinetic constants reflect quite well the algal ability nate. When heavy rainfall and subsequent land runoff to compete for nutrient partition, and that natural have enriched the coastal waters, C. furca is scarce populations are nutrient adapted. Thence, the cor- enough not to be recorded while the diatom blooms. ollary, i.e. the in situ limiting factor of uptake rate Natural populations seem to behave similarly ac- and, hence, of growth, is determined by comparing cording to MacIsaac and Dugdale (1969), who K values and naturally occurring nutrient concen- measured nitrogen uptake in varied seawaters. They trations and/or by comparing actual U and reported that populations from oligotrophic areas (Thomas 1970a, b; Paasche 1973; Smayda 1973; show a K = 0.2 ,tg-at • L- ', in corresponding eutro- MacIsaac et al. 1974, 1979; Perry 1976). However, phic regions K = 1.0 itg-at • L- ', and also K values Carpenter and Guillard (1971) pointed out that con- measured in laboratory cultures correspond to those sideration should be given to the history of the test given for eutrophic regions. strains to verify that kinetic values have not changed Research devoted to the role of carbon is rare during maintenance in culture collection. That leads and usually focuses on freshwater species. Goldman to the question: are the differences observed adapta- et al. (1974) measured the inorganic carbon limited tive or selective? In other words, are the geographic growths of Selenastruin capricornutum and Scene- strains a result of Darwinian selection or are the desmus quadricauda. On the basis of kinetic values algae able to adapt their cell machinery to varied obtained, they predict that, in a mixed culture, S. conditions? Because no physiological measurements quadricauda will eliminate S. capricornutum . An were usually made immediately after isolation, no experiment done in continuous culture proved this experimental data are available to ascertain that prediction correct. After 7 d, the dominant species cultured cells have not changed at all. However, represented 86-93% of the total biomass, when comparisons by Yoder (1979) between the cell di- initial respective biomasses were equivalent. King vision rate of natural populations of the diatom, and Novak (1974), who discussed these data, demon- Slceletonema costatum, grown in dialysis bags, and strated that inorganic carbon would never be rate that predicted from a mathematical model based on limiting for algal growth. However, they suggested laboratory data for cultured strains, supports the va- that a succession of algal types can occur, when lidity of the model and, therefore, indicates that algae species capable of existing on progressively lower remain physiologically unchanged in maintenance concentrations of CO., are present, with an ultimate cultures. Moreover, Maestrini and Kossut (1981) dominance of cyanophyta, apparently the algae demonstrated that the elemental composition, namely (= cyanobacteria) best able to use very low con- C/N, C/proteins, C/P, C/ATP, and C/chlorophyll centrations of inorganic carbon. Such a statement a ratios, is restored promptly to a natural level is also defended by Pruder and Bolton (1979). when cells taken from culture collection are grown However, since they reported that the advantage in situ in oligotrophic waters. given to some species by their capability to directly Comparison of theoretically predicted and take up HCO37 , or by having a functional carbonic observed behavior in phytoplankton communities — anhydrase, will appear only below 2 ,u,M-CO., • L - ' Whereas the preceding authors attempted to apply ( = 0.088 mg CO2 • L- ' ), the so-called advantage laboratory-obtained kinetic constant values to the is certainly rarely acting. Alhgren (1977) also re- algae of natural habitats, others endeavored to use

267 the hyperbolic model to predict the steady-state out- cific kinetic constant values, he established the range come of competition between two or several species of nutfient conditions that allows one or other of the growing in controlled conditions. Generally such competitors to dominate in equilibrium; to do this, experiments were devised to test the validity of the he used the hyperbolic model either with the Monod' s models, and usually did, but frequently the methods (1942) equation (one-compartment model) or the represented an ove-rsimplification, so their conclu- Droop's (1974) model (two-compartment model). sions are not free of criticism. Both models predict that C. meneghiniana is the Ross's (1973) attractive title is in fact a mathe- best competitor when the growth of both species is matical speculation free from considerations of the limited by silicium, whereas A. formosa is dominant physiological patterns of the observed organisms, when both species are phosphate limited. The two a namely diatoms, that makes it useless for our ap- species stably coexist when each is limited by long-term experi- plication. Nevertheless, Ross's model predicts that different nutrient. A series of 76 with high dilution the total biomass of two algae competing in the ments in semicontinuous culture author (Titman = Tilman) to demon- nutrient-balanced medium of a continuous culture rate enables the of is smaller than the sum of the individual biomasses strate that the predictions were valid, even those equation that also explain 75% at equilibrium in separated unialgal cultures. General the simple Monod's of the variance in the relative abundance of the two considerations of Stewart and Levin (1973) and natural silicate-phosphate gradient Taylor and Williams (1975) apply to all micro- species along a lake from which experimental strains were organisms, because they are based on the hyperbolic in the influence kinetics of growth, limited by one nutrient. Taylor isolated. Luxury uptake apparently did not of the competition except and Williams' theoretical work led to several par- the predicted outcome Hence, using ticularly interesting conclusions. They found at boundary ratios and slow flow rates. was probably a bias that that a mixed population growing in a continuous- a fast nutrient replacement flow system could remain in equilibrium if there were offset the role of the luxury uptake and Tilman's application at least as many growth-limiting nutrients as there demonstration is partly negated where were different species. They predicted that aquatic to natural populations is concerned. Nevertheless, have stressed the impor- biota, where a single or a few nutrients are limiting his great contribution is to of the relative proportions, rather than con- growth of microorganisms, will tend to produce tance determining species populations of low species diversity. But they also centrations, of nutrients in this previous contributors did not do. stated that species that are unsuccessful competitors interactions; will be slowly eliminated. Because nitrogen is fre- Moreover, slow algal growth and rapid uptalce quently considered to be the only limiting nutrient of have been observed together, as well as actively waters. This algal growth in seawaters and phosphorus to have the dividing algal cells in nutrient-poor the question of whether analyzed nu- same role in fresh waters, and given the usual com- again raised and how closely plexity of the natural populations of phytoplankton, trients really act as limiting nutrients, are related. The former one is led to question the statements of Taylor and uptake and growth processes et al. 1980) Williams, or the general agreement concerning the problem has been partly solved (Berland recent concepts will be discussed later. The latter nature of the limiting nutrients, or both. Also, one and has generated considerable literature since Caperon should remark that the models cited do not separate and the growth and uptake processes because they essen- (1968) and Droop (1968) proved that uptake the intra- tially refer to bacterial growth: bacteria metabolize growth are uncoupled and suggested that nutrients so rapidly after uptake that separation is cellular nutrient quotas may rule primary algal not essential. Likewise, this shortcut is the main growth. criticism of Tilman's contribution, the first fully Based on such a concept, the model of Grenney dedicated to algae with consideration to physiological et al. (1973) considers an algal cell as a three-compart- features and the consequences for the species com- ment unit: the reservoir of inorganic nutrients , organic petition. Titman (1976) and Tilman (1977) predicted intermediates, and proteins. Variations in the uptake the outcome of the competition of the freshwater rate with environmental concentrations of the limiting diatoms, Asterionella fonnosa and Cyclotella merle- nutrient (nitrogen in Grenney et al.) and assimilation ghiniana , competing for phosphate and silicium. from cell nutrient reservoir into organic compounds The maximal growth rates (it,„) of the two species follow the hyperbolic model. When comparing the are not significantly different, but A. fonnosa has abilities of several species with a slight advantage a K for phosphorus lower than C. meneghiniana in either assimilation or uptalce rates to compete (0.04 ii,g-at • L and 0.25 /..tg-at • L' , respectively) for one limiting nutrient, the model predicts that while the difference in K values for silicium is species with a low K value for growth and high opposite (3.9 ,ug-at • L -1 for A. fonnosa and 1.4 maximum uptake rate will dominate in high flux of Éti,g-at • L for C. meneglziniana). From these spe- nutrient input, while at low-nutrient concentrations

268 species with a low K value for uptake are advantaged. because of the low nutrient levels that follow diatom This was predictable with a more simple model , but bloom. All Hutchinson's observations appeared the model also points out that a short injection of to have been predictable, which enabled Lehman limiting nutrient creates a transient high concentration et al. (1975) to claim that detailed models, when that enables the species with high K for uptake to gain correctly formulated, can predict general patterns a competitive advantage. Grenney et al , pointed out precisely without costly simulation and that the simple that temporal variations in the limiting nutrient model cannot. No model is perfect, in the sense concentration may tender the environment suitable that none can simultaneously satisfy all demands for coexistence of different species. Thus, rather and prerequisites. than the magnitude of the characteristics themselves, Mickelson et al. (1979) experimented on the variations may be the most important factors influ- competition of the diatoms Chaetoceros septen- encing species equilibrium. Indeed, the Grenney trionalis , Skeletonema costatutn, and Thalassiosira et al. model was a critical step; however, it did not gravida grown in nitrogen (NH , )- limited continuous involve light and temperature variations and assumed cultures, to estimate the relative change, if any, all species are limited by the same nutrient. These in the specific growth constants obtained from single- weaknesses were remedied later. species chemostats, that may occur when two species As a matter of fact, Lehman et al. (1975) pro- are grown together. By inoculation of a small sample posed a theoretical framework that avoids the usual of an invader algal culture into a resident one (2-98%) shortcoming of considering a natural population as they demonstrated that both C. septentrionalis and a single unit, without any regard for physiological S. costatum rapidly dominated T. gravida which differences such as differences in nutrient uptake was displaced. Also, the latter strain was not able efficiencies. Their model integrates the nutrient to grow in resident cultures of the two other diatoms. constants of the hyperbolic model, assuming that Behavior of mixed cultures of C. septentrionalis uptake velocities are dependent on both internal and and S. costatum are related to initial proportions, external concentrations, and includes luxury con- i.e. the resident population (98%) displaced the sumption, end-product inhibition of carbon fixation invader, and the two species can coexist when relative and nutrient uptake, and light and temperature initial biomasses are equal. Moreover, temperature response constants as well. It predicts that a species stress (27°C) favored the growth of the less abundant dominance ends and a succession occurs when cell- species. Hence, two modes of dominance may occur; division rates of dominant species are slowed by one based on the different physiological capabilities nutrient shortage or physical limitations. This leads to compete for nutrient partition, and one related these algae to become less resistant to grazing and to species matching with equal success, in which sinking, the primary processes of algal cell disap- take-over occurs only when an additional factor is pearance. The model can also predict the daily pulses detrimental to one species and favors the other. related to light limitation of photosynthesis, and Finally, Mickelson et al. calculated the kinetic con- temporal changes in biomass and species compo- stants by Monod's (1942) equation (one-compartment sition. They did not use data to verify the ability model) and Droop's (1974) model (two-compartment of their model to apply to natural situations. However, model). Comparisons made with data in the literature they fitted the model to data for physiological con- and obtained by unialgal cultures demonstrated that stants and the exact values reported by Hutchinson no notable difference exists between the two sets for a pond he studied, i.e. nutrient contents of 0.1 of data. Hence, this contribution confirms the use ,ug-at-P • I.» with a daily input of 0.0003 it g- of single-species growth kinetics as a basis for pre- at-P • , 8.6 tg-at-N • L-1 and 0.035 pi-at- dicting the outcome of competition between two N • L- ' • , and 142.8 ,ug-at-Si • and 0.125 algae, by derivation of a theory described for bacteria

,tt g-at-Si • L-1 • . Algal densities were 2000 by Veldkamp and Jannasch (1972) and Jannasch cells • for diatoms, 1000 cells • for other taxa. and Mateles (1974). Computation demonstrated that the diatoms are the Validity of constant K-based predictions — first dominant algae. Nevertheless, they are the algae From the main contributions cited, the lcinetic con- most susceptible to sinking due to their siliceous stants of nutrient uptake and growth appear to reflect frustules and absence of motility. They need high the capability of an alga to compete with others for turbulence and rapid division rate to produce heavy nutrient partitioning , and to be valid for prediction populations in pelagic biota. Because of their more of species succession as well. The half-saturation efficient uptake of dissolved phosphorus and slower constants have been frequently used, probably be- sinking rates, the chrysophytes continued to increase cause their dimensions are nutrient concentrations, after diatom blooms collapsed. The other algae, which allow direct comparison with field data. namely chlorophytes and dinoflagellates, and cyano- Contrariwise, the maximal uptake or growth rates phytes never achieved substantial population densities are pratically neglected. Healey (1980) criticized

269 this misuse and recalled that the K values can be that the one species best able to take up and use the readily compared only when maximum rates are limiting nutrient would displace all other species. If the same. To illustrate his argument he replotted this rationale had been correct, the principle of "com- the data of McCarthy (1972) that showed the diatoms petitive exclusion" (as termed by Hardin (1960) Thalassiosira fluviatilis and T. pseudonana to have but known earlier (see Udvardy 1959)) would have the same K value for urea uptake, but a quite dif- been obeyed; all aquatic habitats would contain few ferent U,„. Hence, except at trace levels to which species and each ecological niche a single one. That the algae will respond similarly, the species with is obviously not true; marine and fresh waters usually the higher U,„„, will displace the other. Healey contain several tens of species in apparent competitive proposed to include both K and U,„„, values in one equilibrium. The violation of this principle by the index, in order to get a more general application, phytoplankton was termed "the paradox of the plank- namely the ratio 1.1,„„,:K , which is the slope of ton" by Hutchinson (1961). By using the word Monod's model at lowest nutrient concentrations. "paradox," Hutchinson symbolized the opinion of If the ratio U,„;,,:K is found in uptake studies using the whole community of planktonologists who were cells with minimum cell quota (kJ it should provide all deeply convinced that plankton must obey the an estimate of the U < :K ratio for growth which principle of exclusion. When apparently it does not, remains, the common result of all cellular processes. the explanation must be that environmental pertur- Healey's refinement is promising, indeed, and should bations have interupted and/or masked the normal certainly come into common use in the near future. behavior. Therefore, efforts were made to explain On the other hand, Crowley (1975), who dis- the paradox. cussed the role of natural selection on adjustment Hutchinson (1961) pointed out that the idealized of half-saturation constants to relative ambient sub- assumed in classical theories (i.e. space strate concentrations, demonstrated that the relative conditions uniform, no temporal variations, and all species fitnesses of the alternative evolutionary strategies having the same mortality rate) are violated in a determine these specific constants over evolutionary natural environment whose main characteristics time. In other words, the algae can adapt towards include spatial and temporal complexities (resulting the greater ability to take up the nutrient most bound from light, temperature, and nutrient concentration (affinity strategy) or present at very low levels (effi- variations), while grazing and sinking are species ciency strategy) or adapt to faster growth (velocity variable. Therefore, natural conditions are never strategy). Doyle (1975) emphasized that adapting stable long enough to allow the most efficient species low nutrient concentration does not necessarily to a at any time to eliminate the less successful species be- imply a decrease in the value of U.,. Slawyk (1980) fore the most efficient species loses dominance. Later, confirmed this later by recording high K and Richerson et al. (1970) insisted on the importance values with the phytoplankton of the Costa Rica of patchiness, which results from insufficient rates the half-saturation constant is still dome. However, of mixing compared with the reproductive rates of considered here as reflecting the specific capability phytoplankton. The distinct patches or temporary of an alga to compete for nutrient partition. Gavis niches separated in space represent different biota the rate of diffusion transport of (1976) claimed whose populations are then mixed by the net samplers nutrients within the depleted area around any indi- commonly used. Grenney et al. (1974) discussed vidual may govern nutrient-uptake rate and, hence, this suggestion and demonstrated that fluctuations growth. By comparing the respective K and U„,„, in the supply of limiting nutrient would allow several and diffusion coefficients of the diatoms Coscino- species of algae to persist together. Combined with lineatus (K = 2.8 ,u,g-at-N • L-1 ) and discus Di- other factor variations, they provide an unlimited tylum brightwellii (K = 0.6 ,u,g-at-N L-1 ), he number of combinations of which one or several predicted the latter organism would absorb nutrients may correspond to the respective requirements of more rapidly than the former only at concentrations different algal species. If suitable combinations occur less than 0.75 it g-at-N . This was not pre- with sufficient frequency and duration, a species dictable by comparing only the K values. However, may be able to survive indefinitively. The dynamics no experiment was done to verify whether the pre- of the system could be a major source of species diction pertains to reality. Altogether, there is often diversity. Turpin and Harrison (1979) investigated a significant discrepancy between predictions and this topic and demonstrated that, in marine waters, the real behavior of natural populations, which has and C. constrictus so far resisted elucidation so well that the term the diatoms Chaetoceros socialis in a homogeneous distribution of the lim- "paradox" was applied to it. dominate iting nutrient, while Skeletonema costatum was THE PARADOX OF THE PLANKTON dominant in a patchy distribution, but at the same On the basis of general knowledge on species time that one nutrient is limiting for some species, competition and algal physiology, one would expect other algae may be limited by other nutrients, and

270 then numerous species can coexist in a complex allow a multispecies co-occurrence. environment. O'Brien (1974), who simulated algal growth Williams (1971) suggested that temporal sepa- rate as a function of nutrient concentrations, and ration of nutrient uptake activities of different algae two-species competition, did not refer to the paradox. of the same community could allow numerous species Notwithstanding, his speculations led to interesting to coexist by increasing the number of the niches. remarks which may be relevant to this discussion. Diel periodicity in nitrogen assimilation studied by The cause of a population crash, he wrote, can be Eppley et al. (1971) with Skeletonetna costatum due to a large supply of nutrients in excess of that and Coccolithus huxleyi and diel-phased cell division which allows a steady state, when mortality and of natural populations of Ditylwn brighttvellii ob- growth are balanced. When that occurs, the daily served by Smayda (1975), among others, support rate of nutrient replacement is not fast enough to this statement. So pulses in nutrient supply have maintain the large algal biomass arising from the been associated with daily patterns of incorporation consumption of the whole nutrient pool; hence, a of carbon and nutrients to generate species coei- population crash follows the bloom. O'Brien's dis- istence by restricting the competitive advantage cussion demonstrated the importance of zooplankton of each species to a part of the daily cycle. For and other selective causes of mortality in influencing instance, the introduction of a nitrogen pulse with species composition and succession of phytoplankton. a period greater than 10 d in the model of Grenney He remarked that because K and ,u,„„, are specific et al. (1973) led to prediction of species equilibrium. constants, if the death rate for all species was the same Doyle and Poore (1974), who observed that all the one most able to take up nutrients would displace dinoflagellate species present in the Gulf Stream all others, whereas if the death rates vary within divide simultaneously, suggested that diurnal oscil- competitors several species can coexist at a given lation of nutrient supply could be responsible for nutrient concentration. Subsequently, , because the synchrony, by eliminating from the plankton all grazing is the main mortality factor, it appears that those species not phased with it. However, this effect the more different grazers present, the more variable will be diminished when nutrient availability and the death rates of phytoplankton and, therefore, the grazing rates are both high; the timing of cell division more species able to coexist. This is an alternative and variations in nutrient concentrations will not explanation of the paradox, indeed. be perfectly coupled. Stross and Pemrick (1974) Thus , all included, it seems to be quite clear reported conflicting results by observing a temporal that a natural population of algae could consist of stratification of nutrient niches in a lake that allows an assemblage of species having a little overlap, several functional groups of algae to take up nutrients although they share the same space. Most authors at different times of the day. Weiler and Chisholm who discussed the "paradox" assumed that equilib- (1976) observed that three species of dinoflagellates, rium is an asymptotic situation never fulfilled, be- Ceratium dens, C. furca, and Dinophysis fortii , cause of successive and varied violations of normal divide at different times of day in the waters of Santa behavior resulting from unstable environmental Monica Bay and, hence, minimize respective nutrient features. Recently, a few other authors have claimed privation. It appears, therefore, that temporal patchi- that several species of phytoplankton can co-occur ness results not only from direct responses of cell in a true competitive equilibrium, and others claim machinery to fluctuations of environment, but also that competition for nutrients is a myth. from species-specific physiological rhythms. The Theoretical speculations based on ability of importance of biological clocks in algal physiology species units to take up and assimilate nutrients led was demonstrated long ago (Sweeney and Hastings Stewart and Levin (1973) to state that coexistence 1962) and has become part of the basic background, of two species competing for a single resource can but their role in species competition has been recog- occur when one species grows faster at high concen- nized only recently. Otherwise, zooplankters that tration while the other grows faster at low concen- graze at night are supposed to liberate enough nu- trations of the limiting resource; at one intermediate trient to allow phytoplankton to grow the following concentration, the two relative capabilities equili- day (Wroblewski 1977). Computer simulation and brate. Moreover, the total biomass of both species experiments conducted with the fresh alga Euglena growing together is less than the biomass of either gracilis by Stross et al. (1973) and Chisholm and species in the same conditions; this was also inde- Nobbs (1976) and experiments by Chisholm et al. pendently predicted by Ross (1973). Petersen (1975) (1978) with a natural marine population dominated also simulated an algal competition. The model he by diatoms are persuasive, but do not prove that ran involved the usual kinetic constants plus specific properly phased daily rhythms in nutrient uptalce death rates and remineralization of dead cells as and nutrient pulses can restrict the advantage of each did O'Brien (1974), but it also permitted investigation species to a part of the cycle duration, and hence of the consequences of different specific uptake

271 capabilities. When a plurispecific equilibrium occurs, fended by Fuhs et al. (1972). To clarify this con- that implies (according to Petersen's statements), tention, a didactic language is needed. If one accepts that (i) each species is more adapted than the others the concept that there is a specific minimum cell to acquire at least one of the nutrients that are present quota (Droop's Ic(2 ) of each element to permit pro- in limiting amounts, (ii) each species is less efficient duction of a cell, it becomes obvious that the magni- in obtaining at low concentration the nonlimiting tude of the biomass of a population is given by the nutrient it requires in larger amounts to continue reservoir of the nutrient least available with respect to grow, (iii) the set of respective algal capabilities to the needs. Hence, nutrients limit growth. How- are matched. Subsequently, the model predicted that ever, one can imagine that any individual may find the number of coexisting species cannot exceed the sufficient nutrient supply to divide as fast as its cell total number of limiting nutrients, a result also ob- machinery permits , so nutrients would be not limiting. tained spearately by Taylor and Williams (1975). That would lead to conflicting features — individuals It also predicts that oligotrophic waters should host not growth limited (growth = growth rate), the communities richer in species than eutrophic waters, whole population growth limited (growth = yield). which, in fact, they do. Equilibrium is displaced Is this speculation unrealistic? Certainly it is, if we either by depletion or by heavy enrichment of a refer to unialgal cultures, but not necessarily if daugh- single nutrient; both cases lead to single species ter cells are eliminated as fast as they are produced, dominance. Likewise, the invasion of one species which can occur in natural conditions. most able to take up all nutrients at very low levels Hulburt (1970, 1976, 1977, 1979a, b) conceived will displace all the resident competitors. Thus, a theory whose main feature was that planktonic Petersen's speculations have been really fruitful. algae are not competing for nutrients. This concept Altogether the paradox, if not fully clarified, is can be summarized as follows: (i) Most oceanic no longer a frightening problem. But, on the other species do not vary in abundance over large areas hand, the direct comparison of specific kinetic con- of offshore waters, with the exception of a few like stants (common a decade ago) appears to be a simpli- Coceolithus huxleyi in the western North Atlantic fication presently in extinction, whereas recent Ocean. These species show an appreciable decrease developments have complicated further this im- in shallow waters, while C. huxleyi is more abundant portant scientific topic, by challenging the principle in coastal waters. (ii) This does not result either of nutrient limitation in natural condition. from competition for nutrients or from differential grazing pressure. (iii) Oceanic species grow too slowly to respond to increased nutrient contents of ARE NATURAL PHYTOPLANKTON coastal waters, but their capability to take up nutrients POPULATIONS NUTRIENT LIMITED? at very low concentrations makes them successful That nutrient availability limits algal growth in open ocean. (iv) In open ocean, cell densities of has been implicitly purported for as long as phyto- algal species do not vary very much because, when plankton have been studied. But the word "growth" a mother cell produces two daughter cells, it de- is inadequate here, in that it covers an array of pro- creases nutrients; when a daughter cell is grazed cesses that are sighted by two main consequences — and remineralized it increases nutrients. Therefore, the division rate of individuals and the crop of the nutrients are never lacking, and the oceanic species population. O'Brien (1972, 1973) proposed to clas- are in a permanent equilibrium. sify growth responses into type I, where growth The basis of Hulbures theory belongs to field continues longer and the final yield increases with observations, and refers essentially to the slight increasing concentrations of the limiting nutrient, variations over wide areas of population densities, and type II, where growth rate increases and the and nutrient concentrations, as well. The theory yield remains constant. Holmes (1973) and Kelly evidently applies also to the "paradox of the plank- and Hornberger (1973) severely criticized this concept ton" and provides an alternative explanation. and demonstrated the final yield will depend on the Hulburt (personal communication) wrote: "If there concentration of a limiting nutrient and not neces- is one nutrient concentration which is determined sarily on how fast the algae grow, but a nutrient by every species, then every species determines its addition will usually increase both nutrient growth own different concentration of nutrient. One con- rate and final biomass of natural populations. O'Brien centration of nutrient is maintained by mixing and replied that these figures are unrealistic as popu- extends over distances of a meter or more. Different lations do not have unlimited time to divide. They concentrations of nutrient can be maintained theo- have to face severe causes of mortality that can be retically about each cell for small distances only, compensated only by rapid changes in growth rate, 0.17-0.37 x 10 cm 2 in the Sargasso Sea, for leading to the possibility of both unlimited growth instance. This means that mixing always prevents rate and nutrient-limited yield, a concept also de- any particular species from being in equilibrium 272 with its own nutrient concentration as in a chemostat. up at very low nutrient levels, namely undetectable So, all species but one are not competed out of trace levels. Is that possible? existence by that one which can force the nutrient Steemann Nielsen (1978a, b) brought a new to the lowest concentration, i.e. its own equilibrium insight that could begin to answer the question. He concentration." Yet this formulation is not quite investigated algal growth rate at very low nutrient clear, at least for us. The mechanism evoked is the levels, but he used a special batch technique that same as that proposed by Petersen (1975), Taylor allowed him to record the growth of a population as and Williams (1975), and Titman (1976) who stated rare as 10" cells • L- '. With the fast-growing fresh- that at equilibrium each species is limited by a dif- water algal Selenastrum capricornutum , whose ferent nutrient and, therefore, the number of coex- > 2 divisions • d -1 , he demonstrated that isting species cannot exceed the total number of 75% of it,„„, occurs at concentrations as low as limiting nutrients. Their ideas diverge, however, 0.02 gg-at-N (NH 1 ) • L-'. The K value could not because they assume the phytoplankton is nutrient be determined because concentrations of NH lower limited, which Hulburt denies. By comparing specific than 0.02 ,us-at • L - ' could not be measured and, K values to nutrient concentrations, several authors therefore, the 50% of bc,,,„, never obtained. With have also concluded that the populations they studied NO3 , the nitrogen requirements are higher, and the were nutrient unlimited (Thomas 1969, 1970a, b; K value could be determined to be 0.05 gg-at L - ', MacIsaac and Dugdale 1969; Titman and Kilham while 75% of it., is given by 0.2 lig-at•L' 1976) and, hence, implicitly supported Hulburt's Requirements for phosphorus are also very low: K statement. was found at 7 ng-at • L -1 , which was the lowest The contention will not be solved here; however, concentration that could be produced. Although we we are suspicious about how perfect and how per- can question why the uptake rates are so much dif- manent is the balance between absorption of nutrients ferent for N-NH, and N-NO,, , it remains quite clear for growth and production of nutrients through that in nutrient-poor waters S. capricornutum is remineralization. Because the later process involved a capable of growth as fast as its cell machinery permits, series of mechanisms spread along the food chain, light and temperature permitting. one can expect an algal cell to be able to grow faster Similarly, Goldman et al. (1979) claimed that than the nutrients are regenerated, thence to be nu- growth rates of natural phytoplankton populations trient limited, even though the K value would be in oceanic waters may be near maximal growth and, very low. Moreover, some species are evidently hence, not nutrient limited. They remarked that the nutrient limited as they grow faster as soon as they cell contents of the algae Dunaliella tertiolecta , benefit from nutrient-rich waters. So does Cocco- Monochrysis lutheri (now Pavolova lutheri), and lithus huxleyi in coastal waters as reported by Hulburt Thalassiosira pseudonana 3H approach the Redfield (1979a), for whom he implicitly does not deny that ratio only at high growth rates. Hence, they dis- nutrients limit its growth in oceanic waters. To cussed whether the chemical composition of natural explain why oceanic species decrease in shallow marine phytoplankton is typically characterized by waters, but C. huxleyi increases, Hulburt wrote: this ratio because growth rates in oceanic waters "One simple reason why oceanic species decrease in are quite high (Goldman 1980). Berland et al. (1972, shallow water is that they get stuck on the bottom, p. 343) also remarked the natural populations have stranded there. All cells of all species get stuck there, the same basic cellular composition as cultured cells of course. But the slow growers grow too slowly to growing in nutrient-unlimited conditions, namely the balance loss by stranding; and the fast growers, lilce logarithmic phase of a batch culture in their experi- C. huxleyi, grow faster, due to more nutrient, than ments, and Slawyk et al. (1978) reported that the they get stranded. More nutrient in shallow water is ratio POC: PON of marine phytoplankton of northwest of no use to slow growers, since they are already African upwelling is similar to that found in steady- growing at the maximum in low nutrient deep water." state chemostat cultures. Goldman et al. reconciled Hulburt's original insight implies that all species the steady state of chemostat and the euphotic zone present in oceanic areas are physiologically adapted of the open ocean by the following rationale: (i) In to take up nutrients at very low levels, a capability a chemostat, the growth rate equals dilution rate, but which had not been really explored until recently. remains completely independent of the concentration But it still remains unclear how such an alga as C. of limiting nutrient. (ii) Residual limiting nutrient huxleyi can live in equilibrium with species uniquely flowing into the "sump" are undetectable. (iii) Thus, adapted to the oceanic conditions and can also bloom it is possible to have simultaneously low or unde- in nutrient-rich waters. Such a behavior implies that tectable residual nutrient levels and high growth this alga has at one time, or can adapt rapidly to rates, as is shown in uptake measurements by show, the following physiological capabilities: high McCarthy and Goldman (1979). McCarthy and specific growth rate, low K value, capability to take Goldman demonstrated that the ratio of uptake rate

273 of NH versus growth rate of the diatom T. pseudo- and 16 gg-at-N • ), a part of the NH I could have nana increases from 1.1 to 33.9, when the dilution been absorbed and accounted for with that taken up, rate decreases. In other words, when the nutrient but killed-cell suspensions do not retain activity. concentration decreases to very low levels, the uptake Collos and Slawyk (1980), who replotted the data process is able to sustain a far higher division rate. of Caperon and Meyer (1972), Eppley and Renger Therefore, under conditions of severe starvation, a (1974), and McCarthy and Goldman (1979) obtained cell needs to be exposed to saturation NH I only for with the diatom T. pseudonana , remarked that from a very short period with regard to the doubling time total nutrient depletion to trace concentrations (which to replenish its nutrient reservoir. It can do so by allow a nitrogen cell content of about 0.5 gg-at • passing into microenvironments in which nutrient cell - '), the specific uptake rate increases until a concentrations are elevated, for instance by excretion maximum, then decreases. Hence, the authors who or by remineralization of dead organisms, whose worked with higher values (e.g. McCarthy and Gold- existence is also defended by Wroblewski (1977). man) have missed the left part of the peak and uptake This new concept on adaptation of algal physio- overestimated the algal capability to adapt the logy to conditions of nutrient depletion is surprising at very low nutrient concentrations. In Collos and and leads to some questions. Is it, for instance, totally Slawyk's (1980) opinion, algae react to nitrogen relevant to argue that an open ocean works as a steady limitation by reducing their cell quota (Dugdale U„,„,; but state of a chemostat? In a chemostat, there are two 1977), thereby increasing their specific (Droop 1974). There- privileged points generally separated by the largest there is a minimum cell quota a decrease in distance flooded by the culture medium — the orifice fore, in case of extreme deficiency, for by a parallel decrease of input and the orifice of output; between the two, U,„„, cannot be compensated decrease in nutrients disappear. The cells, which are randomly in nitrogen cell quota; this leads to a distributed by mixing, benefit, from time to time, specific U.,. Collos (1980) confirmed this proposed of the from transient concentrations corresponding to mechanism by investigating the response demon- those at which uptake machinery is efficient. diatom Phaeodactylwn tricornutum and pseudonana. McCarthy and Goldman assume that nutrient-rich strating that it behaves similarly to T. microniches provide similar opportunities, and that Otherwise, Stross (1980) suggested that an phytoplankton might appear to be growing success- alternative arrangement may be conceived for diatoms fully at the expense of virtually undetectable small- living in oligotrophic fresh waters. For such organ- scale nutrient pulses. Jackson (1980) denies the isms, when a small pool of phosphate becomes existence of such niches; his calculations prove the available it can be immediately used (r strategy) excreta of a protozoan are dispersed fast enough because the cell is equipped with efficient enzymes to make the concentration at the center of a 10-gm and excess of RNA, or can be deferred until the plume only 0.1% of the initial concentration within necessary RNA is made available (K strategy), and 0.1 s. Behind a swimming copepod a decrease to 1% permits, in contrast, a larger biomass. of the concentration of the excreta will require only Beyond the problem of stating whether algal 100 s; yet the vicinity of a zooplankter is more often growth is nutrient limited or unlimited, the sub- deadly, the phytoplankton will not encounter high- sequent problem of estimating the importance of nutrient level there, except with macrozooplankters, nutrient partition in species competition and suc- which are very rare. Jackson also remarked that in cession arises. Obviously, if the concepts of Hulburt, the central North Pacific gyre the average division Goldman, and McCarthy pertain to natural popu- rate of the phytoplankton is 0.1 d - '; the diatom lations, the role of nutrients has been greatly over- Thalassiosira pseudonana would maintain such a estimated; hence, most models will have to be growth rate with a NH concentration of only 0.0008 reconsidered. However, prior to any hasty statement, gg-at • Thus, slow phytoplankton growth rates one could remark that only a few species have been are consistent with low nutrient levels and slow involved in experiments that demonstrate the algal zooplankton abundance and filtration rate. Williams capability to take up nutrient very rapidly. Goldman and Muir (1981), on the basis of their own compu- et al. (1979) and McCarthy and Goldman (1979) tations and those of Pasciak and Gavis (1974), refer essentially to data obtained with Thalassiosira established that molecular diffusion is sufficient to pseudonana , while similar capability reported by disperse microzones of excreted nutrients 50 times Conway et al. (1976), Conway and Harrison (1977), too rapidly to be taken up by phytoplankton. En- and Turpin and Harrison (1979) issued from experi- hanced uptake rates at very low substrate concen- ments made with Skeletonema costatum . Both these trations reported by McCarthy and Goldman (1979) diatoms are well known for their active metabolism, can be suspected to have been overestimated. Be- which makes them convenient experimental material; cause their short-term measurements (5 min) were numerous others show less-efficient metabolism. made with relatively high concentrations of NH (8 Also, nitrogen was the only nutrient involved in the

274 studies, while several workers demonstrated the constants for three species of marine phytoplankton. importance of the silicium in seawater and phosphorus Ecology 52: 183-185. in fresh waters. Therefore, it is our opinion that CHISHOLM, S. W., F. AZAM, AND R. W. EPPLEY. 1978. there are so far no convincing reasons to generalize to Silicic acid incorporation in marine diatoms on light: dark cycles: use as an assay for phased celle division. all algal species. Hence, if the concept of nutrient- Limnol. Oceanogr. 23: 518-529. unlimited algal growth has to be restricted to some Cmstiœ,m, S. M., AND P. A. NOBBS. 1976. Simulation species having the ability to assimilate large quantities of algal growth and competition in a phosphate-limitecl of nutrients in a short time duration and to shift their cyclostat, p. 337-355. In R. P. Canale [ed.] Modeling cell machinery to take advantage of any transient biochemical processes in aquatic ecosystems. Ann pulse of nutrients , others would have to face a drastic Arbor Sci. Publ., Ann Arbor, MI. disadvantage. We should have just gone one step CoLLos, Y. 1980. Transient situations in nitrate assimi- more in the discouraging complexity of the phyto- lation by marine diatoms. 1. Changes in uptake para- plankton mechanisms of competition and coexistence; meters during nitrogen starvation. Limnol. Oceanogr. this is , unfortunately, , what we have to expect. 25: 1075-1081. CoLLos, Y., AND G. SLAWYK. 1980. Nitrogen uptake and assimilation by marine phytoplankton, p. 195-211. In P. Falkowsky [ed.] Primary productivity in the Acknowledgments sea. Plenum Press, New York, NY. CONWAY, H. L., AND P. J. HARRISON. 1977. Marine We wish to thank greatly Dr Trevor Platt for diatoms grown in chemostats under silicate or ammo- nium limitation. IV. Transient response of Chaetoceros revising the English manuscript. debilis , Skeletonema costatutn and Thalassiosira gravida to a single addition of the limiting nutrient. Mar. Biol. 43: 33-43. References CONWAY, H. L., P. J. HARRISON, AND C. G. DAVID. 1976. Marine diatoms grown in chemostats under Transient response AHLGREN, G. 1977. Growth of Oscillatoria agardhii in silicate or ammonium limitation. II. to a single addition of the chemostat culture. I. Nitrogen and phosphorus require- of Skeletotzema costatum limiting nutrient. Mar. Biol. 35: 187-199. ments. Oikos 29: 209-224. H. 1975. Natural selection and the Michaelis BERLAND, B. R., D. J. BONIN, P. L. LABORDE, AND CROWLEY, P. J. Theor. Biol. 50: 461-475. S. Y. MAESTRINI. 1972. Variations de quelques constant. W. 1975. Upwelling, clone selection and the facteurs estimatifs de la biomasse, et en particulier DoYLE, R. characteristic shape of nutrient uptalce curves. Limnol. de l'ATP, chez plusieurs algues marines planctoniques. Oceanogr. 20: 487-489. Mar. Biol. 13: 338-345. DOYLE, R. W., AND R. V. POORE. 1974. Nutrient compe- BERLAND, B. R., D. J. BONIN, AND S. Y. MAESTRINI. tition and division synchrony in phytoplankton. J. 1980. Azote ou phosphore? Considérations sur le Exp. Mar. Biol. Ecol. 14: 201-210. "paradoxe nutritionnel" de la mer Méditerranée. DROOP, M. R. 1961. Vitamin B12 and marine ecology: Oceanol. Acta 3: 135-142. the response of Monochrysis lutheri . J. Mar. Biol. BREMER, H., AND P. P. DENNIS. 1975. Transition period Assoc. U.K. 41: 69-76. following a nutritional shift-up in the bacterium 1968. Vitamin B 12 and marine ecology. IV. The Escherichia coli Bir: stable RNA and protein syn- kinetics of uptake, growth and inhibition in Mono- thesis. J. Theor. Biol. 52: 365-382. chrysis lutheri . J. Mar. Biol. Assoc. U.K. 48: 689- BROWN, E. J., AND R. F. HARRIS. 1978. Kinetics of 733. algal transient phosphate, uptake and the cell quota 1974. The nutrient status of algal cells in con- concept. Limnol. Oceanogr. 23: 35-40. tinuous culture. J. Mar. Biol. Assoc. U.K. 54: 825- BROWN, E. J., AND D. L. BUTTON. 1979. Phosphate- 855. limited growth kinetics of Selenastrum capricornutwn DUGDALE, R. C. 1967. Nutrient limitation in the sea: (Chlorophyceae). J. Phycol. 15: 305-311. dynamics, identification and significance. Limnol. BUTTON, D. K. 1978. On the theory of control of microbial Oceanogr. 12: 685-695. growth kinetics by limiting nutrient concentrations. 1977. Modeling, p. 789-806. In E. D. Goldberg Deep Sea Res. 25: 1163-1177. [ed.] The sea: ideas and observations on progress in CAPERON, J. 1967. Population growth in micro-organisms the study of the sea. J. Wiley and Sons, New York, limited by food supply. Ecology 48: 715-721. NY. 1968. Population growth response of Isochrysis DUGDALE, R. C., AND J. J. GOERING. 1967. Uptake of galbana to nitrate variation at limiting concentrations. new and regenerated forms of nitrogen in primary Ecology 49: 866-872. productivity. Limnol. Oceanogr. 12: 196-206. CAPERON, J., AND J. MEYER. 1972. Nitrogen-limited ELBRÂCHTER, M. 1977. On population dynamics in multi- growth of marine phytoplankton. H. Uptake kinetics species cultures of diatoms and dinofiagellates. Helgol. and their role in nutrient-limited growth of phyto- Wiss. Meeresunters. 30: 192-200. plankton. Deep Sea Res. 19: 619-632. EPPLEY, R. W., AND J. L. COATSWORTH. 1968. Uptake CARPENTER, E. J., AND R. R. L. GUILLARD. 1971. of nitrate and nitrite by Ditylum brightwellii - kinetics Intraspecific differences in nitrate half-saturation and mechanisms. J. Phycol. 4: 151-156.

275 EPPLEY, R. W., AND E. H. RENGER. 1974. Nitrogen HARRISON, P. J., H. L. CONWAY, AND R. C. DUGDALE. assimilation of an oceanic diatom in nitrogen-limited 1976. Marine diatoms grown in chemostats under continuous culture. J. Phycol. 10: 15-23. silicate or ammonium limitation. I. Cellular chemical EPPLEY, R. W., J. N. ROGERS, AND J. J. MCCARTHY. composition and steady-state growth kinetics of Ske- 1969. Half-saturation constants for uptake of nitrate letonema costatum. Mar. Biol. 35: 177-186. and ammonium by marine phytoplankton. Limnol. HEALEY, F. P. 1980. Slope of the Monod equation as an Oceanogr. 14: 912-920. indicator of advantage in nutrient competition. Micro- EPPLEY, R. W., J. N. ROGERS, J. J. MCCARTHY, AND biol. Ecol. 5: 281-286. A. SOURNIA. 1971. Light/dark periodicity in nitrogen HOLMES, P. 1973. Phytoplankton algae: nutrient concen- assimilation of the marine phytoplankters Skeletonema trations and growth. Science (Washington, D.C.) 180: costatum and Coccolithus huxleyi in N-limited chem- 1298-1299. ostat culture. J. Phycol. 7: 150-154. HOLMES, R. W., P. M. WILLIAMS, AND R. W. EPPLEY. EPPLEY, R. W., AND W. H. THOMAS. 1969. Comparison 1967. Redwater in La Jolla Bay, 1964-1966. Limnol. of half-saturation constants for growth and nitrate Oceanogr. 12: 503-512. uptake of marine phytoplankton. J. Phycol. 5: 375- HULBURT, E. M. 1970. Competition for nutrients by 379. marine phytoplankton in oceanic , coastal, and estuarine ESTRADA, R. W., AND D. BLASCO. 1979. Two phases regions. Ecology 51: 475-484. of the phytoplankton community in the Baja California 1976. Limitation of phytoplankton species in the upwelling. Limnol. Oceanogr. 24: 1065-1080. ocean off western Africa. Limnol. Oceanogr. 21: FUHS, G. W., S. D. DEMMERLE, E. CANELLI, AND 193-211. M. CHEN. 1972. Characterization of phosphorus- 1977. Coexistence, equilibrium, and nutrient limited plankton algae (with reflexions on the limiting- sharing among phytoplankton species of the gulf of nutrient concept), p. 113-132. In G. E. Likens [cd.] Maine. Am. Nat. 111: 967-980. and eutrophication. Vol. 1. Am. Soc. Nutrients 1979a. An assymetric formulation of the distri- Limnol. Oceanogr. Spec. Symp. bution characteristics of phytoplankton species: an GATES, W. E., AND J. T. MARLAR. 1968. Graphical investigation in interpretation. Mar. Sci. Commun. analysis of batch culture data using the Monod ex- 5: 245-268. pressions. J. Water Pollut. Control Fed. 40: R469- 1979b. Russell's definite description analysis of R476. production and limitation of phytoplankton species. GAUSE, G. F. I932a. Experimental studies on the struggle Mar. Biol. 52: 321-329. for existence. I. Mixed populations of two species of HUTCHINSON, G. E. 1961. The paradox of the plankton. yeast. J. Exp. Biol. 9: 389-402. Am. Nat. 95: 137-145. 1932b. Ecology of populations. Q. Rev. Biol. JACKSON, G. A. 1980. Phytoplankton growth and zoo- 7: 27-46. plankton grazing in oligotrophic oceans. Nature GA VIS, J. 1976. Munk and Riley revisited: nutrient diffusion (London) 284: 439-440. transport and rates of phytoplankton growth. J. Mar. JANNASCH, H. W., AND R. I. MATELES. 1974. Experi- Res. 34: 161-179. mental bacterial ecology studied in continuous culture, GOLDMAN, J. C. 1980. Physiological processes, nutrient p. 165-212. In A. H. Rose and D. W. Tempest [ed.] availability, and the concept of relative growth rate Advances in microbial physiology, Vol. 11. Academic in marine phytoplankton ecology, p. 179-194. In Press, London. P. Falkowsky [cd.] Primary productivity in the sea. KELLY, M. G., AND G. M. HORNBERGER. 1973. Phyto- Plenum Press, New York, NY. plankton algae: nutrient concentrations and growth. Science (Washington, D.C.) 180: 1298-1299. GOLDMAN, J. C., W. J. OSWALD, AND D. JENKINS. KING, D. L., AND J. T. NOVAK. 1974. The kinetics of 1974. The kinetics of inorganic carbon limited algal inorganic carbon-limited algal growth. J. Water growth. J. Water Pollut. Control Fed. 46: 554-574. Pollut. Control. Fed. 46 (7): 1812-1816. GOLDMAN, J. C., J. J. MCCARTHY, AND D. G. PEAVEY. LAW, A. T., AND 1979. Growth rate influence on the chemical compo- D. K. BuTTorq. 1977. Multiple-carbon- sition of phytoplankton in oceanic waters. Nature source-limited growth kinetics of a marine coryneform (London) 279: 210-215. bacterium. J. Bacteriol. 129: 115-123. GRENNEY, W. J., D. A. BELLA, AND H. C. CURL JR. LEHMAN, J. T., D. B. BOTKIN, AND G. E. LIKENS. 1973. A theoretical approach to interspecific compe- 1975. The assumptions and rationales of a computer tition in phytoplankton communities. Am. Nat. 107: model of phytoplankton population dynamics. Limnol. 405-425. Oceanogr. 20: 343-364. GRENNEY, W. J., D. A. BELLA, AND H. C. CURL JR. LIEBIG, J. 1840. Die Chemie in ihrer Anwendung auf 1974. Effects of intracellular nutrient pools on growth Agrikultur und Physiologie. (4th ed., 1847). Taylor dynamics of phytoplankton. J. Water Pollut. Control. & Walton, London. Fed. 46: 1751-1760. LOTKA, A. J. 1920. Analytical note on certain rhythmic GUILLARD, R. R. L., P. KILHAM, AND T. A. JACKSON. relations in organic systems. Proc. Natl. Acad, Sci. 1973. Kinetics of silicon-limited growth in the marine U.S.A. 6: 410-415. diatom Thalassiosira pseudonana Hasle and Heimdal 1934. Théorie analytique des associations bio- (= Cyclotella nana Hustedt). J. Phycol. 9: 233-237. logiques, p. 1-45. In G. Tessier [cd.] Exposés de HARDIN, G. 1960. The competitive exclusion principle. biométrie et de statistique biologique. IV. Actual. Sci, Science (Washington, D.C.) 131: 1292-1297. Ind. 187.

276 MCCARTHY, J. J. 1972. The uptake of urea by marine PRICE, C. A., AND J. W. QUIGLEY. 1966. A method for phytoplankton. J. Phycol. 8: 216-222. determining quantitative zinc requirements for growth. MCCARTHY, J. J., AND J. C. GOLDMAN. 1979. Nitro- Soil Sci. 101: 11-16. genous nutrition of marine phytoplankton in nutrient- PRUDER, G. D., AND E. T. BouroN. 1979. The role of depleted waters. Science (Washington, D.C.) 203: CO2 enrichment of aerating gas in the growth of an 670-672. estuarine diatom. Aquaculture 17: 1-15. MACISAAC, J. J., AND R. C. DUGDALE. 1969. The QASIM, S. Z., P. M. A. BHATTATHIRI, AND V. P. kinetics of nitrate and ammonia uptake by natural DEVASSY. 1973. Growth kinetics and nutrient require- populations of marine phytoplankton. Deep Sea Res. ments of two tropical marine phytoplankters. Mar. 16: 45-57. Biol. 21: 299-304. MACISAAC, J. J., R. C. DUGDALE, S. A. HUNTSMAN, RICHERSON, P., R. ARMSTRONG, AND C. R. GOLDMAN. AND H. L. CONWAY. 1979. The effect of sewage on 1970. Contemporaneous disequilibrium, a new hypo- uptake of inorganic nitrogen and carbon by natural thesis to explain the "paradox of the plankton." Proc. populations of marine phytoplankton. J. Mar. Res. Natl. Acad. Sci. U.S.A. 67: 1710-1714. 37: 51-66. RILEY, G. A. 1963. Theory of food chain relation in the MACISAAC, J. J., R. C. DUGDALE, AND G. SLAWYK. ocean, p. 438-463. In M. N. Hill [ed.] The sea. Vol. 1974. Nitrogen uptake in the northwest Africa up- 2. Interscience Pub!., New York, NY. welling area: results from the CINECA-Charcot Ross, G. R. 1973. A model for the competitive growth cruise. Tethys 6: 69-76. of two diatoms. J. Theor. Biol. 42: 307-331. MAESTRINI, S. Y., AND M. G. KOSSUT. 1981. In situ SCHAECHTER, M. 1968. Growth: cells and populations, cell depletion of some marine algae enclosed in dialysis p. 136-162. In J. Mandelstam and K. McQuillen sacks and their use for the determination of nutrient [ed.] Biochemistry of bacterial growth. J. Wiley and limiting growth in Ligurian coastal waters (Mediter- Sons, New York, NY. ranean Sea). J. Exp. Mar. Biol. Ecol, 50: 1-19. SLAWYK, G. 1980. L'absorption de composés azotés par MICHAELIS, L., AND M. M. L. MENTEN. 1913. Die le phytoplancton marin: rôle dans la production pri- kinetik der Invertinwirkung. Biochem. Z. 49: 333- maire, relations avec la photosynthèse et les variables 369. du milieu extra et intracellulaire. Thèse Doct. ès Sc., MICKELSON, M. J., H. MASKE, AND R. C. DUGDALE. Univ. Aix-Marseille II. 213 p. 1979. Nutrient-determined dominance in multispecies SLAWYK, G., Y. COLLOS, M. MINAS, AND J. R. GRALL. chemostat cultures of diatoms. Limnol. Oceanogr. 1978. On the relationship between carbon-to-nitrogen 24: 298-315. composition ratios of the particulate matter and growth MONOD, J. 1942. Recherches sur la croissance des cultures rate of marine phytoplankton from the northwest bactériennes. 2nd ed. Hermann, Paris. 210 p. African upwelling area. J. Exp. Mar. Biol. Ecol. 33: O'BRIEN, W. J. 1972. Limiting factors in phytoplankton 119-131. Algae: their meaning and measurement. Science SMAYDA, T. J. 1973. The growth of Skeletonema costatum (Washington, D.C.) 178: 616-617. during a winter-spring bloom in Narragansett Bay, 1973. Phytoplankton algae: nutrient concen- Rhode Island. Norw. J. Bot. 20: 219-247. trations and growth. (Reply to comments by Holmes 1975. Phased cell division in natural populations and Kelly and Hornberger). Science (Washington, of the marine diatom Ditylum brightwellii and the D.C.) 180: 1298-1300. potential significance of diel phytoplankton behavior 1974. The dynamics of nutrient limitation of in the sea. Deep Sea Res. 22: 151-165. phytoplankton algae: a model reconsidered. Ecology STEELE, J. H. 1959. The quantitative ecology of marine 55: 135-141. phytoplankton. Biol. Rev. Cambridge Philos. Soc. PAASCHE, E. 1973. Silicon and the ecology of marine 34: 129-158. plankton diatoms. II. Silicate-uptalce kinetics in five STEEMANN NIELSEN, E. 1978a. I. Productivity and diatom species. Mar. Biol. 19: 262-269. modeling. Growth of the unicellular alga Selenastrum PARSONS, T. R., AND J. D. H. STRICKLAND. 1962. On capricornutum as a function of P. With some infor- the production of particulate organic carbon by hetero- mation also on N. Verh. Int. Ver. Limnol. 20: 38-42. trophic processes in seawater. Deep Sea Res. 8: 211- 1978b. Growth of plankton algae as a function 222. of N-concentration, measured by means of a batch PASCIAK;W. J., AND J. GAVIS. 1974. Transport limitation technique. Mar. Biol. 46: 185-189. of nutrient uptake in phytoplankton. Limnol. Oceanogr. STEWART, F. M., AND B. R. LEVIN. 1973. Partioning 19: 881-888. of resources and the outcome of interspecific compe- PERRY, M. J. 1976. Phosphate utilization by an oceanic tition: a model and some general considerations. Am. diatom in phosphorus-limited chemostat culture and Nat. 107: 171-198. in the oligotrophic waters of the central North Pacific. STROSS, R. G. 1980. Growth cycles and nutrient-limited Limnol. Oceanogr. 21: 88-107. photosynthesis in phytoplankton. Limnol. Oceanogr. PETERSEN, R. 1975. The paradox of the plankton: an 25: 538-544. equilibrium hypothesis. Am. Nat. 109: 35-49. STROSS, R. G., S. W. CHISHOLM, AND T. A. DOWNING. PRICE, C. A., AND E. F. CARELL. 1964. Control by iron 1973. Causes of daily rhythms in photosynthetic rates of chlorophyll formation and growth in Euglena of phytoplankton. Biol. Bull. Woods Hole. 145: gracilis. Plant Physiol. 39: 862-868. 200-209.

277 STROSS, R. G., AND S. M. PEMRICK. 1974. Nutnent DDYARDY, M. F. D. 1959. Notes on the ecological concepts uptake kinetics in phytoplankton: a basis for niche of habitat, biotope and niche. Ecology 40: 725-728. separation. J. Phycol. 10: 164-169. VELDKAMP, H., AND H. W. JANNASCH. 1972. Mixed SWEENEY, B. M., AND J. W. HASTINGS, 1962. Rhythms, culture studies with the chemostat. J. Appl. Chem. p. 687-700. In R. A. Lewin [ed.] Physiology and Biotechnol. 22: 105-123. biochemistry of algae. Academic Press, New York, VOLTERRA, V. 1926. Fluctuations in the abundance of a NY. species considered mathematically. Nature (London) TAYLOR, P. A., AND P. J. LE B. WILLIAMS. 1975. 118: 558-560. Theoretical studies on the coexistence of competing VOLTERRA, V., AND U. D'ANCONA. 1935. Les asso- species under continuous-flow conditions. Can. J. ciations biologiques au point de vue mathématique. Microbiol. 21: 90-98. In V. G. Tessier [cd.] Exposés de biométrie et de THOMAS, W. H. 1969. Phytoplankton nutrient enrichment statistique biologique. Actual. Sci. Inc. 243. experiments off Baja California and in the eastern WEILER, C. S., AND S. W. CHISHOLM. 1976. Phased tropical Pacific Ocean. J. Fish. Res. Board Can. cell division in natural populations of marine dino- 26: 1133-1145. flagellates from shipboard cultures. J. Exp. Mar. Biol. 1970a. On nitrogen deficiency in tropical pacific Ecol. 25: 239-247. oceanic phytoplankton: photosynthetic parameters in WILLIAMS, F. M. 1971. Dynamics of microbial popu- poor and rich water. Limnol. Oceanogr. 15: 380-385. lations, p. 197-267. In B. C. Patten [cd.] Systems 1970b. Effect of ammonium and nitrate concen- analysis and simulation in ecology. Vol. I. Academic tration on chlorophyll increases in natural tropical Press, New York, NY. pacific phytoplankton populations. Limnol. Oceanogr. WILLIAMS, P. J. LE B., AND L. R. MUIR. 1981. Diffusion 15: 386-394. as a constraint on the biological importance of micro- TILMAN, D. 1977. Resource competition between plane- in the sea. Proc. 12th Colloq. Ocean Hydro- tonic algae: an experimental and theoretical approach. zones Ecology 58: 338-348. dynam. Liège. TITMAN, D. 1976. Ecological competition between algae: WROBLEWSKI, J. S. 1977. Vertically migrating herbivorous experimental confirmation of resource-based compe- plankton. Their possible role in the creation of small tition theory. Science (Washington, D.C.) 192: 463- scale phytoplankton patchiness in the ocean, p. 817- 465. 847. In N. R. Andersen and B. J. Zahuranec [cd.] TITMAN, D., AND P. KILHAM. 1976. Sinking in fresh- Plenum Press, Oceanic sound scattering prediction. water phytoplankton: some ecological implications New York, NY. of cell nutrient status and physical mixing processes. YODER, J. A. 1979. A comparison between the cell division Limnol. Oceanogr. 21: 409-417. rate of natural populations of the marine diatom TURPIN, D. H., AND P. J. HARRISON. 1979. Limiting Skeletonemn costatum (Greville) Cleve grown in nutrient patchiness and its role in phytoplankton dialysis culture and that predicted from a mathematical ecology. J. Exp. Mar. Biol. Ecol. 39: 151-166. model. Limnol. Oceanogr. 24: 97-106.

278 Importance of Organic Nutrients for Phytoplankton Growth in Natural Environments: Implications for Algal Species Succession

DANIEL J. BONIN AND SERGE Y. MAESTRINI

Station Marine d'Endowne, Chemin de la Batterie des Lion, F-I3007 Marseille

Introduction The significance of the utilization of organic compounds is very complex and varies with the One probable role of organic compounds in algae (Droop 1974; Neilson and Lewin 1974; Hastings natural waters is to support the growth of phyto- and Thomas 1977; Hellebust and Lewin 1977). Ex- plankters when major inorganic nutrients are totally perimentally, it has been demonstrated that many lacking or are present at a very low concentration, algae can grow in the total dark if an organic or when the light intensity is very low. The ecological compound is added in the medium. This has been importance of these phenomena has been strongly observed for algae of all taxonomie groups. These debated, in spite of the fact that heterotrophic assimi- compounds are used as energy source and some- lation capabilities have been observed in numerous times also as carbon source. For example, it has freshwater and marine algae. As a matter of fact, been shown that when CO 2 is lacking in the medium, the ability of many planktonic algae to use diverse Navicula pelliculosa (Lewin 1953) can grow in organic nutrients in laboratory cultures is well docu- presence of light if glycerol is added in the medium. mented. However, the kinetics of such uptake often Likewise, Euglena gracilis (Murray et al. 1970) can appear to be quite inadequate for the low concen- grow if glycolate, glycine, or serine is provided as trations of organic compounds found in natural single carbon source. On the other hand, it is men- environments. The high half-saturation (K) values tioned that organic nutrients stimulate the growth observed for organic nutrient assimilation by phyto- of some obligatory photoautotrophs maintained in plankters led to the conclusion that, typically, algae very dim light (Cheng and Antia 1970; Cooksey cannot compete with bacteria for the dilute organic and Chansang 1976; Morrill and Loeblich 1979). substrates usually present in natural waters; and algal But no ecological significance can be attributed heterotrophy is then considered as a laboratory artifact to these capabilities until more is known of the arising from the axenic conditions in cultures and performance of the algae under natural conditions. from the very high concentrations of substrates used Not only the concentrations and types of organic in most experiments. The dependence of the phyto- materials available in natural waters must be deter- plankters on organic materials is not obligatory; it mined, but also the kinetics of uptake of organic is facultative heterotrophy. The importance of this compounds by the algae must be measured. Only way of growth depends greatly on the general nu- then can the laboratory studies be extrapolated to tritional conditions, and differs according to the natural environmental conditions to determine if the element given in organic form: carbon, nitrogen, algae can successfully compete with bacteria for or phosphorus. dissolved organic compounds. The first experiments made in this topic were not very encouraging. For example, Sloan and Strickland (1966) in their con- Carbon clusions pointed out that, even if it is possible to find some heterotrophic capability in Cyclotella HISTORICAL AND PHYSIOLOGICAL FEATURES cryptica and Thalassiosira rotula, it takes a lag phase for the algae to adjust, and the uptake is too From mud samples collected during the Gala- low, in the range of substrate concentrations in natural thea Expedition, in 1951, Wood (1956) found well- waters, to attach any ecological significance to the preserved benthic diatoms, belonging to autochtonous heterotrophy. Wright and Hobbie (1965) distin- populations and adapted to live at very great depths guished two different mechanisms in the uptake of (7000-10 000 m). Bernard (1958) observed some organic materials by organisms. One responds to Chrysomonads in the Mediterranean Sea in very an enzymatic transport system which is probably deep layers (between 250 and 2000 m) where ab- unique to bacteria. The second one is linear: uptalce solutely no light can penetrate. These organisms by passive diffusion increases proportionally with were in apparent good health. Obviously, in such increasing substrate concentration. It should be conditions, if growth can occur, it should be de- performed by organisms other than bacteria, such pendent totally on organic material. as phytoplankton. If so, obviously, the algae would

279 be strongly disadvantaged relative to bacteria for Undoubtedly, facultative heterotrophy is of ecological the uptake of such compounds at low concentrations. benefit to these diatoms when they settle out of the Somewhat later, Hellebust and Guillard (1967) photic layer into dimly lit or dark bottom waters, demonstrated that the transport of organic compounds or in muds, rich in organic materials. They can across the membrane of the cell is an active phe- remain viable there, even for long periods, until nomenon depending on the external concentration exposed to conditions suitable for photoautotrophic of the metabolite and responding to a Michaelis- growth. And these algae are able to reoccupy the Menten type equation. Using this approach, it has illuminated upper layers when the conditions become been possible to characterize the affinities of the cells better. On the other hand, instead of total hetero- for different substrates. The algae show different trophy, , a light-stimulated organic nutrient uptake characteristics in regard to this uptake. For instance, (photoheterotrophy) may confer a biological ad- Lewin and Hellebust (1975, 1976, 1978) demon- vantage to species living preferentially in dim light. strated that Nitzschia angularis , Nitzschia laevis , Bristol Roach (1928) already demonstrated that, for and Navicula pavillardi are able to utilize glutamate a soil alga exposed to a very attenuated light, auto- for heterotrophic growth but differ greatly in their trophic and heterotrophic uptakes are not exclusive heterotrophic capabilities with respect to glucose processes, but, on the contrary, may be cumulative. utilization: Navicula pavillardi is absolutely unable Even if we have observed that some centric diatoms to utilize glucose; Nitzschia angularis takes up are capable of such an heterotrophic growth, most glucose as a substrate, in the presence of glumatate species showing these capabilities belong to the as energy source; Nitzschia laevis can grow when pennate group. Quantitative measurements for other in presence of glucose as sole substrate. No lag phase algae are scarce and it is difficult to establish a goocl is observed before growth takes place after trans- estimate of the real influence of this uptake by un- ferring cells from autotrophic to heterotrophic con- studied species of algae relative to the uptake by ditions. This finding confirms the presence of a bacterial populations in the water column. transport system in light-grown cells. The authors With differential filtration and autoradiographic pointed out that it is not a rule and other algae be- methods, it has been possible to characterize the have differently. For instance, Cyclotella coptica uptake of labeled organic compounds by various and Cylindrotheca fusiformis show a lag phase of components of the pelagic ecosystem. The results about 2 d when transferred to the dark. There is a obtained at sea and in lakes point out that the fraction considerable variability in the characteristics of the that retains the most part of the radioactivity is of responses of these diatoms to organic substrates. a very small size (under 1-2 gm) and likely mainly Furthermore, the affinity of the uptake system in belongs to bacterial populations. Many works have Nitzschia laevis is rather high for glutamate (K = been published on this topic (Williams 1970; Allen 0.03 mM), glucose (K = 0.03 mM), and alanine 1971; Berman 1975; Azam and Hodson 1977; Berman (K = 0.02 mM), but fairly low for lactate (K = and Stiller 1977; Chrétiennot-Dinet and Vacelet 0.4 mM). The ability to take up organic carbon 1978). These results are not surprising because, as can vary also from dark to light exposure. For in- stated before, the half-saturation constants of algae stance, Hellebust (1971) showed that glucose uptake for uptalce of the organic compounds are usually by Cyclotella cryptica .is induced in the dark and high. Bennett and Hobbie (1972) mentioned with stopped during exposure to light. Various physical glucose, K = 5 mg I, in Chlamyclomonas sp.; factors may react on this uptake capability, not only this concentration is much higher than those observed the light intensity and the temperature, but also the in natural waters; consequently, such a transport salinity. A shift in the salt balance may modify the system cannot in general be effective under natural characteristics of the uptake. Thus, Hellebust (1978) conditions. the transport system for glucose demonstrate that However, the concentrations of all the naturally and amino acids is strongly related to the NaCI occurring organic compounds increase significantly relationship between uptake and concentration. The in peculiar situations: at the surface of organic aggre- NaCI concentration is hyperbolic. gates (Riley 1963), in muddy sediments near the bottom, or in, areas submitted to hypertrophication. ECOLOGICAL IMPLICATIONS The heterotrophic potential of the algae cannot be Acclimation to a shift in the energy source, neglected in these conditions, since there is often a light or organic carbon, can have considerable good correspondence between their heterotrophic ecological significance because it is more or less capabilities and the organic trophic level of the easy for the facultative heterotrophs. For example, ecosystem to which they belong. This has been White (1974) pointed out that Coscinodiscus sp. and observed more frequently in benthic diatoms Cyclotella cryptica, grown in total darkness for 1 yr, (Admiraal and Peletier 1979) and less often in phyto- keep their chlorophyll a and chlorophyll c contents. plankton. For instance, Mahoney and McLaughlin

280 (1977) found that dinoflagellates, abundant during photosynthetic apparatus for very long periods. These the blooms in the hypertrophicated New York Bay, observations lead the authors to assert that hetero- are highly versatile and use more than half the 21 trophic capability helps the cells to maintain good organic compounds tested, mainly sugars and de- health in aphotic conditions, but does not contribute rivatives. Obviously, the two dinoflagellates studied, significantly to the total production of the water Massartia rotundata and Prorocentrum micans , and column. the chrysophyte Olithodiscus luteus have a nutri- In conclusion, when organic compounds are tional advantage in this environment. Because these taken up by phytoplankters, they are used mainly organisms can reach densities over 200 million as additional energy sources and less frequently as cells per litre, it is possible that the CO2 becomes carbon source. Furthermore, it is doubtful that this a limiting factor. In fact, one observes an increase kind of heterotrophy plays a major role in the trophic of growth in these samples when bicarbonate is added. strategies of the greatest part of phytoplankters, in For these species, the use of organic carbon sources most natural environmental conditions. But, prin- would likely help to mitigate the CO2 deficiency. cipally in locally enriched areas, heterotrophy and The role of organic compounds as alternative carbon the photoheterotrophy might increase the survival sources has also been postulated in very soft water potential of organisms during periods of low light lakes, inadequatly buffered, where, at some period intensity and might give them an advantage relative of the day, the intensive photosynthesis brings the to their exclusive photoautotrophic competitors. pH abruptly to very high levels (Allen 1972). But, such improvements of growth by organic carbon compounds that act as a carbon source may be not Nitrogen very common and are mainly observed in very rich waters. As a matter of fact, it has been demonstrated INTRODUCTION AND GENERAL FEATURES that the availability of inorganic carbon from bicar- For many years, inorganic nitrogen was con- bonate complex is sufficient in most natural waters sidered the sole potential source for the algae in to be much less limiting than other nutrients (Gold- natural waters. This tendency was so strong that man et al. 1974; Caperon and Smith 1978). even the role of ammonia was omitted completely The real demonstration and the final proof of in the computation of the available forms of nitrogen heterotrophy in the natural environments, as stated for the primary production. But, obviously, the Vincent and Goldman (1980), require the following: metabolism of organic matter by animals, algae, and (i) analytical description of the ambient concen- bacteria releases in the medium many organic forms trations and rates of supply of usable substrates; of nitrogen and phosphorus, which are more or (ii) in situ evidence that algae have a high affinity less rapidly oxidized to nitrates and phosphates. transport system which enables them to take up When the Redfield's (1958) ratio nitrogen: phos- organic compounds at the in situ concentrations; phorus is close to 15:1, the conclusion that organic (iii) demonstration that these algae are biochemically forms do not interfere directly in plant growth could equipped to use these substrates for growth; (iv) be accepted. But, often in neritic waters and at the estimation of the relative contribution of dissolved end of the blooms in the open waters, and much organic substances versus other nutrient or energetic more commonly in lakes, this ratio can vary con- factors. To get simultaneous information on these siderably. Its variations are greater when only the four points is very difficult indeed. But the same amounts of nitrates and nitrites are taken into account authors, Vincent and Goldman (1980), demonstrated and not the other forms of nitrogen (Harris and Riley that, in an oligotrophic lake (Lake Tahoe, USA), 1956; Vaccaro 1963). Even if such organic com- it is possible to find two species, Monoraphidiutn pounds probably have only a slight influence on the contortum and Friedmannia sp., that are able to total production of the waters, it is difficult to deny take up organic compounds in the deeper layers of their role on phytoplankton populations, at least in the lake but are usually not abundant in the surface peculiar conditions. euphotic zone. Furthermore, these species are capable The concept of Dugdale and Goering (1967) of significant acetate transport, which varies with has been very useful in this respect. These authors environmental conditions and ranges from 25 to proposed the terms of "regenerated production" for 247% compared with the photosynthetic carbon up- the production due to rapidly recycled forms of take. But, at the bottom of the photic layer, the ratio of nitrogen in the upper layers, and "new production" the organic carbon utilization by total phytoplankton related to forms of nitrogen brought from the deep is only 1.5-15.0% of the photosynthetic carbon layers or from the atmosphere. When the hydro- uptake. On the other hand, in culture, these species logical conditions lead to an increase in the function have an extremely slow growth in the dark but are of regenerated production relative to new production, able to remain viable and to retain a fully operational the algae of natural assemblages that are able to

281 take up and assimilate organic forms of nitrogen (1962) and Antia et al. (1975) did not observe any will be helped in their growth, and this capability growth on glycine, the opposite results to that of could be an important factor in determining species Kapp et al. (1975). Liu and Hellebust (1974) observed succession. This role will be more important in successful growth of Cyclotella cryptica on urea oligotrophic waters where the concentrations of but Antia et al. (1977) did not. Obviously, all ex- inorganic forms of nitrogen are low. perimental conditions (light, temperature, salinity Indeed, algae are able to grow with organic strength) may react on the capacity for one strain nitrogen compounds as sole nitrogen source and this to grow on a given organic compound. (iii) Another information is very old. Ternetz (1912) was probably cause of discrepancies may lie in the fact that, for the first to give such results, using Euglena gracilis some species—organic compound associations, an Since then, these potentialities have been observed adaptation phase may occur when the alga is trans- in many algae (Syrett 1962; Van Baalen 1962; Guil- ferred from a medium with inorganic source to a lard 1963; Lewin 1963; McCarthy 1972a; Wheeler medium with an organic source. For instance, Antia et al. 1974; Antia et al. 1975, and many others). et al. (1975) observed a very long period of adaptation The aim of the earliest studies was often to charac- (18-23 d) with Prasinocladus marinus , Isochrysis terize biochemically some strains or species, to galbana , and Hemiselmis virescens cultivated on differentiate them by their growth patterns which hypoxanthine before growth began; afterwards, might be of some taxonomic utility. Each investigator these species appeared to utilize the purine as effi- chose his own set of experimental conditions, parti- ciently as they did ammonium or urea. On the other cularly the nitrogen source concentration. Unfortu- hand, it has been demonstrated that the capability nately, it was frequently 100-1000 times higher of taking up such compounds may vary as a function than the values usually found in natural waters, and of chemical conditions of the medium. North and the results may not reflect the behavior of phyto- Stephens (1967, 1969, 1971, 1972) mentioned plankters in their environment. Moreover, illumi- uptake of amino acids by Platynionas subcordifonnis nation conditions (intensity, quality, light—dark and Nitzschia (walls . They observed that a previous cycles) used in these experiments were also quite restriction of the nitrogen availability in the culture different. Because of that, these studies were not greatly accelerates the rate of amino acid uptake. always directly useful for the ecologist, even if As a matter of fact, it appears that the growth they offered a valuable insight into the metabolic potentials of the nitrogen compounds depend on potential of algae for utilization of various nitrogen the nutritional history of the cells. (iv) Furthermore, sources. Anyhow, it is possible to retain some the concentration of the nutrient plays a very im- general features. (i) The responses greatly differ with portant role in this capability of growth (Berland the species and, even within one species, with the et al. 1976, 1979; Antia et al. 1977). An increase strains. All organic compounds tested (urea, uric of the concentration enhances not only the frequency acid, amino acids, amino sugars, purines) do not of appearence of the potentiality, but also the growth show the same capability of growth for each species. rate for several species which develop poorly at For instance, among 26 species belonging to all lower concentrations. Sometimes, it is necessary to taxonomic groups present in the plankton, and grown reach very high levels (12.5 or 25 mM'L of on urea, glycine, glucosamine, and hypoxanthine, N compound) to observe any growth. But, such 88% showed a more or less good growth on urea, high values can be toxic for other strains and we 69% on hypoxanthine, 50% on glycine, and 42% have to be very cautious in extending such remarks on glucosamine (Antia et al. 1975). Obviously, urea to the ecological content, since these high concen- is a good potential nitrogen source for almost all trations are never found in natural waters. the species studied and this is pointed out by many In fact, some works demonstrate that many authors. Amino acids are used very differently by algae are able to effectively take up such compounds the different species and, in one species, by various as urea and amino acids at concentrations rather clones (Cain 1965). Usually, glycine, glutamic acid, close to the amounts existing in the waters. asparagine, and tryptophane are more frequently used than lysine, proline, phenylalanine and tyrosine AMINO ACIDS (data from Turner 1979). Pyrimidic and puric bases are less frequently taken up. But hypoxanthine is Wheeler et al. (1974) studied 25 strains culti- proved as a good source of nitrogen (Antia et al. vated on nine amino acids, and observed that 75% 1975). (ii) It is necessary to point out that the results of the strain—substrate combinations allowed a fairly may vary with the culture conditions, and conflicting good growth at a concentration of 1 p,g-at • of results are not unusual. For example, the cyanophyta N substrate. One conclusion was that, in most cases, Agmenellum quadruplicatum strain PR6 shows dif- species that could grow in culture on amino acids ferent behaviors in different studies. Van Baalen at high concentrations, could also take up amino

282 acids at low concentrations. Therefore, these results (1978) showed the existence of amino acid specific showed that the high concentrations frequently used carriers in various Chlorophyceae. These carriers in culture often provide valuable and useful infor- can be loaded with several different amino acids mation on the real uptake potentialities of the algae or other compounds like urea, that would compete - in their natural environment. More generally, it for the occupation of one site. In a given species, must be concluded that many phytoplankters have the carrier of one amino acid can be saturated by transport systems that allow them to accumulate another and then can transport this latter, but with and assimilate amino acids at concentrations com- a lower efficiency. This discovery is very important monly found in natural waters. Such patterns appear because it implies that different organic compounds preferentially in species that normally occur in could act synergistically to modify the response inshore and littoral habitats where amino acid con- observed with one compound alone; this synergistic centrations may be higher (Clark et al. 1972). action is not always positive. Likely, such systems Another question which has to be resolved exist in other groups of algae. Since the uptalce concerns the capability of these algae to take up systems may differ among species, the amino acid and assimilate such compounds when different ni- composition of the waters could preferentially en- trogen sources are present simultaneously in the hance the growth of some species in comparison medium. Early research on the subject led to the with the other components of natural phytoplankton conclusion that high concentrations of nitrate or assemblages. ammonia did not interfere with amino acid uptake Once taken up, all amino acids do not have the (North and Stephens 1971). But, further experiments same evolution inside the cells. Wheeler and Stephens gave opposite results. North and Stephens (1972) (1977) demonstrated that arginine, alanine, and demonstrated with Nitzschia ovalis that the transport lysine absorbed by Platymonas N-limited cells of all amino acids is repressed by high concentrations rapidly entered both anabolic and catabolic pathways. of nitrate in the medium. Likewise, Wheeler (1977) In N-nonlimited cells, all the soluble alanine, but with Platymonas subcordiformis pointed out that only a part of the arginine and lysine were available ammonia is the most efficient N source in the re- for protein systhesis, suggesting two different meta- pression of amino acid uptake. As the assimilation bolic pools of amino acids in the cells. In other of nitrate is inhibited by ammonia in the medium, words, all the amino acids seem not to be similarly the uptake of amino acids seems directly related to used for protein synthesis, but we do not know enough the presence of inorganic nitrogen sources. Such about these metabolic pathways to assert that there results are scarce, but if they were generalized, they is no difference, in qualitative needs for the amino would lead us to believe that amino acid uptake acids, between algae belonging to different taxonomic could only occur with nitrogen-limiting conditions, groups. contrary to what was earlier thought to be the case. This problem becomes much more complex if Stephens and North (1971) observed that the we consider all the organisms in the ecosystem algae are able to retain the amino moiety of the able to take up amino acids. Bacteria from fresh molecule after taking it up and to extrude the carbon and oceanic waters are the best equipped in this part. This negates the hypothesis of Algeus (1948) competition and develop very rapidly in culture, on that amino acids are extracellularly deaminated. almost all the amino acids. This is verified by size The mechanism of the amino acid uptake is really fractionation, microautoradiographic, and kinetics much more complex than initially expected. North studies in natural communities (Hobbie et al. 1968; and Stephens (1972) demonstrated by kinetic studies Williams 1970; Crawford et al. 1974; Paerl 1974; that Nitzschia ovalis possesses at least three amino Hollibaugh 1976; and others). The main argument acid uptake systems, specific for transporting acidic, for this assessment is that bacterial uptake of organic polybasic, and neutral amino acids. The efficiencies substrates proceeds more rapidly and at a lower of the three systems are not similar. Acidic and substrate concentrations than algal uptake. Further- neutral amino acids are taken up only by nitrogen- more, bacteria very rapidly adjust their uptake rates depleted cells at the end of growth in batch cultures. to high concentrations of amino acids by metabolic On the contrary, the polybasic amino acid sites adaptations or activation involving the synthesis of appear to be present throughout all phases of the inducible enzymes (Hollibaugh 1979). Very often, growth. Likewise, Hellebust (1970) found at least with the autoradiographic studies, it appears that three amino acid transport sites in the diatom Melosira the phytoplankton cells are not appreciably labeled. nummuloides . Because several different compounds On the contrary, other studies do deny a total lack can compete for one site when analogues are present of influence of the amino acids on the growth of in the medium, they can prevent the uptake of amino the phytoplankton in defined conditions. For exam- acids as Pedersen and Knudsen (1974) demonstrated ple, Schell (1974) described situations in marine in Chlorella fusca . In the same way, Kirk and Kirk waters, southeast of Alaska, where the uptake of

283 N- and C-labeled glycine and glutamic acid showed plankters. Healey (1977) observed similar values for utilization by natural assemblages at low but detect- K constants in freshwater algae Scenedesmus quadri- able rates compared with those of nitrate and ammonia. cauda and Pseudoanabaena catenata . Since the urea Consequently, the uptakes of the dissolved free concentrations in lakes and fresh waters are often amino acids may contribute to a small percentage higher than in the sea, it led the same author to of the nitrogen required by phytoplankton in surface assert that, in fresh waters, ammonia and urea are waters severely depleted in inorganic nitrogen. The the main nitrogen sources and that, most often, author even explains variations in turnover times nitrate and nitrite are not taken up by algae. of dissolved free amino acids by differences in the As observed with amino acids, bacteria can composition of the phytoplankton populations or compete with algae for urea uptake, but it has been by cell adaptations to more efficient use of organic demonstrated that, in eutrophic areas where the urea compounds. At the entrance of Newport Bay (USA), concentrations are high, the phytoplankters are Wheeler et al. (1977) observed the occurrence of responsible for the major part of its decomposition. activity in the nanoplankton size fraction, whereas Remsem et al. (1972) showed this in estuaries of only a slight uptake was coupled with the small the Georgia coast (USA) where the urea concen- particles of bacteria size. This fact could be explained tration can reach 8.9 I.L.A/ • In these brackish by the activity of the bacteria attached to big particles waters, the uptake by the cells is more efficient which would be able to retain a great part of the for marine algae than for freshwater ones, and higher amino acids; but, according to the authors, such an for the large cells than small ones. In coastal waters hypothesis is untenable in the water they studied off southern California, McCarthy (1972b) found and the role of uptake by nanoplankton cannot be that urea, particularly in the upper layers at stations neglected. where the amount of nitrates are very low, could constitute between 30 and 50% of the total nitrogen UREA used by natural assemblages. In a polluted bay of the The influence of urea on the phytoplankton Hawaian Islands, Harvey and Caperon (1976) gave average that ac- production has been easier to demonstrate. The such findings with an urea uptake For attention has been more easily focused on this com- counted for 53.5% of the total nitrogen uptake. pound because it has been possible to observe very several stations, this percentage can reach 100%. high concentrations of urea in several areas. Further- The real function of urea in the phytoplankton more, the enzymes urease and ATP: urea amidolyase production is not easy to appreciate because it depends (UAL-ase) used in the initial step in urea metabolism on the nature and concentrations of other nitrogen are well known (Roon and Levenberg 1968), whereas sources in the medium. McCarthy and Eppley's the mechanism of uptake of amino acid is still almost (1972) results on natural populations showed that unknown. It appeared that the two enzymes are not urea and ammonia uptakes occurred simultaneously, present together in the same alga (Leftley and Syrett but at different rates according to their relative con- 1973). Each organism has either one enzymatic centrations. Urea inhibits nitrate uptalce but at a activity or the other. According to the studies carried lower level than ammonia. In Chesapeake Bay, over out so far, it seems that Chlorophyceae contain 13 mo, McCarthy et al. (1977) always observed a UAL-ase and the other algae, urease. The published high phytoplankton preference for ammonia and data of K values with extracted urease are much urea over nitrate; urea was used after ammonia but higher than with UAL-ase (Syrett and Leftley 1976). before nitrate in order of preference. In another But, in organisms, it appears that the activities of experiment, Eppley et al. (1971) did not observe the two enzymes are very close and algae that possess a significant difference in the total crop of natural UAL-ase are not at such an advantage to take up assemblages obtained after 3 d in outdoor cultures the urea at low concentrations as would be initially supplied with nitrate, ammonia, or urea at the same expected. For example, McCarthy (1972a) demon- concentration. The average doubling rate of the cells strated that the marine diatoms Cyclotella nana, is almost similar with the three substrates. But the sp., Skeletonema Ditylum brightwelli , Lauderia behavior of the algae might differ from one species costatum , and Thalassiosira fluviatilis have urea to the other when in presence of urea and other uptake K values ranging from 0.42 to 1.70 tc,g-at • inorganic nitrogen sources. In fresh waters, Healey N. Furthermore, Skeletonema costatum and (1977) gave the results as follows: Scenedesmus Ditylum brightwelli have urea uptake K values not quadricauda and Pseudoanabaena catenata do not really different from those observed with ammonia. respond in the same way when they have both urea The fact that half-saturation constants for urea uptake and ammonia at their disposal. When the two sub- are equal or lower than the urea concentrations fre- strates are present together, the rate of the urea uptake quently measured in the oceans suggests that urea appears to dominate over ammonia uptake in P. should be a valuable nitrogen source for many phyto- catenata . On the contrary, S. quadricauda generally

284 takes up ammonia two or three times faster than 1980). Steemann Nielsen (1978) observed good urea. The introduction of ammonia to a S. quadri- growth of Selenastrum capricornutum at very low cauda culture severely depressed urea uptake whereas concentrations of ammonia. He pointed out that the introduction of urea caused a small depression such concentrations are very often observed in oligo- of ammonia uptake. It is the opposite in P. catenata. trophic parts of the oceans and he asserted that, in Even if all the uptake mechanisms are not yet these areas, it is wrong to conclude that nitrogen completely elucidated in natural waters, the utilization is the most limiting factor on the primary production of urea as a nitrogen source must be interpreted, only on the basis of the low concentrations of inor- in general, as a common and very important process ganic nitrogen. in natural environments, the opposite of amino acid Furthermore, algae have to compete with uptake. It might be possible that the presence of bacteria for phosphorus uptake. The role of bacteria urea in the medium not only allows the crop to in the consumption of inorganic phosphorus cannot reach higher levels, but also may significantly help be neglected. Harrison et al. (1977) pointed out some species to outgrow their competitors. In other that the fraction of inorganic phosphorus taken up words, we agree totally with the statement of Butler by bacteria was always important in Saanich Inlet, et al. (1979): "There is now considerable evidence British Columbia; at least 50% of the "2P uptake is that some species of phytoplankton utilize at least associated with particles under 1 ,um. For the size part of the dissolved organic nitrogen directly. It fraction over 1 ,u,m, the amount of phosphorus therefore seems a reasonable hypothesis that as NO,, fixed by bacteria seems more important than for is exhausted the phytoplankton population changes algae. Likewise, Paerl and Lean (1976), using the so that the species capable of utilizing other forms 32 P autoradiographic technique, found that, in lakes, of N become dominant." bacteria are more strongly labeled by "P-labeled Phosphorus orthophosphates than are algae. Even if it seems that, in certain conditions, the affinity towards orthophos- phates is lower for the bacteria than for the algae PHOSPHORUS AS A LIMITING FACTOR (Rhee 1972), there is a strong evidence that, in some FOR ALGAL GROWTH natural waters, bacteria can successfully compete The utilization of organic phosphorus com- with algae for the utilization of inorganic phosphorus. pounds has been studied with less attention than Otherwise, it is usually assumed that the depri- nitrogen organic materials. Generally, phosphorus vation of external phosphorus does not immediately has been considered sufficient, relative to the ni- damage the algal metabolism because of the internal trogen concentrations in seawater. On the contrary, pool of orthophosphates and polyphosphates stored in lakes, it frequently becomes the first limiting inside the cell. Effectively, the algal capacity to factor. Many works support this conclusion. For store internal phosphorus is usually great. It is also example, Healey and Hendzel (1980) measured observed that phosphorus starvation is followed by physiological indicators of nitrogen and phosphorus an increase of the phosphorus uptake after the re- deficiency of many lakes in the center of Canada, plenishment of the medium. In cells newly enriched and concluded that the lakes severely or moderately with phosphorus, the proportion of polyphosphates deficient in phosphorus are much more numerous in the cell can reach very high values (Aitchison than nitrogen-deficient ones. Several other authors and Butt 1973) and this form of phosphorus can give similar conclusions. contribute as much as 50% of the total cell phos- Recently, the position of several scientists on phorus (Perry 1976). Obviously, this reservoir dra- the impossibility of phosphorus limiting the pro- matically enhances the capability of the algae to duction in the sea has been reconsidered. Effectively, remain viable during phosphorus depletion and helps if ammonia and urea are included as available ni- them to compete successfully with bacteria in which trogen in the Redfield's (1958) N:P ratio, it often this property is less well developed. Such a demon- becomes much more higher than 15:1 and the role stration has been made experimentally by Rhee of phoeorus in the limitation of the primary pro- (1972) in comparing the growth of the green alga duction would be emphasized. An example is given Scenedesmus sp. with that of a strain of Pseudo- by Sander and Moore (1979) who observed a shift tnonas in different environmental conditions. The of the N:P atomic ratio from 9.8:1.0 to 28.8:1.0 algae also have the capability to take up polyphos- in the first 100 m of the water column near Barbados, phates in the medium. Such observations have often if ammonia is included. In some areas like the been made in cultures. Solorzano and Strickland Mediterranean Sea, the ratio frequently reaches (1968) mentioned it in Skeletonema costatum and values over 20 (McGill 1965), and there, at many Amphidiniwn carteri. Several authors also explain stations, it is possible to verify that phosphorus the high potential of the algae to keep their viability, becomes more limiting than nitrogen (Berland et al. even in phosphorus-deficient media, by a very rapid

285 turnover of the ambient phosphate (Fitzgerald and organic phosphorus are positively related demon- Nelson 1966). strates both the importance of the phosphatases in It is possible to believe that the total inorganic the uptake of DOP and the role of phosphomonoesters phosphorus may become limiting, at least in parti- as a supplementary source of phosphorus. Such cular ecological conditions, and consequently, that results showed the real potentiality of many algae organic compounds could be used as an alternative to utilize DOP, but, because of the high concen- source of phosphorus. trations used in the experiments, it was difficult to assert categorically that such utilization can really DISSOLVED ORGANIC MATERIALS AS AN occur in natural environments. ALTERNATIVE SOURCE OF PHOSPHORUS As a matter of fact, the concentration of DOP Steiner (1938) is the first to have foreseen is often very low in natural waters. In San Diego the possible ecological importance of these com- Harbor (USA) Solorzano and Strickland (1968) found pounds for phytoplankton growth. Later, Chu (1946) concentrations ranging from 0.12 to 0.70 iLg-at• demonstrated experimentally that some sources of In Chesapeake Bay, Taft et al. (1975, 1977) gave Seto dissolved organic phosphorus, as phytin and glyce- values between 0.12 and 0.53 itg-at • In the rophosphate, support a fairly good growth of Phaeo- Inland Sea (Japan), Matsuda et al. (1975) mentioned -1 . dactylum tricornutum. It was proved very quickly concentrations always lower than 0.75 ilg-at • L that many species of algae are capable of such growth In the Dutch Wadden Sea, De Jonge and Postma g-at • L-1 ; in in the absence of orthophosphates but in the presence (1974) indicated an average of 0.3 it of phosphoesters (Provasoli 1958). Therefore, it was these waters, the DOP can be present at higher con- assumed and verified later that algae, like bacteria, centrations than the inorganic phosphorus, particu- can produce a phosphatase that hydrolyzes the larly in summer. Kobori and Taga (1979) gave similar Japan. Moreover, all ester in, or on, the cell membrane, when the inorganic conclusions for some bays in not available at phosphorus becomes limiting. Kuenzler and Perras forms of inorganic phosphorus are waters (Lean 1973) (1965) demonstrated that this enzyme exists in 27 the same level. Data from fresh and Solôrzano 1966) show strains of algae belonging to Chrysophyceae, Bacil- and seawaters (Strickland contains lariophyceae, Cryptophyceae, Cyanophyceae, Dino- that two general types of DOP exist. One phyceae, and Chlorophyceae. They showed that the the high molecular weight compounds, slowly de- activity of the phosphatase could vary greatly ac- gradable and supposed to be mainly constituted of easily biodegradable cording to the species. Furthermore, the optimum nucleic acids. The second part, pH conditions for the highest enzymatic activity and attacked by the phosphatases, includes all the differs for each species. These results, with others phosphomonoesters. The latter is often found at dedicated to freshwater algae (Brandes and Elston relatively low concentrations regarding total DOP: 1956), were very provocative. As a matter of fact, 19% in Loch Creran, Scotland (Solôrzano 1978); since the activity of phosphatase, when present, could between 13 and 75% in Sagami Bay, Japan (Kobori easily degrad- vary according to species and environmental con- and Taga 1979). But the fact that the is usually at low concen- ditions, this capability might ensure the algae which able fraction of the DOP fraction has less possess these enzymes a decisive advantage over tration does not mean that this than the other species, when inorganic phosphorus concen- influence on the phytoplankton production of the DOP in two trations in the medium are low, compared with those more stable one. The separation easily, the other of the esters. fractions of inequal importance, one an excessive simpli- Kuenzler (1970) verified the influence of these slowly degradable, is probably artificial pro- compounds on the species succession in an experi- fication. As a matter of fact, using an mental ecosystem. He found that Cyclotella myptica, cedure (pH = 9; temperature = 37°C) Francko Phaeodactylum tricornutum, Dunedin tertiolecta, and Heath (1979), in two lakes in Ohio, found that Chlorella sp., Rhodomonas lens, Coccolithus 60-98% of the particulate phosphorus is hydrolyzed alkaline phosphatase. Ob- huxleyi, and Synechococcus sp. are capable of grow- to orthophosphate by the effectively ing on filtrate from cultures of other strains but with viously, the easily hydrolyzable fraction various success. The diatoms were the most efficient released by the organisms is much higher than that in the in the uptake of organic phosphorus. On the contrary, expected from concentrations usually found Dunaliella and Syneclzococcus, which have an in- water. ternal but no external phosphatase, and Chlorella The problem concerning the capability of algae and Rhodomonas, which have no efficient phos- to take up such phosphomonoesters is whether they phatase in the experimental conditions, cannot take are able to do it at very low concentrations. Taft et up most compounds released in the medium by the al. (1977) observed that in the Chesapeake Bay other algae. In these experiments, the fact that the natural assemblages have uptalce characteristics that amount of enzymes and the rate of uptake of dissolved permit utilization of phosphoesters at very low con-

286 centrations (K below 0.5 p.g-at • with glucose-6- b). In these conditions, if the inorganic phosphorus phosphate). In this uptake, the role of the bacteria is not completely exhausted in the medium, the algae seems to be less than 15% of the total absorption. are not phosphorus-depleted and the phosphatase This K value is close to those observed for the uptake measurement has no real significance. On the con- of the inorganic phosphorus by various species: trary, when the amounts of inorganic phosphorus K (tg-at • L-1 ) is 0.12 for Asterionella japonica are very small, it is impossible to ignore the potential (Thomas and Dodson 1968); 0.15 for Ceratiwn furca role of bacteria in the measurements. Using the (Qasim et al. 1973); 0.6 for Scenedestnus sp. (Rhee phosphatase activity as an index of the nutrient 1973); 0.6 for Thalassiosira pseudonana (Perry depletion of phytoplankton becomes more difficult. 1976); values ranging from 0.11 to 1.72 for natural populations in the Chesapeake Bay (Taft et al. 1975). It is possible to find direct and good correlations Such comparisons of K values allow us to assert between the chlorophyll a concentration and the without great risk that, in many algae, the organic activity of the alkaline phosphatase. Thus, in natural compounds may be hydrolyzed by the phosphatase populations, Taft et al. (1977) demonstrated that the at concentrations not too different from the natural size fraction ranging from 0.8 to 5 gm, which cor- ones. responds to the major part (78%) of the plant biomass, is the most efficient (70%) in the uptake of the glu- ALKALINE PHOSPHATASE ACTIVITY AS AN cose-6-phosphate. Obviously, in this case, organic INDEX OF PHYTOPLANKTON GROWTH ON phosphorus is essentially taken up by algae and not ORGANIC PHOSPHORUS by bacteria. But, generally there is no direct evidence for the simultaneous occurrence of phosphatase Another approach for investigating the limitation activity and esterphosphate uptake by algae. The of inorganic phosphorus and the potential utilization presence of the phosphatase only indicates that the by algae of DOP as monoesters is to measure the orthophosphate supply has dropped below a critical activity of the alkaline phosphatase in the plankton, level, which corresponds to a very high internal since the activity of this enzyme has been reported N:P ratio and, therefore, that additional pathways to be directly related to inorganic phosphorus limi- for phosphorus uptake are in use. Usually, the critical tation. High phosphatase activities are induced or value for the N:P ratio is about 30 (indications of activated by low inorganic phosphorus concentra- Rhee 1973 on Scenedesmus sp. and M011er et al. tions. Conversely, phosphatases are repressed or 1975 on Chaetoceros affinis). As a matter of fact, inhibited by high inorganic phosphorus concentra- the presence of the alkaline phosphatase in the plank- tions (Kuenzler and Perras 1965). Thus, it seems ton does not prove that the inorganic part of the possible to associate the presence of the enzyme dissolved phosphorus in the medium is utilized by with (i) a lack of available phosphorus, (ii) conse- algae. For example, Perry (1976) demonstrated that quently a limitation of the algal growth, and (iii) Thalassiosira pseudonana cultivated in a chemostat, an eventual utilization of the organic phosphorus on a medium made with inorganic, phosphorus-poor to compensate for the lack of inorganic phosphorus water sampled in the central North Pacific Ocean, in the environment. In fact, it is absolutely impossible did not use naturally occurring dissolved organic to assert that the presence of the enzyme is directly phosphorus in spite of the presence of phosphatase coupled with the utilization of phosphomonoesters at the surface of the cells. In this experiment, all by algae, and it now seems that the real significance natural organic phosphorus compounds in the water of the phosphatase has not always been adequately were not hydrolyzable by the phosphatase. On the evaluated in natural waters. other hand, it has been demonstrated that phosphatase One difficulty in interpreting the significance activity responds to diel rhythms. Rivkin and Swift of the presence of the enzyme is that bacteria are (1979) showed it in the dinoflagellate Pyrocystis also able to use phosphomonoesters in natural waters. noctiluca , where the activity of the enzyme is directly For example, the phosphatase-producing bacteria coupled to the light-dark cycles, suggesting an represent more than 40% of the total bacteria studied endogenous control. Indeed, the meaning of the in some bays in Japan (Kobori and Taga 1979), and phosphatase occurrence in natural phytoplankton the profiles of the enzyme activity are well corre- populations appears to be very complex. lated with bacteria densities at various depths. Even more, because the phosphatase activity is significantly Even if the metabolic action of the phosphatase correlated with the number of bacteria, and the is presèntly fairly well known in culture, "the actual number of bacteria directly related to the quantity role and importance of alkaline phosphatase in the of organic materials in the water, several ecologists mineral nutrition and ecology of marine phytoplank- use the phosphatase activity as an index of the trophic ton is not yet known." This sentence written by level of the environment, both in seawaters (Taga Perry (1972) is still valid. It is still difficult to ap- and Kobori 1978) and in fresh waters (Jones 1972a, preciate the real advantage in species competition

287 held by those algae able to take up organically bound Lake Kinneret by multiple isotopic labelling and dif- phosphorus. ferential filtration. Microb. Ecol. 3: 279-288. BERNARD, F. 1958. Données récentes sur la fertilité élé- mentaire en Méditerranée. Rapp. P.V. Cons. Cons. Acknowledgments Int. Explor. Mer 144: 103-108. BRANDES, D., AND R. N. ELSTON. 1956. An electron microscopical study of the histochemical localization We thank Dr Trevor Platt for revising the of alkaline phosphatase in the cell wall of Chlorella English manuscript and offering helpful suggestions. Mewls. Nature (London) 177: 274-275. BRISTOL ROACH, B. M. 1928. On the influence of light and of glucose on the growth of a soil alga. Ann. Bot. References London 42: 317-345. BUTLER, E. I., S. KNOX, AND M. I. LIDDICOAT. 1979. ADMIRAAL, W., AND H. PELETIER, 1979. Influence The relationship between inorganic and organic nu- of organic compounds and light limitation on the trients in sea water. J. Mar. Biol. Assoc. U.K. 59: growth rate of estuarine benthic diatoms. Br. Phycol. 239-250. J. 14: 197-206. CAIN, J. 1965. Nitrogen utilization in 38 freshwater AITCHISON, P. A., AND V. S. BUTT. 1973. The relation Chlamydomonas algae. Can. J. Bot. 43: 1367-1378. between the synthesis of inorganic polyphosphate and CAPERON, J., AND D. F. Sm ITH. 1978. Photosynthetic phosphate uptake by Chlorella vulgaris . J. Exp. Bot. rates of marine algae as a function of inorganic carbon 24: 497-510. concentration. Limnol. Oceanogr. 23: 704-708. ALGEUS, S. 1948. The deamination of glycocoll by green CHENG, J. Y., AND N. J. ANTIA. 1970. Enhancement algae. Physiol. Plant. 1: 382-386. by glycerol of phototrophic growth of marine plank- ALLEN, H. L. 1971. Dissolved organic carbon utilization tonic algae and its significance to the ecology of in size-fractionated algal and bacterial communities. glycerol pollution. J. Fish. Res. Board Can. 27: 335- Int. Rev. Gesamten Hydrobiol. 56: 731-749. 346.

1972. Phytoplankton photosynthesis, micro- CHRETIENNOT - DINET, M. J., AND E. VACELET. 1978. nutrient interactions, and inorganic carbon availability Séparation fractionnée du phytoplancton et estimation in a soft-water Vermont lake, p. 63-80. In Nutrients de l'assimilation autotrophe et hétérotrophe de "C. and eutrophication. Spec. Symp. Vol. I, Am. Soc. Oceanol. Acta 1: 407-413. Limnol. Oceanogr. CHU, S. P. 1946. The utilization of organic phosphorus ANTIA, N. J., B. R. BERLAND, D. J. BONIN, AND S. Y. by phytoplankton. J. Mar. Biol. Assoc. U.K. 26: MAESTRINI, 1975. Comparative evaluation of certain 285-295. organic and inorganic sources of nitrogen for photo- CLARK, M. E., G. A. JACKSON, AND W. J. NORTH. trophic growth of marine microalgae. J. Mar. Biol. 1972. Dissolved free amino acids in southern California Assoc. U.K. 55: 519-539. coastal waters. Limnol. Oceanogr. 17: 749-758. 1977. Effects of urea concentration in supporting COOKSEY, K. E., AND H. CHANSANG. 1976. Isolation growth of certain marine microplanktonic algae. Phy- and physiological studies on three isolates of Amphora cologia 16: 105-111. (Bacillariophyceae). J. Phycol. 12: 455-460. AZAM, F., AND R. E. HODSON. 1977. Size distribution CRAWFORD, C. C., J. E. HOBBIE, AND K. L. WEBB. and activity of marine microheterotrophs. Limnol. 1974. The utilization of dissolved free amino acids Oceanogr. 22: 492-501. by estuarine microorganisms. Ecology 55: 551-563. BENNETT, M. E., AND J. E. HOBBIE. 1972. The uptake DE JONGE, V. N., AND H. POSTMA. 1974. Phosphorus of glucose by Chhunydomonas sp. J. Phycol. 8: compounds in the Dutch Wadden Sea. Neth. J. Sea 392-398. Res. 8: 139-153. BERLAND, B. R., D. J. BONIN, O. GUERIN - ANCEY, DROOP, M. R. 1974. Heterotrophy of carbon, p. 530-559. AND N. J. ANTIA. 1979. Concentration requirement In W. D. P. Stewart [ed.] Algal physiology and bio- of glycine as nitrogen source for supporting effective chemistry. Blackwell Scientific Publications, Oxford. growth of certain marine microplanktonic algae. Mar. DUGDALE, R. C., AND J. J. GOERING. 1967. Uptake Biol. 55: 83-92 , of new and regenerated forms of nitrogen in primary BERLAND, B. R., D. J. BONIN, AND S. Y. MAESTRINI. productivity. Limnol. Oceanogr. 12: 196-206. 1980. Azote ou phosphore? Considérations sur le EPPLEY, R. W. ET AL. 1971. Phytoplankton growth and "paradoxe nutritionnel" de la mer Méditerranée. composition in shipboard cultures supplied with nitrate, Oceanol. Acta 3: 135-142. ammonium, or urea as the nitrogen source. Limnol. BERLAND, B. R., D. J. BONIN, S. Y MAESTRINI, M. L. Oceanogr. 16: 741-751. LIZÀRRAGA-PARTIDA, AND N. J. ANTIA. 1976. FITZGERALD, G. P., AND T. C. NELSON. 1966. Extrac- The nitrogen concentration requirement of D-gluco- tive and enzymatic analyses for limiting or surplus samine for supporting effective growth of marine phosphorus in algae. J. Phycol. 2: 32-37. microalgae. J. Mar. Biol. Assoc. U.K. 56: 629-637. FRANCKO, D. A., AND R. T. HEATH. 1979. Functionally BERMAN, T. 1975. Size fractionation of natural aquatic distinct classes of complex phosphorus compounds populations associated with autotrophic and hetero- in lake water. Limnol. Oceanogr. 24: 463-473. trophic carbon uptalce. Mar. Biol. 33: 215-220. GOLDMAN, J. C., W. J. OSWALD, AND D. JENKINS. BERMAN, T., AND M. STILLER. 1977. Simultaneous 1974. The kinetics of inorganic carbon limited algal measurement of phosphorus and carbon uptake in growth. J. Water Pollut. Control Fed. 46: 554-574. 288 GUILLARD, R. R. L. 1963. Organic sources of nitrogen KAPP, R., S. E. STEVENS JR., AND J. L. Fox. 1975. for marine centric diatoms, p. 93-104. In C. H. A survey of available nitrogen sources for the growth Oppenheimer red.] Symposium on marine micro- of the blue-green alga, Agmenellum quadruplicatwn. biology. C. C. Thomas, Springfield, IL. Arch. Microbiol. 104: 135-138. HARRIS, E., AND G. A. RILEY. 1956. Oceanography KIRK, D. L., AND M. M. KIRK. 1978. Amino acid and of Long Island Sound, 1952-1954. VIII. Chemical urea uptake in ten species of Chlorophyta. J. Phycol. composition of the plankton. Bull. Bingham. Ocea- 14: 198-203. nogr. Coll. 15: 315-323. KOBORI, H., AND N. TAGA. 1979. Phosphatase activity HARRISON, W. G., F. AZAM, E. H. RENGER, AND R. W. and its role in the mineralization of organic phos- EPPLEY. 1977. Some experiments on phosphate phorus in coastal sea water. J. Exp. Mar. Biol. Ecol. assimilation by coastal marine plankton. Mar. Biol. 36: 23-39. 40: 9-18. KUENZLER, E. J. 1970. Dissolved organic phosphorus HARVEY, W. A., AND J. CAPERON. 1976. The rate of excretion by marine phytoplankton. J. Phycol. 6:7-13. utilization of urea, ammonium, and nitrate by natural KUENZLER, E. J., AND J. P. PERRAS. 1965. Phospha- populations of marine phytoplankton in a eutrophic tases of marine algae. Biol. Bull. Woods Hole 128: environment. Pac. Sci. 30: 329-340. 271-284. HASTINGS, J. L., AND W. H. THOMAS. 1977. Qualitative LEAN, D. R. S. 1973. Movements of phosphorus between requirements and utilization of nutrients: algae, p. its biologically important forms in lake water. J. Fish. 87-163. In M. Rechcigl Jr. [cd.] Handbook series Res. Board Can. 30: 1525-1536. in nutrition and food, section D: Nutritional requi- LEFTLEY, J. W., AND P. J. SYRETT. 1973. Urease and rements. Vol. 1. CRC Press, Cleveland, Ohio. ATP: urea amidolyase activity in unicellular algae. HEALEY, F. P. 1977. Ammonium and urea uptake by J. Gen. Microbiol. 77: 109-115. some freshwater algae. Can. J. Bot. 55: 61-69. HEALEY, F. P., AND L. L. HENDZEL. 1980. Physiological LEWIN, J. C. 1953. Heterotrophy in diatoms. J. Gen. indicators of nutrient deficiency in lake phytoplankton. Microbiol. 9: 305-313. Can. J. Fish. Aquat. Sci. 37: 442-453. 1963. Heterotrophy in marine diatoms, p. 229- HELLEBUST, J. A. 1970. The uptake and utilization of 235. In C. H. Oppenheimer [cd.] Symposium on organic substances by marine phytoplankters, p. 223- marine microbiology. C. C. Thomas, Springfield, 256. In D. W. Hood [ed.] Symposium on organic IL. matter in natural waters. Inst. Mar. Sci., Univ. Alaska, LEWIN, J., AND J. A. HELLEBUST. 1975. Heterotrophic Occas. Publ. No. 1. nutrition of the marine pennate diatom Navicula 1971. Glucose uptake by Cyclotella cryptica: dark pavillardi Hustedt. Can. J. Microbiol, 21: 1335-1342. induction and light inactivation of transport system. 1976. Heterotrophic nutrition of the marine J. Phycol. 7: 345-349. pennate diatom Nitzschia angularis var. affinis. Mar. 1978. Uptake of organic substrates by Cyclotella Biol. 36: 313-320. cryptica (Bacillariophyceae): effects of ions, iono- 1978. Utilization of glutamate and glucose for phores and metabolic and transport inhibitors. J. heterotrophic growth by the marine pennate diatom Phycol. 14: 79-83. Nitzschia laevis . Mar. Biol. 47: 1-7. HELLEBUST, J. A., AND R. R. L. GUILLARD. 1967. Liu, M. S., AND J. A. HELLEBUST, 1974. Uptake of Uptalce specificity for organic substrates by the marine amino acids by the marine centric diatom Cyclotella diatom Melosira nummuloides . J. Phycol. 3: 132-136. cryptica . Can. J. Microbiol. 20: 1109-1118. HELLEBUST, J. A., AND J. LEWIN. 1972. Transport MAHONEY, J. B., AND J. J. A. MCLAUGHLIN. 1977. systems for organic acids induced in the marine pennate The association of phytoflagellate blooms in lower diatom Cylindrotheca fusiformis . Can. J. Microbiol. New York Bay with hypertrophication. J. Exp. Mar. 18: 225-233. Biol. Ecol. 28: 53-65. 1977. Heterotrophy, p. 169-197. In D. Werner MCCARTHY, J. J. 1972a. The uptake of urea by marine [ed.] The biology of diatoms. Blackwell Scientific phytoplankton. J. Phycol. 8: 216-222. Publications, Oxford. 19726. The uptake of urea by natural populations HOBBIE, J. E., C. C. CRAWFORD, AND K. L. WEBB. of marine phytoplankton. Limnol. Oceanogr. 17: 1968. Amino acid flux in an estuary. Science (Wash- 738-748. ington, D.C.) 159: 1463-1464. MCCARTHY, J. J., AND R. W. EPPLEY. 1972. A comp- HOLLIBAUGH, J. T. 1976. The biological degradation rison of chemical, isotopic, and enzymatic methods of arginine and glutamic acid in seawater in relation for measuring nitrogen to the growth of phytoplankton. Mar. Biol. 36: 303- assimilation of marine phyto- 312 . plankton. Limnol. Oceanogr. 17: 371-382. 1979. Metabolic adaptation in natural bacterial MCCARTHY, J. J., W. R. TAYLOR, AND J. L. TAFT, populations supplemented with selected amino acids. 1977. Nitrogenous nutrition of the plankton in the Estuarine Coastal Mar. Sci. 9: 215-230. Chesapeake Bay. I. Nutrient availability and phyto- JONES, J. G. I972a. Studies on freshwater bacteria-asso- plankton preferences. Limnol. Oceanogr. 22: 996- ciation with algae and alkaline phosphatase activity. 1011. J. Ecol. 60: 59-75. McGILL, D. A. 1965. The relative supplies of phosphate, 1972b. Studies on freshwater microorganisms- nitrate and silicate in the Mediterranean Sea. Comm. phosphatase activity in lakes of differing degrees of Int. Explor. SCi. Mer. Medit., Rapp. P.V. Réun. eutrophication. J. Ecol. 60: 777-791. 18: 737-744.

289 MOLLER, M., S. MYKLESTAD, AND A. HAUG. 1975. RIVKIN, R. B., AND E. SWIFT. 1979. Diel and vertical Alkaline and acid phosphatases of the marine diatoms patterns of alkaline phosphatase activity in the oceanic Chaetocaros affinis var. ivillei (Gran) Hustedt and dinoflagellate Pyrocystis noctiluca Limnol. Oceanogr. Skeletonema costanun (Grey.) Cleve. J. Exp. Mar. 24: 107-116. Biol. Ecol. 19: 217-226. ROON, R. J., AND B. LEVENBERG. 1968. An adenosine MORRILL, L. C., AND A. R. LOEBLICH III. 1979. An triphosphate-dependent, avidin-sensitive enzymatic investigation of heterotrophic and photoheterotrophic cleavage of urea in yeast and green algae. J. Biol. capabilities in marine Pyrrhophyta. Phycologia 18: Chem. 19: 5213-5215. 394-404. SANDER, F., AND E. MOORE. 1979. Significance of MURRAY, D. R., J. GIOVANELLI, AND R. M. SMILLIE. ammonia in determining the N:P ratio of the sea water 1970. Photoassimilation of glycolate, glycine and off Barbados, West Indies. Mar. Biol. 55: 17-21. serine by Euglena gracilis . J. Protozool. 17: 99-104. SCHELL, D. M. 1974. Uptalce and regeneration of free NEILSON, A. H., AND R. A. LEWIN. 1974. The uptake amino acids in marine waters of southeast Alaska. and utilization of organic carbon by algae: an essay Limnol. Oceanogr. 19: 260-270. in comparative biochemistry. Phycologia 13: 227-264. SLOAN, P. R., AND J. D. H. STRICKLAND. 1966. Hetero- NORTH, B. B., AND G. C. STEPHENS. 1967. Uptake and trophy of four marine phytoplankters at low substrate assimilation of amino acids by Platymonas . Biol. concentrations. J. Phycol. 2: 29-32. Bull. Woods Hole 133: 391-400. SOLÔRZANO, L. 1978. Soluble fractions of phosphorus 1969. Dissolved amino acids and Platymonas compounds and alkaline phosphatase activity in Loch nutrition. Proc. Int. Seaweed Symp. 6: 263-273. Creran and Loch Etive, Scotland. J. Exp. Mar. Biol. 1971. Uptake arid assimilation of amino acids Ecol. 34: 227-232. by Platymonas . II. Increased uptake in nitrogen- SOLÔRZANO, L., AND J. D. H. STRICKLAND. 1968. deficient cells. Biol. Bull. Woods Hole 140: 242-254. Polyphosphate in seawater. Limnol. Oceanogr. 13: 1972. Amino acid transport in Nitzschia ovalis 515-518. Amott. J. Phycol. 8: 64-68. STEEMANN NIELSEN, E. 1978. Growth of plankton algae PAERL, H. W. 1974. Bacterial uptake of dissolved organic as a function of N-concentration, measured by means matter in relation to detrital aggregation in marine and of a batch technique. Mar. Biol. 46: 185-189. freshwater systems. Limnol. Oceanogr. 19: 966-972. STEINER, M. 1938. Zur Kenntnis des Phosphatkreislaufes PAERL, H. W., AND D. R. S. LEAN. 1976. Visual obser- vations of phosphorus movement between algae, in Seen. Naturwissenchaften 26: 723-724. bacteria, and abiotic particles in lake waters. J. Fish. STEPHENS, G. C., AND B. B. NORTH. 1971. Extrusion Res. Board Can. 33: 2805-2813. of carbon accompanying uptake of amino acids by marine phytoplankters. Limnol. Oceanogr. 16: 752- PEDERSEN, A. G., AND G. KNUTSEN. 1974. Uptake of L-phenylalanine in synchronous Chlorella fusca 757. Characterization of the uptake system. Physiol. Plant. STRICKLAND, J. D. H., AND L. SOLÔRZANO. 1966. 32: 294-300. Determination of monoesterase hydrolysable phosphate PERRY, M. J. 1972. Alkaline phosphatase activity in sub- and phosphomonoesterase activity in sea water, p. tropical Central North Pacific waters using a sensitive 665-674. In H. Barnes [ed.] Some contemporary fluorometric method. Mar. Biol. 15: 113-119. studies in marine science. Allen and Unwin, London. 1976. Phosphate utilization by an oceanic diatom SYRETT, P. J. 1962. Nitrogen assimilation, p. 171-188. in phosphorus-limited chemostat culture and in the In R. A. Lewin [cd.] Physiology and biochemistry oligotrophic waters of the Central North Pacific. of algae. Academic Press, New York, NY. Limnol. Oceanogr. 21: 88-107. SYRETT, P. J., AND J. W. LEFTLEY. 1976. Nitrate and PROVASOLI, L. 1958. Nutrition and ecology of protozoa urea assimilation by algae, p. 221-234. hi N. Sun- and algae. Annu. Rev. Microbiol. 12: 279-308. derland [ed.] Perspectives in experimental biology, QASIM, S. Z., P. M. A. BHATTATHIRI, AND V. P. Vol. 2: Botany. Pergamon Press, Oxford. DEVASSY. 1973. Growth kinetics and nutrient re- TAFT, J. L., M. E. LOFTUS, AND W. R. TAYLOR. 1977. quirements of two tropical marine phytoplankters. Phosphate uptake from phosphomonoesters by phyto- Mar. Biol. 21: 299-304. plankton in the Chesapeake Bay. Limnol. Oceanogr. REDFIELD, A. C. 1958. The biological control of chemical 22: 1012-1021. factors in the environment. Am. Sel. 46: 205-221. TAFT, J. L., W. R. TAYLOR, AND J. J. MCCARTHY. REMSEN, C. C., E. J. CARPENTER, AND B. W. SCHROE- 1975. Uptake and release of phosphorus by phyto- DER. 1972. Competition for urea among estuarine plankton in the Chesapeake Bay estuary, USA. Mar. microorganisms. Ecology 53: 921-926. Biol. 33: 21-32. RHEE, G.-Y. 1972. Competition between an alga and an TAGA, N., AND H. KOBORI. 1978. Phosphatase activity aquatic bacterium for phosphate. Limnol. Oceanogr. in eutrophic Tokyo Bay. Mar. Biol. 49: 223-229. 17: 505-514. TERNETZ, C. 1912. Beitrâge zur Morphologie und Physio- 1971 A continuous culture study of phosphate logie der Euglena gracilis Klebs. Jahr. Wiss. Bot. uptake, growth rate and polyphosphate inSceitedesmus 51: 435-514. sp. J. Phycol. 9: 495-506. THOMAS, W. H., AND A. N. DODSON. 1968. Effects of RILEY, G. A. 1963. Organic aggregates in sea water and phosphate concentration on cell division rates and yield the dynamics of their formation and utilization. of a tropical oceanic diatom. Biol. Bull. Woods Hole Limnol. Oceanogr. 8: 372-381. 134: 199-208.

290 TURNER, M. F. 1979. Nutrition of some marine microalgae WHEELER, P. A., B. B. NORTH, AND G. C, STEPHENS. with special reference to vitamin requirements and 1974. Amino acid uptake by marine phytoplankters. utilization of nitrogen and carbon sources. J. Mar. Limnol. Oceanogr. 19: 249-259. Biol. Assoc. U.K. 59: 535-552. WHEELER, P. A., AND G. C. STEPHENS. 1977. Metabolic VACCARO, R. F. 1963. Available nitrogen and phosphorus segregation of intracellular free amino acids in Platy- and the biochemical cycle in the Atlantic off New monas (Chlorophyta). J. Phycol. 13: 193-197. England. J. Mar. Res. 21: 284-301. WHITE, A. W. 1974. Growth of two facultatively hetero- VAN BAALEN, C. 1962. Studies on marine blue-green trophic marine centric diatoms. J. Phycol. 10: 292— algae. Bot. Mar. 4: 129-139. 300. VINCENT, W. F., AND C. R. GOLDMAN. 1980. Evidence WILLIAMS, P. J. LE B. 1970. Heterotrophic utilization for algal heterotrophy in Lake Tahoe, Califomia- of dissolved organic compounds in the sea. I. Size Nevada. Limnol. Oceanogr. 25: 89-99. distribution of population and relationship between WHEELER, P. A. 1977. Effect of nitrogen source on respiration and incorporation of growth substrates. J. Platymonas (Chlorophyta) cell composition and amino Mar. Biol. Assoc. U.K. 50: 859--870. acid uptake rates. J. Phycol. 13: 301-308. WOOD, E. J. 1956. Diatoms in the ocean deeps. Pac. Sci. WHEELER, P., B. NORTH, M. LITTLER, AND G. STE- 10: 377-381. PHENS. 1977. Uptake of glycine by natural phyto- WRIGHT, R. T., AND J. E. HOBBIE. 1965. The uptake plankton communities. Limnol. Oceanogr. 22: 900— of organic solutes in lake water. Limnol. Oceanogr. 910. 10: 22-28.

291 Some Processes and Physical Factors that Affect the Ability of Individual Species of Algae to Compete for Nutrient Partition

D. J. BONIN AND S. Y. MAESTRINI

Station marine d'Endoutne, Chemin de la Batterie des Lions, F - I3007 Marseille, France

AND J. W. LEFTLEY Dunstaffnage Marine Research Laboratory, P.O. Box 3, Oban, Argyll PA34 4AD, Scotland

Introduction of metres, or a few metres, or less in areas where light-absorbing particles are especially dense. Fur- Optimal algal anabolism requires nutrient suf- thermore, nutrients are also vertically distributed; ficiency, optimal illumination, and temperature; on usually nutrient concentrations decrease from the the other hand, there are extreme values at which surface to the bottom of the oceans and deep lakes. growth stops. At intermediate levels, algae are as- Ultimately, , both spatial and temporal distributions sumed to grow according to parameters established of energy source, nutrient supply, and temperature on the basis of one-species and one-factor experiments may vary so that they are differentially favorable. (see reviews by Williams 1971; Soeder and Stengel For instance, nutrient-rich waters are often cold and 1974; Yentsch 1974; Finenko 1978). However, in nutrient-poor waters well illuminated; but algae have an algal cell, all single-factor effects are combined to react to variations in all the physical and chemical to give an ultimate integrated result: the actual growth. parameters at the same time. The way by which These three parameters (nutrients, light, and tem- an alga balances the different effects is species de- perature) vary both spatially and temporally and also pendent and, therefore, helps us understand why govern simultaneously the species succession; there- some species dominate and regularly succeed others. fore, it is impossible to discriminate completely The importance of the interactions between the between their respective effects. However, in spite effects of light intensity, temperature, and nutrient of the fact that considering them separately has concentrations on algal growth rate have been dis- obvious weaknesses, there is practically no other cussed (Eppley and Strickland 1968; Middlebrooks way to discuss their roles. and Porcella 1971; Di Toro et al. 1971; MacIsaac and Dugdale 1972) and modeled (Kiefer and Enns Field observations indicate that temperature 1976; Platt et al. 1977; Kremer and Nixon 1978; the horizontal distribution of many species controls Nyholm 1978). But recent observations may render algae; convincing evidence has been gathered for of obsolete the early concepts as it has been shown that freshwater phytoplankton (Hutchinson 1967) and motile algae do not conform to the models established marine phytoplankton (Raymont 1963; Gessner 1970; on the sole basis of fixed kinetic constants recorded Guillard and Kilham 1977). Moreover, experimen- with single species in laboratory cultures; such algae tally recorded ranges of temperatures suitable for represent a fascinating adaptation to the competition species growth indicate that natural phytoplankton for life. is usually well adapted to the local conditions and that death may occur when these conditions are experimentally modified (Hulburt and Guillard Temperature 1968). However, temporal fluctuations over a broad range of temperature, especially in temperate waters, TEMPERATURE-GOVERNED SPECIES may be such that the local species have to adapt and develop under rapidly changing conditions. As such SUCCESSIONS populations show the most marked and characteristic Whereas physiological adaptations for main- species successions, one might suspect that temper- taining the cell machinery at its highest efficiency ature controls these successions. On the other hand, when nutrients and light are not optimum have been light is the energy source of photosynthesis and frequently studied and discussed (see below), the subsequently that of the whole food chain. Day length ways by which algae adapt to grow at different and light intensity are largely dependent on the lati- temperatures remain often unclear, although resist- tude and climate. Moreover, light is progressively ance to extreme and lethal temperatures has been the absorbed by the water with depth, so that the so- subject of much research. Frequently the role of called euphotic zone may be spread over some tens temperature is expressed simply by ranges of values

292 within which responses are maximal, or possible (e.g. the algal photosynthetic ability at higher tempera- Jitts et al. 1964). Nevertheless, an imperfect overlap tures, but does not enhance the ability to photosyn- of temperature ranges suitable for growth can govern thesize at low temperatures. "The generally accepted the competition and/or the temporal succession of • hypothesis of temperature adaptation by algae is some species. Such an effect is, for instance, that of untenable" concluded Morris and Glover. Thus, a the tropical dinoflagellate Gymnodinium breve and simple equation could be written by Eppley (1972) the cyanophyte Gomphosphaeria aponina; the dino- to describe the kinetics of maximum expected growth flagellate, the growth of which is inhibited by external rate versus temperature lower than 40°C. metabolites released by the cyanophyte, blooms at relatively low temperatures, i.e. 22-23°C, because the range of temperature suitable for its own growth TEMPERATURE AND NUTRIENTS (17-30°C) extends towards lower values than that Assuming the validity of the Eppley model for (24-29°C) of its antagonistic companion species (Eng- all species would mean that temperature would govern Wilmot et al. 1977). Likewise, the dominance of the competition and species succession only through the flagellate Olisthodiscus luteus during summer blooms different specific ranges of temperature suitable for in Narragansett Bay has been ascribed recently to growth, as predicted by many other models (e.g. temperature (Tomas 1980; for more details see Goldman and Caipenter 1974; Kiefer and Enns 1976; Maestrini and Bonin 1981). Goldman and Ryther Kremer and Nixon 1978); this is denied by various (1976) reported that the diatom Skeletonema costatum observations. As a matter of fact, Goldman and dominates outdoor continuous mass culture at tem- associates (Goldman 1977, 1979; Goldman and peratures below 10°C whereas Phaeodactylum tri- McCarthy 1978; Goldman and Mann 1980) dem- cornutum outgrows and replaces this species at higher onstrated that temperature strongly influences cellular temperatures. This displacement of naturally domi- chemical composition and that each species responds nant species by high-temperature-adapted algae in somewhat differently. At the lowest temperature power plant discharge plumes is also well established at which growth occurs , Dunaliella tertiolecta , (Briand 1975). In this way, cyanophytes are often Phaeodactylwn tricornutum , Skeletonema costatum , regarded as organisms favored by high temperature and Pavlova luth cri increase their cellular nitrogen because they are the only oxygen-evolving photosyn- content, while the particulate organic nitrogen (PON) thetic organisms occurring in hot springs (Brock per cell remains constant in Thalassiosira pseudo- 1967) and because they are more abundant in tropical nana . Yoder (1979) reported conflicting results in than in temperate waters (Fogg et al. 1973). showing that temperature does not affect the cellular content of POC and PON in the diatom Skeletonema PHYSIOLOGICAL ADAPTATIONS TO costatunt , but influences the chlorophyll a content. SUBOPTIMAL TEMPERATURES Because cell division rates and nutrient uptake rates It is evident that temperature governs algal are uncoupled with respect to temperature, and growth and species competition; but this statement because extreme and optimal temperatures are species does not establish whether algal response to tem- related, there is evidence that nutrient uptalce rates are perature is direct. not equally temperature-governed in all species. Early work carried out with the diatom Skele- As a matter of fact, in competition experiments tonema costatum prompted JOrgensen and Steemann by Goldman and co-workers, S. costatum was the Nielsen (see JOrgensen and Steemann Nielsen 1965; most efficient species at assimilating inorganic JOrgensen 1968; Steemann Nielsen and JOrgensen nitrogen and was the dominant species at 10°C 1968a, b) to suggest that phytoplankton adapt to whereas D. tertiolecta dominated the other algae at nonoptimal temperature by increasing their enzyme 30°C. In another outdoor mass culture, the routine content to permit the rate of photosynthesis to remain S. costatum dominance at winter low temperatures constant. Due to the increase in the amount of all was prevented by an excess of nitrogen and phos- enzymes at low temperatures, the total amount of phorus while silicon was lacking, and led to a per- organic matter per cell also increases. This pioneer manent dominance of P. tricot-nutum through the statement was first supported by Morris and Farrell whole range of temperatures. Finally, Goldman (1971) who demonstrated that the green alga Duna- hypothesized that algae would be separated into a liella tertiolecta behaves like S. costatum; a real group of fast-growth species versus temperature, adaptation to resist unfavorable temperatures seemed which includes S. costatum and Thalassiosira to have been confirmed. But, a further study made pseudonana , while other species, such as Pavlova on Phaeodactylum tricornututn , Nitzschia closterium , lutheri , would comprise a group of algae with a low and D. tertiolecta led Morris and Glover (1974) growth rate versus temperature. The first group to question the previous statement. Their experiments would correspond to the curve of Eppley (1972) demonstrated that growth at low temperatures reduces whereas the second one would not obey the model,

293 especially at low temperatures. Other research, unadapted cells. For instance, from 25 to 10°C, summarized by Li (1981), also indicated that the cells adapted to lower temperatures exhibit enhanced processes of cell division, production, and accumu- photosynthesis at that temperature, compared with lation of some cell materials, for instance protein cells adapted to high temperatures. At temperatures and POC, are affected differently according to the below 10°C, adapted cells do not exhibit enhanced different species. Conversely, temperature alterations photosynthesis but have a reduced optimal photo- of chlorophyll a and PON contents might be similar synthetic capability. On the other hand, Hitchcock in all species. (1980) found that the temperature dependence of Whether and/or how temperature changes the growth rate is similar at light-limiting or light-sat- nutrient uptake rate is poorly documented, probably urated levels for the oceanic isolate Thalassiosira because the growth kinetic parameters, such as K pseudonana 13-1, while for other neritic species, and are implicitly considered to be species such as Detomda confervacea Skeletonetna costa- constants. The early work by McCombie (1960) nun , and Thalassiosira nordenskioldii , the changes showed that the temperature optimum for growth in division rates along with increasing temperatures of the freshwater alga Chlamydomonas reinhardii differ when cells are adapted to high light intensities. decreases when the nutrient concentration increases, Indeed, all these insights lead to the opinion whereas Maddux and Jones (1964) reported that that the role of temperature in phytoplankton growth such an increase of nutrient supply widens the opti- and succession is not as simple as previously assumed, mum temperature range of the marine algae Nitzschia and further complexity may be expected as research closterium and Tetraselmis sp. Otherwise, it was progresses. repotted that, at low temperatures, higher phosphate concentrations are required to produce the same biomass of freshwater algae (Borchardt and Azad Illumination 1968). Further research by Eppley et al. (1969) and Reports that light is the major factor governing Thomas and Dodson (1974) showed that the lower algal species competition have been hitherto rare. the temperature, the lower the K constant for nitrogen More often the authors' attention has been focused uptake in Skeletonema costatum , Dunaliella sp., and on specific metabolic adaptations, with special ref- Gymnodinium splendens; Ahlgren's (1978) results erence to pigment composition and the photosynthetic with the cyanophyte Oscillatoria agardhii were rate of the total crop. Species competition has been similar. Hence, cells growing at higher temperatures a secondary and subsequent topic of interest; hence show a higher requirement for the limiting nutrient. the literature cited here on the subject is rather short. Nutrient-rich waters such as winter waters or newly upwelled marine waters are generally cold; these waters become warmer and nutrient-depleted while LIGHT-SHADE AND LIGHT-SUN the algae grow. Therefore, the result is that algal PHYSIOLOGICAL ADAPTATIONS IN nutrient requirements increase when available nu- PHYTOPLANKTON SPECIES: trients become scarce. The algae which profit in MAIN EACKGROUND 3 these conflicting situations are those able to migrate vertically and so take advantage of the beneficial Early observations made with several species effects of the two kinds of waters (see below). of the dinoflagellate Ceratium led Steemann Nielsen (1954, in Ryther and Menzel 1959) to remark that COMBINED TEMPERATURE some species are characteristically found in surface AND LIGHT EFFECTS waters whereas others are confined below the euphotic zone of the oceans (100 m in his example). On this The combination of light and temperature effects basis, he termed "sun" species those growing in on algal photosynthesis leads to a secondary role for highly illuminated surface waters, and "shade" temperature in differentiating the specific capabilities species those growing in the dim light prevailing of adapting to variable environmental conditions. For in deep waters. Later on, the phytoplankton as a instance, it has been demonstrated that low temper- whole was sorted into sun and shade populations, ature decreases the optimum light intensity for both in fresh waters (Rhode et al. 1958) and in growth, and increases the inhibitory effect of high marine waters (Ryther and Menzel 1959). It has been light intensities, in Detonula confervacea (Smayda demonstrated by the latter authors that temporal 1969), Coptomonas ovata (Cloern 1977), and C. variations and mixing may modify significantly the erosa (Morgan and Kalff 1979). Otherwise, Li and conditions goveming light adaptation. In winter, Morris (1981) reported that Phaeodactylum tricornu- nun cells behave differently according to their history of adaptation; cells adapted to grow at high temper- "For further cytophysiological details, see atures have a higher rate of photosynthesis than Prézelin (1981).

294 when waters are isothermal and mixed below the firmed and explained by Falkowski and Owens euphotic zone, all phytoplankton forms are sun (1978). Likewise, Berseneva et al. (1978) demon- adapted. In summer, only surface phytoplankton is strated that Ditylutn brightwellii and Gyrodinium sun adapted; the deep components are shade adapted fissutn adapt to varying illumination by changing the because of the strong stratification of the water. rate of their enzymatic reactions while cellular chlo- Steemann Nielsen and Hansen (1959) demonstrated rophyll content remains constant, whereas other that planktonic algae from the deeper waters are better species such as Chaetoceros curvisetus , C. socialis , able to utilize low-intensity light and , on the other Gytnnodiniutn kovalevskii Peridinium trochoideutn , hand, that a reduced light intensity for optimal and Scenedesmus quadricauda photoadapt by mo- photosynthesis does not mean an enhanced ability to difying the chlorophyll concentration. Prézelin and utilize very low light. Data obtained by Ichimura Matlick (1980) also observed similar reactions with (1960) and Aruga and Ichimura (1968) are consistent the dinoflagellate Glenodinium sp., which shows with this concept. Conversely, Yentsch and Lee photosynthetic changes during light—dark cycles un- (1966) questioned the reality of this adaptation and related to variations in its pigmentation, and Prézelin assumed that natural phytoplankton populations in and Sweeney (1979) demonstrated that the freshwater suboptimal conditions do not adapt to maintain dinoflagellate Peridinium cinetum increases the total highest rates of growth, but simply respond to a number of PSU per cell while pigmentation is un- physiologically inferior environment, i.e. light changed. responses of algae at very low illumination are Algal adaptation to new light conditions is characterized by a low rate of carbon uptake, which is usually swift. For example, a variation in cellular saturated at low light intensity. This rate remains low chlorophyll a corresponding to a change in light when light intensities increase because there are not intensity can occur within 12h at 18°C in Skeletonema enough enzyme molecules present to utilize the costatum (Glooschenko et al. 1972) and a similar increased illumination. observation has been made with Phaeodactylum However, on the basis of unialgal culture ex- tricornutum at 20°C (Beardall and Morris 1976). periments , several other authors have claimed that Indeed, pigments, particularly chlorophyll a, very algae adapt their physiology to different light inten- often have a short turnover time. Thus, Riper et al. sities. Steemann Nielsen et al. (1962) and JOrgensen (1979) reported such turnover times ranged from (1964) described two types of responses: the "Chlo- 3 to 10 h for chlorophyll a with Skeletonema costa- rella type" which adapts to a new light intensity tum , depending on the growth conditions. Grumbach by changing the pigment content, and the "Cyclotella et al. (1978) mentioned a turnover time of only type" which adapts only by changing the light satu- 1 h for chlorophyll a in the green alga Chlorella rating rate, suggesting a shift in the concentration pyrenoidosa . The second photoadaptative strategy of the enzymes involved in the dark reactions of of phytoplankters, which belongs to, or is close photosynthesis. Later, JOrgensen (1969) focused to the "Cyclotella type," and consists of modifying attention on the fact that the two adaptation types the photosynthetic units without new chlorophyll are not sharply separated and that transition responses synthesis, can be performed within an even shorter can occur between the two extreme types. Usually, time period. That is , it can be achieved after only marine phytoplankton respond to decreased light th the generation time according to Prézelin and intensities by increasing photosynthetic pigment Matlick (1980). Yet the ability to adjust the cell content, but the responses differ greatly among the physiology to new light conditions, as reflected by the species, which agrees well with JOrgensen's idea. time required to show new photosynthetic patterns, Usually, chlorophytes have a more marked ability differs greatly with the species. Such cell adaptability, to modify their pigment concentrations than dino- usually high, is peculiar to algae; such fast light- flagellates and diatoms. For example, Lauderia governed adaptations are not observed in higher borealis shows very slight variations in its pigment plants. concentration when exposed to different light (Marra Otherwise, responses to illumination are not free 1978). Brooks (1964 unpublished, in Prézelin and from the influence of other physical and chemical Matlick 1980) suggested that Dunaliella tertiolecta factors. For instance, diatoms near their temperature and Skeletonema costatum may also photoadapt (1) optimum reach their maximum rates within only 24h, by a modification of the pigment concentration and when illumination increases to optimal intensities, (2) by a variation in the efficiency of the photosyn- but, conversely, when they grow at their temperature thetic units (PSU) because the pigment concentration minimum, their division rates require more than 1 d to per cell varies much less than the shifts in optimal respond to variation in light intensity (Hitchcock light intensities. Such results suggest that these algae 1981). photoadapt partly by increasing PSU size. The data In natural waters, diel illumination variations concerning Skeletonema costatum have been con- and turbulence could also generate a light—shade

295 adaptation by modifying, on a short-term basis, the intense light absorption; with such low illumination, illumination received by an organism growing in the the green alga cannot grow, although the cyanophyte mixing layer. Waves, for instance, are able to change does. Likewise, Foy et al. (1976) explained the the illumination conditions sufficiently to control dominance of Oscillatoria redekei over three other the photosynthetic patterns of the natural population: blue-green algae in an Irish lake during early spring at high light intensity the waves depress the photo- by its ability to grow fastest under low light con- synthetic rate; conversely, they increase photosyn- ditions. Competition between diatoms might also thesis at low light intensities (Frechette and Legendre be governed by light intensity, in association with 1978). Furthermore, the cells do not remain at the nutrient levels and temperature. For instance, Phaeo- same depth within a diel light—dark cycle. To sum- dactyhen tricomutum has been reported to dominate marize, a cell will be able to photoadapt only if its outdoor mass cultures when temperature ranges from transport rate is slow enough compared with its adap- 10 to 20°C, while the same species is poorly repre- tation time to adjust to the new light conditions. sented in all natural waters at the same temperature. Subsequently, "if mixing-processes occur on a time- Otherwise, many species such as the minute Thalas- scale shorter than it takes the cells to adapt to the siosira pseudonana , which has high division rates, variations in the light-regime, the vertical distribution and which would grow well in natural waters in of the light-dependent physiological characteristics the same 10-20°C range, is usually outgrown by would be expected to be more uniformly distributed" P. tricomutunt in mass cultures. Further experiments (Falkowski 1981). Hence, in a natural population, performed by Nelson et al. (1979) demonstrated that the fast-adapting algae are obviously at an advantage P. tricornutum is able to sustain unusually high in the species competition. In any case, chlorophyll growth rates at low light intensities; this property turnover times lower than one generation are of a contributes greatly to its success in nutrient-rich great ecological significance for phytoplankters and highly turbid mass cultures, systems in which potentially exposed to considerable variations in light only this diatom dominates. Such examples show the intensity during periods of a few hours (Riper et al. importance of photoadaptation in species compe- 1979). However, when the well-mixed layer extends tition. But light intensity is not the only aspect of beyond the 1% light penetration depth, as for instance illumination; day length and light quality might also in coastal upwelling, the phytoplankton populations have an important role. stay within the euphotic zone for a shorter average time than they do in thermally well-stratified marine DAY LENGTH AND TEMPERATURE GOVERNED or fresh waters. Hence, not only the light regime SPECIES SUCCESSIONS but also the light-exposure history play a critical role Early studies made by Guillard and Ryther in controlling the availability of nutrients (Nelson (1962) demonstrated that the diatom Detonula and Conway 1979). confervacea outgrows another diatom, Skeletonema when temperature is low (4°C) and the LIGHT INTENSITY AND NUTRIENT CONTROLLED costatum , duration of illumination is long (12-16 h). Further SPECIES SUCCESSIONS research confirmed this finding and demonstrated The effects of illumination on species succession that D. confervacea grows well at low light intensities have been differentiated from the action of other (Holt and Smayda 1974). Comparison between its factors only under experimental conditions, and physiological characteristics and the field conditions reports on this subject are still rare. Recently, Mur suggests that the ambient light intensities during et al. (1977, 1978) found that the maximal growth the winter bloom occurring in Narragansett Bay favor rate of the green alga Scenedesmus protubemns is its growth, while the concomitant short day lengths far higher than that of the cyanophyte Oscillatoria are unfavorable and would become a limiting factor. agardhii , 1.58 and 0.86 divisions per day, , respec- At the beginning of the bloom, D. confervacea tively. Nevertheless, the cyanophyte periodically grows faster than Skeletonetna costatum and Thalas- dominates the natural populations of many lakes. siosira nordenskioldii; its rate of photosynthesis is Light-effect studies demonstrated that the cyanophyte higher and its compensation light levels for photo- prefers low intensities and has the ability to select synthesis and growth are the lowest. Conversely, suitable conditions by modifying its buoyancy. Thus, for some reason still unclear, D. confervacea a simple light-governed mechanism enables the dominance ends while the bloom is progressing. cyanophyte to dominate. In nutrient-poor waters, Durbin (1974) reported that a cultivated Thalassiosira high illumination prevents O. agardhii from blooming nordenskioldii strain showed an optimal temperature in surface waters; dense populations occur only in which varied according to day length: at a longer deeper layers. In eutrophic waters, this organism day length, the optimum temperature and the optimum may multiply because, the growth potential being light intensities were lower than those required for higher, the biomass becomes denser and leads to an a shorter day length, which is consistent with early

296 results obtained with the same species (Jitts et al. flagellates and other motile algae are favored by 1964). Thus, though studies performed on this subject low intensities. Therefore, combining both light- have been few, such algal responses related to variable intensity and light-color effects, Wall and Briand day length and light intensities may be common. (1979) obtained responses consistent with distribution of different algae in the water column: cyanophytes, LIGHT QUALITY AND LIGHT diatoms, and green algae were favored by high INTENSITY GOVERNED SPECIES intensities and red light; these conditions occur in COMPETITION AND SUCCESSION surface waters of many lakes where these algae IN DOMINANCE dominate. In contrast, high intensities and red light induced the displacement of dinoflagellates from The selective attenuation of incident radiation the top to the bottom of the euphotic layer during by particulate and dissolved organic and inorganic daytime. However, the most common and, hence, compounds, and by the water itself, induces a change the most successful species belong to a group of in the spectral composition of illumination with minute and motile microflagellates depth, while light intensity is also attenuated. Green in which no preference for a given wavelength of light has been and blue light penetrate to the greatest depths. Thus, recorded; the foremost light-induced cell adaptation the algae reaching the bottom of the euphotic layer could lie in this motility that allows them to maintain will be illuminated mainly by blue light, particularly their vertical position in a layer corresponding to an in oligotrophic waters. The dependence of algal optimal light regime. photosynthetic response on wavelength of light has already been studied in detail since the early contri- butions by Emerson and Lewis (1943) and Haxo Buoyancy and Motility and Blinks (1950), while the "complementary chro- matic adaptation," that is, the variation of pigment PHYTOPLANKTON STRATEGIES TO REMAIN composition with light quality, is one of the oldest IN NUTRIENT-RICH LAYERS fields of algal research (Engelmann and Gaidukav 1902, in Halldal 1958); pertinent reviews on the topic The success of a phytoplankter largely depends can be found in Halldal (1970). Regarding this early on its ability to remain suspended in the photic zone attention, it is surprising that little has been done to where sufficient light is available for the photosyn- investigate whether light quality might govern, in thetic reactions. Unfortunately, the upper layers are part, the species competition in phytoplankton popu- not the richest in the water column. The nutrient lations, in spite of the fact that the vertical zonation of richness often increases below the mixed layer at the benthic seaweeds has been frequently related to the bottom of the euphotic zone. In the upper nutrient- presence or the absence of pigments such as phyco- limited layers, the availability of light relative to the bilins which preferentially absorb short wavelengths photochemical requirements of the alga exceeds that and allow these plants to live in deeper waters (e.g. of inorganic nutrients relative to the requirements Levring 1966). Nevertheless, recent findings mitigate of the enzymatic processes. Conversely, in deeper some of these early oversights (see for instance Ramus layers, production is controlled primarily by avail- et al. 1976, 1977). ability of substrates for the photochemical system However, Wall and Briand (1979) investigated relative to its needs. Between these two extremes, whether light intensity and color can play a major represented by the upper and lower bounds of the role in determining the vertical gradient observed photic zone, the relative availability of substrates in phytoplankton species composition in Heney Lake, varies systematically for a species according to the Québec Natural populations were cultivated in situ in light and nutrient supply rate (Bienfang and Gunder- plexiglass cubes of different colors. Under red light, sen 1977) and differs in all species according to the proportions of cyanophytes, diatoms, and green their biochemical characteristics. It explains why algae increased, while those of dinoflagellates de- the chlorophyll maximum is not, generally, at the top creased. Chrysophytes and cryptophytes tended of the water column but at depths where the avail- to be relatively enhanced by blue light. However, abilities of light and nutrients are still high enough to only four species were reported to react significantly allow good production (Eppley et al. 1977). to light at specific wavelengths; blue light enhanced In this respect, phytoplankters belong to two the growth of the chrysophyte Chtysocapsa sp. and large families. Motile cells, such as dinoflagellates, depressed that of the cyanophytes Dactylococcopsis are able to perform vertical migrations of a phototactic smithii and Rhabdoderma sigmoides; under red light, nature and to choose an optimum depth, at least the growth of the cyanophyte Gloeotrichia echinulata between some limits (Seliger et al. 1970; Blasco decreased. Responses to given light intensities are 1978). On the other hand, nonmotile cells can consistent with predictions, i.e. the diatoms and green counterbalance the natural movements of the water algae dominate under high illumination whereas dino- only by regulating their buoyancy, either by changing

297 their own density as diatoms do, or by producing compounds, and particularly its glycerol content, gas vacuoles, as cyanophytes do (Walsby 1972). according to the intensity of different factors (par- It is obvious that the ability of all flagellates to ticularly external molarity, light intensity, and migrate actively is an important advantage in ob- quality). A positive correlation between the glycerol taining sufficient nutrients and light (Gran 1929). content and the age of the culture was found. The species sinking rate is an important eco- Usually, senescent cells sink faster than actively al. 1967). It has been logical factor in understanding succession, as stressed growing ones (Eppley et buoyancy is often by Hutchinson (1967) who postulated that much of verified that, in the sea, a positive in active growth, particularly the seasonal species succession is due to the resulting associated with cells at the beginning of blooms. But it is also possible effect of turbulence and sinking, and O'Brien (1974) to find high sinking rates during blooms (Lânnergren who included the species specific loss rates as a major 1979). Such high sinking rates may be interpreted parameter in his nutrient-limitation models. An as adaptations to a shift of the nutritional conditions. experimental demonstration of its importance was As a matter of fact, when the nutrients become less carried out by Knoechel and Kalff (1975), who concentrated in the upper layers, an increase in the concluded a study based on the track autoradiography rate leads the organisms to reach lower layers technique by stating that the summer decline in a sinking and, thus, facilitates nutrient Canadian lake of the diatom population, dominated richer in nutrients uptake. Such a phenomenon has been demonstrated by Tabellaria fenestra , and the increase of the cyano- by Kahn and Swift (1978), who showed that Pyro- phyte population, especially Anabaena planctonica, cystis noctiluca is neutrally buoyant when nutrient was mainly due to the lower sinking rate of the latter. depleted, and positively buoyant after taking up In situ sinking rates are strongly affected by water the bottom of the photic layer. Such mass stability. They depend on the size, shape, nutrients at explanations might be given to explain in part why density of organisms, and also on the physiological the large diatoms usually appear during blooms, when characteristics of the cells (Smayda 1970; Sournia concentrations are high. Under these con- 1981). It is well known that buoyancy differs during nutrient ditions, without a variable buoyancy, it would be light—dark cycles; for instance, Anderson and absolutely impossible for such large cells to remain Sweeney (1978) showed that the diatom D113,111,71 in the photic layer because of an unfavorable surface/ brightwellii changes its ionic balance and is able to volume ratio at other times of the bloom. Later modify its density and, consequently, its buoyancy on, such algae are eliminated from the upper layers between day and night. The settling rate is lower replaced by smaller cells, and in the dark. A very similar result is given by SchCine by sinking and are rates or (1972); the specific gravity of Thalassiosira rontla particularly flagellates, with lower sinking is proportional to light intensity: with increasing even a positive buoyancy. Usually, cells in deeper light intensity, the sinking rate increases. Scheme layers find better conditions in regard to nutrients, gave no metabolic explanation for this mechanism. but worse conditions for growth if we consider also But, for the alga such a buyoyancy control may the physical factors, namely dim light and often lower temperature. There is, therefore, a very com- be a direct response to an excessively bright light; plex interrelationship between these physical factors T. rotula seems to be a shade cell and prefers low physiological processes of nutrient uptake by light intensities. Recently, Burns and Rosa (1980) and the algae, and it is difficult to study any individual studied the settling velocities. of the main species element of the relationship in isolation from the of a lake phytoplankton population, and reported that others. velocities of nonliving particles were about equal for diurnal variation each size range and showed little EFFICIENCY OF NUTRIENT CAPTURE algae showed considerable whereas, in contrast, the IN DEEP LAYERS response to changing light intensity during the day- time. Some flagellates migrated downward at sunset; Various results, obtained both in the field and a species of diatom settled more slowly during day- in the laboratory, support the hypothesis that the light than at night; a blue-green alga showed maximum migration of dinoflagellates to nutrient-enriched deep buoyancy at midnight; and some green algae changed waters is effective because the cells are able to take up from sinking on a bright day to upward movement nutrients at night. Such an uptake by natural popu- on a dull day. These observations clearly indicate lations or clones has often been observed and seems that settling rate is a result of various physiological common even if rates are very low (Dugdale and processes and subsequent species-dependent changes Goering 1967; Holmes et al. 1967; Eppley and in cell content, but such changes have been little Harrison 1975). It has also been verified that the studied. However, Jones and Galloway (1979) uptalce in the dark is the response of algae to drastic demonstrated experimentally that Dunaliella tertio- N starvation (Syrett 1962). The more nutrient-de- lecta is able to modify its internal pool of organic pleted are the cells, the more important is uptake

298 at night (Malone et al. 1975). But, besides slight Reaching such layers, if they are rich in nutrients, dark uptake, it has been pointed out that, usually, is undoubtedly an important advantage in the compe- maxima for nitrate and ammonia uptake by natural tition for growth. Reshkin and Knauer (1979) phytoplankton populations occur during daylight measured uptake of phosphate as a function of light hours (Dugdale and Goering 1967; MacIsaac and intensity by natural phytoplankton populations and Dugdale 1972; Packard and Blasco 1974). Because found a relationship which approximated to the of the nocturnal migrations of the dinoflagellates, Michaelis-Menten equation. Light "half-saturation particularly during red tides, it had been suggested constants" (K 1,), corrected for dark uptake, were that only these organisms were able to take up nu- found to be 11.3-18.3% of the surface light intensity, trients in the dark, in contrast with the diatoms which which is comparable to K1, values for light-dependent could not do so. Harrison (1976) demonstrated that, uptake of nitrate and ammonium by phytoplankton in culture, Gonyaulax polyeclra has the ability for found by MacIsaac and Dugdale (1972). The possible both uptake and assimilation of nitrate in the dark importance of dark uptake of phosphate by phyto- and would be capable of meeting 50-100% of its plankton was also pointed out by Reshkin and Knauer daily nitrogen requirement for growth from dark (1979). They measured rates of dark uptake of phos- assimilation only. In the same way, it has been phate that were 29.1-33.4% of the maximum uptake verified that two dinoflagellates, Pyrocystis fusi- rate in the light. Although dark uptake of phosphate fonnis and Dissodinium lunula, take up nitrate and in natural waters has been attributed to bacteria, ammonium at almost equal rates both day and night these workers suggest that, because Thalassiosira when preconditioned on 12:12 h night-day cycle fluviatilis has been shown to accumulate phosphate at illuminations between 42 and 67 it, E • m-2 • s- ' (Bho- in the dark, dark uptake by phytoplankton in general vichitra and Swift 1977). However, the ability of may also significantly affect calculations of KI,. algae to take up nitrogen sources in the dark varies The low K1, values calculated for phosphate, nitrate, greatly; for instance, Coccolithus huxleyi does so but and ammonia are in good agreement with the average the diatom Skeletonetnu costatutn does not. incident light level in the photic zone of high-pro- duction areas. For example, Huntsman and Barber Unfortunately, field surveys are not often con- (1977) showed that, in the northwest Africa up- sistent with an ecological extension of these obser- welling , phytoplankton had an effective exposure to vations. For example, MacIsaac (1978), comparing a mean light intensity of 10% relative to the surface her own data with those of MacIsaac and Dugdale value. (1972) on the waters off California and Peru, asserts that there is no evidence that dinoflagellates have a competitive advantage over diatoms under low light NUTRIENT UPTAKE ENHANCEMENT BY conditions. The dark uptake of nitrate is apparently CELL DISPLACEMENT characteristic of both starved dinoflagellates and The migrations of small flagellates or the sinking diatoms. But the migratory capability of dinoflagel- of other algae have another very important effect lates may help them to occupy the most favorable on their ability to take up nutrients: the availability levels of the water column where the concentration of dilute nutrients is enhanced by passive or active of nutrients is higher and may help them to persist movements of the cells, or by movements of the in the blooms. This is the biggest advantage of the medium (Munk and Riley 1952). In fact, the uptake dinoflagellates over the diatoms, which cannot avoid rate of a nutrient by a cell can be limited by the sinking when turbulence is decreased. It explains inability of the nutrient to diffuse to the cell as rapidly why strong upwellings, with high nutrient conditions, as the cell can assimilate it at its ambient concen- are accompanied by diatom dominance, while a tration. It has been demonstrated that the effect of relaxation of the system leads rapidly to dinofla- transport limitation was decreased by mixing and gellate dominance even at reduced nutrient levels that diffusion transport can influence uptake rates (MacIsaac 1978). This has been verified experi- at low-nutrient concentrations (Pasciak and Gavis mentally by Eppley et al. (1978). More than dark 1974, 1975). Hence, it is assumed that diffusion uptake, the ability to take up nutrients at very low transport limitation may be so great that it can lead light levels is very important for phytoplankters. It an alga to lose the competitive advantage that it is reported that light saturation of nitrate assimilation might have had, given its uptake capabilities (Gavis is much lower than that of photosynthesis (Hattori 1976). This concept is very complex and depends 1962). For instance, it has been observed that this on many factors (Hulburt 1970) — the effect of light saturation for Pyrocystis sp. was about 10 flow determined by cell size, shape, density, self times lower than that of photosynthesis (Swift and movement, the "affinity" of the nutrient for the Meunier 1976; Bhovichitra and Swift 1977). In clear surface of the cell, the surface/volume ratio, and oceanic waters, such phytoplankters are able to take so on. The lower the nutrient concentration in the up nitrate at the maximum rate at a depth of 100 m. medium, the greater is the beneficial action of the

299 sinking. As explained by Canelli and Fuhs (1976), water ponds, salt marshes, or rock pools) have to an increase in sinking rate has two antagonistic adapt their physiological processes rapidly to be able actions: (1) the loss of organisms under a mixed to maintain their growth as constant as possible. and illuminated layer is linear; (2) the increase in Likewise, freshwater algae reaching estuaries can nutrient uptake follows an hyperbolic curve versus survive only if they are somewhat euryhaline. Thus, the sinking velocity. Thus, there is an optimum value in many coastal and usually nutrient-rich environ- of sinking rate for each nutrient condition; for each ments, salinity might be expected to shift the species species there is a sinking rate that allows an optimum dominance in a community reaching a variable of production for given nutrient concentrations. All salinity area, and to be responsible for the peculiar species do not have the same characteristics and do composition of resident populations (Lohmann 1908; not find the best conditions for maximum production Vâlinkangas 1926; Carter 1938; Droop 1953; Halim at the same time. 1960; Hulburt 1963; Hulburt and Rodman 1963; Hobson 1966; Saugestad 1971; Podamo 1972; Chré- ABILITY OF MOTILE CELLS TO CROSS OVER tiennot 1974; Charpy-Roubaud and Charpy 1981). THE THERMOCLINE BARRIER However, laboratory studies have shown that some species of algae are able to tolerate very large vari- One other factor cannot be dissociated from ations in salinity and this has stimulated studies on buoyancy and the ability of phytoplankters to talce the physiological mechanisms allowing algae to resist up nutrients at lower depths in the water column: osmotic stresses. the strong density gradient of the thermocline which slows down the sinking of all particles; very often it is possible to observe an increase of cell numbers ALGAL RESISTANCE TO SALINITY VARIATIONS at this level. But some migrations of dinoflagellates Allen and Nelson (1910) and Allen (1914) first traverse large temperature gradients and cross this remarked that some marine algae, e.g. the diatoms hurdle. It has been shown that many freshwater Biddulphia mobiliensis, Coscinodiscus excentricus, dinoflagellates can cross this density hurdle, although Skeletonema costatum, and Thalassiosira gravida, there is considerable variation in their ability to do were able to grow in different media, the salinity so. For instance, Ceratium hirundinella can cross of which varied within a broad range, namely a 2°C temperature gradient (Talling 1971), Peri- 17-50%0 for T. gravida Braarud (1951) extended dinium westii , 4°C (Berman and Rhode 1971), this concept to many phytoplankton species, espe- marssonii, 6°C (Soeder 1967), and cially dinoflagellates, which he studied in detail. By Pyrodinium bahamense, 5°C (Seliger et al. 1970). measuring the growth of cultured algae versus salin- The ability of marine phytoplankters to cross these ity, he demonstrated that the optimum salinity for gradients is somewhat less but they can often cross growth is usually lower than that of the waters from over a 2°C gradient and sometimes more (Cachonina which natural populations come, and that species niai migrates across 5°C). In very stratified waters are separated by the lowest salinity at which they can such phytoplankters can cross the thennocline barrier grow. From the available data (Braarud 1961; Wil- and gain access to subthermocline nutrient pools. liams 1964; Nakanishi and Monsi 1965; Smayda This ability may be of a great importance for these 1969; Ignatiades and Smayda 1970; Schae 1974; species because it has been pointed out that, below Paasche 1975; Grant and Horner 1976; White 1978), a temperature characteristic of each latitude, plant it appears that such optimal values range from 15 to nutrients and temperature are inversely related 20%0 for algae as different as coccolithophorids (Kamykowski and Z,entara 1977). (e.g. Coccolithus huxleyi and Syracosphaera car- The problem in such situations is that, some- terae), dinoflagellates (e.g. several Ceratiwn spp., times, the temperature below the thermocline is low Exuviaella baltica, Peridinium trochoideum), cryp- enough to be under the temperature optimum and tomonads (e.g. Hemiselmis virescens), chrysomonads even below the temperature limit for the species (e.g. Pavlova lutheri), and diatoms (e.g. Asterionella growth. In these conditions, the benefit of higher japonica , Skeletonema costatum). Furthermore, nutrient concentrations is greatly diminished by the McLachlan (1960) reported that the green alga decrease in the rate of all the metabolic reactions. Dunaliella tertiolecta can grow in media having a Furthermore, the ability for the cells to reach such salinity between 3.7 and 120%0 and Takano (1963) deeper layers is strongly modified by the reduction claimed that diatoms can also adapt to grow between of their motility. broad-ranging values such as 6 and 45%0, e. g. Cerataulina pela gica, Chaetoceros radians, Lepto- Salinity cylindrus &minis, and Thalassiosira decipiens , while Skeletonema costatum can grow even at 3%0. Algae living in an environment where salinities This concept was supported by Bonin (1969), who may change rapidly (such as estuaries, brackish reported that with Chaetoceros ends growth can 300 occur within a 5-50%0 range, and Tomas (1978), tance of salinity in controlling phytoplankton distri- who demonstrated that the chrysophyte Olisthodiscus bution (Provasoli 1958; Braarud 1961; Carpelan luteus tolerates a salinity range of 2-50%0. 1964; Wood 1965; Riley 1967; Gessner and Schramm 1971; Rice and Ferguson 1975). But speculations EFFECTS OF SUBOPTIMAL SALINITIES based on both field evidence and laboratory studies Thus, the ability of algae to grow at various are rather rare and open to criticism. For instance, Nakanishi and Monsi (1965) recorded the relation- salinities is a long-established concept and seems to be shared by many species belonging to different ships between salinity and photosynthesis of phyto- plankton sampled taxa. However, the effect of suboptimal salinities in Tokyo Bay and the brackish Lake Hinuma, and compared the data they obtained is poorly documented: often the only quantitative with the responses of three cultured algal strains, criterion used to record algal responses in such experi- ments is the cell density measurement, with some- Skeletonema costatum , Chaetoceros sp., and Chlo- times the duration of lag phase, and few results are rella ellipsoidea, with the aim of explaining the dominance of some species in natural waters. They given on the physiological characteristics of the cells. noted that the diatom S. costatum can grow down Yet, McLachlan (1961) demonstrated that the to 3%0, with optimum growth at 18%o, whereas chlorophyll a content per cell of algae such as Platymonas sp. and Porphyridium sp. can be affected Chaetoceros sp. shows its optimum at 34%o. This by suboptimal salinity, while growth itself is unaf- led them to remark that most species living in Tokyo fected. For other species such as Amphidinium carteri Bay are euryhaline and have their optimum salinities and Olisthodiscus sp., growth and chlorophyll a markedly different from the salinities of their habitats. content show a respective maximum rate for the same However, their discussion is quite unclear and failed optimal salinity value. This author also reported that to demonstrate that salinity governs the phytoplankton distribution in the studied areas. The main Syracosphaera carterae , Pavlova lutheri , and reason Thalassiosira decipiens show a very wide range is that natural population responses were compared with responses of of salinity values for optimal growth, which conflicts cultured strains not reported to slightly with previous results given for these latter be dominant, whereas the dominant species have not been put species. Otherwise, the chlorophyll content of the into culture. green alga Platymonas subcordiformis (Kirst 1975) Mahoney and McLaughlin (1979) observed and of the diatoms Chaetoceros affinis (Bonin 1969) that, in New York harbor, estuarine and oceanic and Cyclotella cryptica (Liu and Hellebust 1976a) waters are dominated primarily by several flagellates has been reported not to vary very much in the range whose behavior seems to be closely related to salinity. of optimal salinities, while it decreases sharply for As a matter of fact, in this area, algal blooms fre- values above and below this range. Other researchers quently develop in the harbor estuarine waters and using photosynthetic carbon uptake rate (Curl and then spread to both ocean and rivers; occasionally, McLeod 1961; Nakanishi and Monsi 1965) or oxygen the algae bloom in the river waters and then extend production and respiration measurements (Ehrhardt to estuarine and oceanic ones. However, in one and Dupont 1972) reported results similar to those such case, a dominant species in the tidal river obtained by recording cell-division rates or chloro- waters (with salinity between 15 and 21%o), Mas- phyll and production values. Otherwise, White sartia rotundata, lost its dominance in the marine (1978) demonstrated that maximum growth and toxin waters in favor of Prorocentrum micans or Oils- production in Gonyaulax excavata occur at different thodiscus luteus . Laboratory experiments demon- salinities, 30.5 and 37%0, respectively. But, more strated that M. rotundata shows the best growth in generally, little effort has hitherto been made to the range 24-30%o, O. luteus in the range 10-36%o, characterize the different responses of cell physio- and P. micans within 27-36%o. Thus, the three logical processes to salinity variations. In particular optimum ranges for growth sufficiently match the nothing is known yet about inorganic nutrient uptake. salinities characteristic of the studied waters (17- However, Liu and Hellebust (1976b) and 32%), eliminating salinity tolerance as a major factor Hellebust (1978) opened a new area of research in limiting the development of these species in these when they demonstrated that the marine diatom natural conditions. On the other hand, experiments Cyclotella cryptica increases its uptake rate of organic carried out to study the influence of the change from nutrients like glucose and amino acids, according to brackish to ocean water salinities allow the authors the increase of NaCl concentrations, from 0 mM to point out that these stress effects inhibit growth, (no uptake) to 100 mM (optimum uptake rate). especially in M. rotundata and O. luteus , but that the latter species is able to recover its growth SALINITY-GOVERNED SPECIES DISTRIBUTION rate faster. Indeed, Mahoney and McLaughlin's inter- On the basis of field observations and experi- pretation of their experimental data fit well with the mental data, many authors have stressed the impor- observed natural distribution of these phytoflagel- 301 lates. But the authors did not give any physiological Only a few algae have been studied with regard explanation of the three different species responses. to the biochemical changes that occur in response In fact, all early reports left unclear the physio- to osmotic stress. Kauss (1974, 1979) has shown logical mechanisms by which the cells adapt to that, in Ochromonas , increasing the concentration osmotic stress, although some speculations on the of external solutes causes an increase in the internal subject had been formulated (see review by Guillard concentration of isofloridoside (a-1,1,-galactosyl- 1962). However, more recently, promising research glycerol) and, to a lesser extent, potassium and free has been initiated in this field. amino acids. Decreasing the external osmotic value stimulates the rapid transformation of isofloridoside PHYSIOLOGICAL AND BIOCHEMICAL to a reserve glucan. ADAPTATIONS TO OSMOTIC STRESS The pathway of isofloridoside synthesis has also been studied: initiation of biosynthesis appears to It has been shown that many algae respond to involve the proteolytic activation of an inactive increasing concentrations of external solute by in- precursor of the enzyme isofloridoside—phosphate creasing the intracellular concentration of certain synthase in response to changes in external water osmotic solutes, particularly polyols such as glycerol, potential. Kreuzer and Kauss (1980) have also shown mannitol, and sorbitol, and the amino acid proline that there is de novo synthesis of the enzyme a- (Hellebust 1976a, b; Liu and Hellebust 1976a, b; galactosidase in response to hyperosmotic conditions. Brown and Hellebust 1978b; Kauss 1974, 1977, The regulatory signal and regulatory sites of the iso- 1978, 1979; Kirst 1979; Schobert 1979). The change floridoside pathway have yet to be elucidated. in concentration of these solutes has been interpreted The osmoregulatory mechanism in Cylindro- as being a simple osmotic mechanism causing water theca closterium seems to parallel that in Ochro- reflux and balancing the difference in water potential Intracellular concentrations of free mannose between the inside and the outside of the cell. How- monas and polymannose in this diatom appear to be in ever, although any intracellular solute (other than dynamic equilibrium and change in response to the dominant extracellular solute) which responded changing external water potential (Paul 1979). In- appropriately to changes in extracellular water creasing salinity shifted the equilibrium towards activity could function as an osmoregulator, it is production of free mannose, and the rate of CO 2 remarkable that the major solutes accumulated under also increased. Decreasing salinity stimulated fixation water stress by a wide range of algae and higher the polymerization of mannose. It was also found plants are usually of only two types, polyols and that the enzymes involved in the mannose/poly- proline. Schobert (1977, 1979, 1980a) and Brown mannose conversion appeared to respond to changes and Borowitzka (1979) have argued that polyols and in water potential rather than to changes in any proline are "compatible solutes" which, because of particular ion, as varying the water potential with their unique properties, have a special role in water sorbitol was equally as effective as using inorganic regulation other than simply acting as osmoregulators. "Compatible solute" was a term originally salts. responds to in- coined by Brown and Simpson (1972), and may be Phaeodaciyhun tricornutum creasing osmolarity of the external medium by accu- defined as a solute that is compatible with cellular mulating proline: this accumulation appears to be metabolism because, even when present at very high due to inhibition of proline catabolism (Schobert concentrations, there is minimal inhibition of en- 1980b). zymes and, hence, biochemical reactions. The reasons why the physico-chemical properties of The osmotically stimulated production of glyc- polyols and proline make them eminently suitable erol in Dunaliella has been intensively studied. The as compatible solutes have been examined in detail relevant literature has been reviewed by Brown and by Schobert (1977, 1979, 1980a). It is not possible Borowitzka (1979). Dunaliella is remarkably adapt- to reiterate this work fully here without a lengthy able to a wide range of salt concentrations. For discussion of the physico-chemical properties of example, D. tertiolecta is able to grow within a water and the compatible solutes. Suffice it to say range of NaCI concentrations from 0.06 to > 3.6 M that, under conditions of water stress, the reduction (Borowitzka et al. 1977). If transferred to concen- in cellular water activity is assumed to impair the trated salt solutions, Dunaliella responds immediately hydration sphere of cellular constituents, particularly by initiating synthesis of glycerol, which continues macromolecules. It is proposed that the compatible for about 90 min, after which the normal cell volume solutes function primarily by improving the solubility is regained (Kessly unpublished results, quoted in of macromolecules by interacting with them to pre- Brown and Borowitzka 1979). Therefore, there must serve the surrounding water structure, thus preventing be some other mechanism that allows certain meta- precipitation of these molecules and irreversible bolic functions to continue during this period while damage to the algal cell. the osmotic equilibrium is being restored. The energy 302 for glycerol synthesis seems to come primarily from SALINITY-GOVERNED SPECIES respiration. Dunaliella is able to control closely COMPETITION: PROSPECTS its glycerol content in response to salinity and it is Although there are many publications describing thought that a probable regulatory site is an NADP- one specific glycerol dehydrogenase, but the exact nature species based ecophysiological experiments that stress the importance of salinity as an ecological of the regulatory signal that initiates changes in factor in estuarine and coastal waters, it is surprising glycerol metabolism is not yet understood (Brown to find a lack of and Borowitzka 1979). experimental studies and in situ observations of the potential role of salinity in nutrient The discussion so far has centered on adaptation partition between species and, subsequently, in phyto- to decreasing water potential, but phytoplankters plankton succession. Nevertheless, the shift in must also be able to adapt quickly to increasing species dominance from tidal river to brackish and water potential (osmotic downshock). Algae which marine waters, reported by Mahoney and McLaughlin are downshocked are faced with the problem of (1979), and the failure of oceanic strains to dominate increasing their internal water potential and various in coastal waters as a result of their inability to types of response have been observed. For example, increase their growth rate at higher nutrient concen- Ochromonas malhamensis (Kauss 1974) and Cylin- trations (Hulburt 1979), indicate that the relationships drotheca fusiformis (Paul 1979) respond to mild between dominant species and osmotic strength in osmotic downshock by repolymerizing the compatible salinity-variable environments are not yet at all solute. In gdunaliella the effect of dilution shock clear, at least as far as ecophysiological processes is to cause a metabolic dissimilation of intracellular are concerned. glycerol, a process that begins within minutes of As a matter of fact, in such studies it is abso- downshock and is complete within about an hour lutely necessary to consider separately in the natural (Kessly unpublished, in Brown and Borowitzka environment the action (1) of the total dissolved 1979). The metabolic dissimilation of glycerol is not salts, (2) of the minimum requirements for predomi- fully understood, but must obviously involve con- nant anions and cations, and (3) the influence of the version to nonosmotically active compounds, most different ionic ratios in the water. In many estuaries, likely starch and/or CO 2 . Even with a relatively where the salt concentration is low enough and the large downshock, very little glycerol is lost by leakage ionic ratios may vary considerably depending on the from Dunaliella spp. (Brown and Borowitzka 1979). origin of the water, such parameters, among many In contrast, some algae respond to downshock by others, may explain the composition of species as- rapid ejection of low molecular weight compounds semblages that cannot otherwise be understood only from the cell, followed by a period of metabolic in terms of the direct influence of the total osmotic adjustment of various processes — respiration, strength. Since it has been demonstrated that the photosynthesis, and membrane permeability. Algae concentration of particular ions directly governs the that show this type of adaptation are Pavlova ability to take up some organic molecules by algae, (= Monochrysis) lutheri (Craigie 1969), Platytnonas it is possible to believe that, more generally, the suecica (Hellebust 1976b), Platymonas subcordi- concentrations of certain conservative ions might formis (Kirst 1979), and Phaeodactylum tricornutum affect the uptake rate of inorganic nutrients. Hence, (Schobert 1980b). This mechanism allows the algae research into salinity and ion balance paralleling that to reduce very quickly the internal concentration of already carried out with combined variations of osmotically active solutes, thereby obviating pos- nutrients and illumination and/or temperature might sible lysis of the cell due to greatly increased turgor give exciting results and bring a new insight in pressure. estuarine phytoplankton ecology. Moreover, Brown and Hellebust (1978a) demon- strated the very important role of the concentration of particular ionic species in deplasmolysis. They focused attention on KC1, the only electrolyte capable Acknowledgment of allowing deplasmolysis of Cyclotella cryptica cells when supplied alone in the medium. On the other We wish to thank Dr Trevor Platt for a critical hand, Ca2+ , Me+ , and Na+ , which are involved in the review of the manuscript. energy-dependent uptake of K+, strongly enhance the ability of the alga to adapt at various salinities. Furthermore, it has been observed that the energy References used in this regulation has not always the same origin. Some algae, like Cyclotella cryptica , are able to AHLGREN, G. 1978. Growth of Oscillatoria agardhii in deplasmolyze in the dark, whereas others, like Scene- chemostat culture. 2. Dependence of growth constants desmus obliquus (Wetherell 1963) need light to do so. on temperature. Mitt. Int. Ver. Limnol. 21: 88-102.

303 ALLEN, E. J. 1914. On the culture of the plankton diatom BROWN, A. D., AND J. R. SIMPSON. 1972. Water-relations Thalassiosira gravida Cleve, in artificial sea-water. of sugar-tolerant yeasts: the role of intracellular polyols. J. Mar. Biol. Assoc. U.K. 10: 417-439. J. Gen. Microbiol. 72: 589-591. ALLEN, E. J., AND E. W. NELSON. 1910. On the artificial BROWN, L. M., AND J. A. HELLEBUST. 1978a. Ionic culture of marine plankton organisms. J. Mar. Biol. dependence of deplasmolysis in the euryhaline Assoc. U.K. 8: 421-474. diatom Cyclotella cryptica. Can. J. Bot. 56: 408-412. ANDERSON, L. W. J., AND B. M. SWEENEY. 1978. I978b. Sorbitol and proline as intracellular Role of inorganic ions in controlling sedimentation osmotic solutes in the green alga Stichococcus bacil- rate of a marine centric diatom Ditylum laris. Can. J. Bot. 56: 676-679. J. Phycol. 14: 204-214. BURNS, N. M., AND F. ROSA. 1980. In situ measurement ARUGA, Y., AND S. E. ICHIMURA. 1968. Characteristics of the settling vlocity of organic carbon particles and of photosynthesis of phytoplankton and primary pro- 10 species of phytoplankton. Limnol. Oceanogr. duction in the Kuroshio. Bull. Misaki Mar. Biol. 25: 855-864. Inst. Kyoto Univ. 12: 3-20. CANELLI, E., AND G. W. FUHS. 1976. Effect of the sinking BEARDALL, J., AND I. MORRIS. 1976. The concept of rate of two diatoms (Thalassiosira spp.) on uptake light intensity adaptation in marine phytoplankton: from low concentrations of phosphate. J. Phycol. 12: some experiments with Phaeodactylum tricornutum. 93-99. Mar. Biol. 37: 377-387. CARPELAN, L. H. 1964. Effects of salinity on algal distri- BERMAN, T., AND W. RHODE. 1971. Distribution and bution. Ecology 45: 70-77. migration of Peridinhun in Lake Kinneret. Verh. Int. CARTER, N. 1938. New or interesting algae from brackish Ver. Limnol. 19: 266-276. water. Arch. Protistenk. 90: 1-68.

CHARPY - BERSENEVA, G. P., L. M. SERGEYEVA, AND Z. Z. ROUBAUD, C. J., AND L. J. CHARPY. 1981. FINENKO. 1978. Adaptation of marine planktonic A biomassa fitoplanctonica do Saco Justino (32°S, algae to light. Oceanology 18: 197-201. Lagoa dos Patos, Rio Grande do Sul). Bolm Inst. Oceanogr. São Paulo. (In press) BHOVICHITRA, M., AND E. SWIFT. 1977. Light and CHRETIENNOT, M.-J. 1974. Nanoplancton de flaques dark uptake of nitrate and ammonium by large oceanic dinoflagellates: Pyrocystis noctiluca , Pyrocystis supralittorales de la région de Marseille. I. Étude fizsiformis , and Dissodinium hinul a. Limnol. Ocea- qualitative et écologie. Protistologica 10: 469-476. nogr. 22: 73-83. CLOERN, J. E. 1977. Effects of light intensity and tem- perature on Cryptomonas ovata (Cryptophyceae) BIENFANG, P., AND K. GUNDERSEN. 1977. Light effects growth and nutrient uptake on nutrient-limited, oceanic primary production. Mar. rates. J. Phycol. 13: 389- Biol. 43:•187-199. 395. CRAIGIE, J. S. 1969. Some salinity-induced changes in BLASCO, D. 1978. Observations on the diel migration of growth, pigments, and cyclohexanetetrol content of marine dinoflagellates off the Baja California coast. Monochlysis lutheri. J. Fish Res. Board Can. 26: Mar. Biol. 46: 41-47. 2959-2967. BONIN, D. J. 1969. Influence de différents facteurs éco- CURL, H., AND G. C. McLEon. 1961. The physiological logiques sur la croissance de la diatomée marine ecology of a marine diatom Skeletonenza costatum Chaetoceros affinis Lauder en culture. Téthys 1: (Grey.) Cleve. J. Mar. Res. 19: 70-88. 173-238. DI TORO, D. M., D. J. O'CONNER, AND R. V. THOMANN. BORCHARDT, J. A., AND H. S. AZAD. 1968. Biological 1971. A dynamic model of the phytoplankton popu- extraction of nutrients. J. Water Pollut, Control Fed. lation in the Sacramento-San Joaquim Delta, p. 40: 1739-1754. 131-180. In J. D. Hem [cd.] Nonequilibrium systems BOROWITZKA, L. J., D. S. KESSLY, AND A. D. BROWN. in natural water chemistry. Adv. Chem. Ser. 106. 1977. The salt relations of Dunaliella. Further DROOP, M. R. 1953. On the ecology of flagellates from observations on glycerol production and its regulation. some brackish and fresh water rockpools of Finland. Arch. Microbiol. 113: 131-138. Acta Bot. Fenn. 51: 1-52. BRAARUD, T. 1951. Salinity as an ecological factor in DUGDALE, R. C., AND J. J. GOERING. 1967. Uptake of marine phytoplankton. Physiol. Plant. 4: 28-34. new and regenerated forms of nitrogen in primary 1961. Cultivation of marine organisms as a means productivity. Limnol. Oceanogr. 12: 196-206. of understanding environmental influences on popu- DURBIN, E. G. 1974. Studies on the autecology of the lations, p. 271-298. In M. Sears [ed.] Oceanography. marine diatom Thalassiosira nordenskedii Cleve. Am. Assoc. Adv. Sci. Publ. 67. I. The influence of daylength, light intensity and BRIAND, F. J. P. 1975. Effects of power-plant cooling temperature on growth. J. Phycol. 10: 220-225. systems on marine phytoplankton. Mar. Biol, 33: EHRHARDT, J. P., AND J. DUPONT. 1972. Effets de la 135-146. salinité du milieu de culture sur les activités photo- BROCK, T. D. 1967. Micro-organisms adapted to high synthétiques et respiratoires de Dunedin salina temperatures. Nature (London) 214: 882-885. (Dunal) Theodoresco. C. R. Soc. Biol. 166: 516-519. BROWN, A. D., AND L. J. BOROWITZKA. 1979. Halo- EMERSON, R., AND C. M. LEWIS. 1943. The dependence tolerance of Dunaliella , p. 139-190. In M. Levandow- of the quantum yield of Chlorella photosynthesis on sky and S. H. Hutner [ed.] Biochemistry and phys- wave length of light. Am. J. Bot. 30: 165-178.

iology of protozoa. 2nd ed. Vol. 1. Academic Press, ENG - WILMOT, D. L., W. S. HITCHCOCK, AND D. F. New York, NY. MARTIN. 1977. Effect of temperature on the prolif-

304 eration of Gymnodinium breve and Gomphosphaeria GESSNER, F. 1970. Temperature. Plants, p. 362-406. In apottina . Mar. Biol. 41: 71-77. O. Kinne [ed.] Marine ecology. Vol. 1: Environmental EPPLEY, R. W. 1972. Temperature and phytoplankton factors, part I. Wiley Interscience, London. growth in the sea. Fish. Bull. 70: 1063-1085. GESSNER, F., AND W. SCHRAMM. 1971. Salinity. Plants, EPPLEY, R. W., AND W. G. HARRISON. 1975. Phys- p. 705-820. In O. Kinne [ed.] Marine ecology. Vol. I: iological ecology of Gonyaulax polyedra , a red water Environmental factors, part 2. Wiley Interscience, dinoflagellate of Southern California, p. 11-22. In London. V. R. LoCicero [ed.] Proc. 1st Int. Conf. Toxic Dino- GLOOSCHENKO, W. A., H. CURL, AND L. F. SMALL JR. Flagellate Blooms, Nov. 1974. Massachusetts Science 1972. Diel periodicity of chlorophyll-a concentration and Technology Foundation, Wakefield, MA. in Oregon coastal waters. J. Fish. Res. Board Can. 29: EPPLEY, R. W., R. W. HOLMES, AND J. D. H. STRICK- 1253-1259. LAND. 1967. Sinking rates of marine phytoplankton GOLDMAN, J. C. 1977. Temperature effects on phyto- measured with a fluorometer. J. Exp. Mar. Biol. Ecol. planIcton growth in continuous culture. Limnol. 1: 191-208. Oceanogr. 22: 932-936. EPPLEY, R. W., P. KOELLER, AND G. T. WALLACE JR. 1979. Temperature effects on steady-state growth, 1978. Stirring influences the phytoplankton species phosphorus uptake, and the chemical composition of composition within enclosed columns of coastal sea a marine phytoplankton. Microb. Ecol. 5: 153-166. water. J. Exp. Mar. Biol. Ecol. 32: 219-239. GOLDMAN, J. C., AND E. J. CARPENTER. 1974. A kinetic EPPLEY, R. W., J. N. ROGERS, AND J. J. MCCARTHY. approach to the effect of temperature on algal growth. 1969. Half-saturation constants for uptake of nitrate Limnol. Oceanogr. 19: 756-766. and ammonium by marine phytoplankton. Limnol. GOLDMAN, J. C., AND R. MANN. 1980. Temperature- Oceanogr. 14: 912-920. influenced variations in speciation and chemical EPPLEY, R. W., J. N. ROGERS, J. J. MCCARTHY, AND composition of marine phytoplankton in outdoor mass A. SoURNIA. 1971. Light/dark periodicity in nitrogen cultures. J. Exp. Mar. Biol. Ecol. 46: 29-39. assimilation of the marine phytoplankters Skeletonema GOLDMAN, J. C., AND J. J. MCCARTHY. 1978. Steady costatum and Coecolithus huxleyi in N-limited chem- state growth and ammonium uptake of a fast-growing ostat culture. J. Phycol. 7: 150-154. marine diatom. Limnol. Oceanogr. 23: 695-703. EPPLEY, R. W., J. H. SHARP, E. H. RENGER, M. J. GOLDMAN, J. C., AND J. H. RYTHER. 1976. Temperature- PERRY, AND W. G. HARRISON. 1977. Nitrogen as- influenced species competition in mass cultures of similation by phytoplankton and other microroganisms marine phytoplankton. Biotechnol. Bioeng. 18: in the surface waters of the Central North Pacific 1125-1144. Ocean. Mar. Biol. 39: 111-120. GRAN, H. H. 1929. Investigation of the production of EPPLEY, R. W., AND J. D. H. STRICKLAND. 1968. plankton outside the Romsdalsfjord 1926-1927. Rapp. Kinetics of marine phytoplankton growth, p. 23-62. P.V. Réun. Cons. Int. Explor. Mer 56: 1-112. In M. R. Droop and E. J. Ferguson-Wood [ed.] GRANT, W. S., AND R. A. HORNER. 1976. Growth Advances in microbiology of the sea. Vol. 1. Academic responses to salinity variation in four arctic ice diatoms. Press, New York, NY. J. Phycol. 12: 180-185. FALKOWSKI, P. G. 1981. Light-shade adaptation in marine GRUMBACH, K. H., H. K. LICHTENTHALER, AND K. H. phytoplankton, p. 99-119. In P. G. Falkowski [ed.] ERISMANN. 1978. Incorporation of "CO 2 in photo- Primary productivity in the sea. Environmental Science synthetic pigments of Chlorella pyrenoidosa . Planta Research Series. Vol. 19 Plenum Press, New York, 140: 37-43. NY. GUILLARD, R. R. L. 1962. Salt and osmotic balance, p. FALKOWSKI, P. G., AND T. G. OWENS. 1978. Effects of 529-540. In R. A. Lewin [ed.] Physiology and bio- light intensity on photosynthesis and dark respiration chemistry of algae. Academic Press, New York, NY. in six species of marine phytoplankton. Mar. Biol. GUILLARD, R. R. L., AND P. KILHAM. 1977. The ecology 45: 289-295. of marine phytoplanktonic diatoms, p. 372-469. In FINENKO, Z. Z. 1978. Production in plant populations, D. Werner [ed.] The biology of diatoms. Blackwell, p. 13-87. In O. Kinne [ed.] Marine ecology. Vol. 4: Oxford. Dynamics. Wiley Interscience, London. GUILLARD, R. R. L., AND J. H. RYTHER, 1962. Studies FOGG, G. E., W. D. P. STEWART, P. FAY, AND A. E. of marine planktonic diatoms. I. Cyclotella nana WALSBY. 1973. The blue-green algae, Academic Hustedt and Detonula confervacea (Cleve) Gran. Can. Press, London. 459 p. J. Microbiol. 8: 229-239. FOY, R. H., C. E. GIBSON, AND R. V. SMITH. 1976. HALIM, Y. 1960. Observations on the Nile bloom of phyto- The influence of daylength, light intensity and tem- planIcton in the Mediterranean. J. Cons. Int. Explor. perature on the growth rates of planktonic blue-green Mer 26: 57-67. algae. Br. Phycol. J. 11: 151-163. HALLDAL, P. 1958. Pigment formation and growth in FRECHETTE, M., AND L. LEGENDRE. 1978. Photosyn- blue-green algae in crossed gradients of light intensity thèse phytoplanctonique : réponse à un stimulus simple, and temperature. Physiol. Plant. 11: 401-420. imitant les variations rapides de la lumière engendrées HALLDAL, P. [ed.] 1970. Photobiology of microorganisms. par les vagues. J. Exp. Mar. Biol. Ecol. 32: 15-25. Wiley Interscience, London. 479 p. GA VIS, J. 1976. Munk and Riley revisited: nutrient diffusion HARRISON, W. G. 1976. Nitrate metabolism of the red transport and rates of phytoplankton growth. J. Mar. tide dinoflagellate Gonyaulax polyedra Stein. J. Exp. Res. 34: 161-179. Mar. Biol. E,col. 21: 199-209.

305 HATTORI, A. 1962. Light induced reduction of nitrate, and temperature. J. Fish. Res. Board Can. 21: 139- nitrite and hydroxylamine in a blue-green alga Ana- 157. baena cylindrica . Plant Cell Physiol. 3: 355-369. JONES, T. W., AND R. A. GALLOWAY. 1979. Effect HAXO, F. T., AND L. R. BLINKS. 1950. Photosynthetic of light quality and intensity on glycerol content in action spectra of marine algae. J. Gen. Physiol. 33: Dunaliella tertiolecta (Chorophyceae) and the relation- 389-422. ship to cell growth/osmoregulation. J. Phycol. 15: HELLEBUST, J. A. 1976a. Osmoregulation. Annu. Rev. 101-106. Plant Physiol. 27: 485-505. JORGENSEN, E. G. 1964. Adaptation to different light 1976b. Effect of salinity on photosynthesis and intensities in the diatom Cyclotella meneghiniana mannitol synthesis in the green flagellate Platymonas Kütz. Physiol. Plant. 17: 136-145. suecic a Can. J. Bot. 54: 1735-1741. 1968. The adaptation of plankton algae. Il. 1978. Uptake of organic substrates by Cyclotella Aspects of the temperature adaptation of Skeletonema cryptica (Bacillariophyceae): effects of ions, iono- costatum . Physiol. Plant. 21: 423-427. phores and metabolic and transport inhibitors. J. 1969. The adaptation of plankton algae. IV. Light Phycol. 14: 79-83. adaptation in different algal species. Physiol. Plant. HITCHCOCK, G. L. 1980. Influence of temperature on the 22: 1307-1315. growth rate of Skeletonema costatum in response to JORGENSEN, E. G., AND E. STEEMANN NIELSEN. 1965. variations in daily light intensity. Mar. Biol. 57: The adaptation in plankton algae. Mem. 1st. Ital. 261-269. Idrobiol. 18 (suppl.): 37-46. HOBSON, L. A. 1966. Some influences of the Columbia KAHN, N., AND E. SWIFT. 1978. Positive buoyancy river effluent on marine phytoplankton during January through ionic control in the nonmotile marine dino- 1961. Limnol. Oceanogr. 11: 223-234. flagellate Pyrocystis noctiluca Murray ex Schuett. HOLMES, R. W., P. M. WILLIAMS, AND R. W. EPPLEY. Limnol. Oceanogr. 23: 649-658. 1967. Red water in La Jolla Bay, 1964-1966. Limnol. KAMYKOWSKI, D., AND S. J. ZENTARA. 1977. The Oceanogr. 12: 503-512. diurnal vertical migration of motile phytoplankton HOLT, M. G., AND T. J. SMAYDA. 1974. The effect of through temperature gradients. Limnol. Oceanogr. daylength and light intensity on the growth rate of the 22: 148-151. marine diatom Detonula confervacea (Cleve) Gran. Knuss, H. 1974. Osmoregulation in Ochromonas , p. J. Phycol. 10: 231-237. 90-94. In U. Zimmermann and J. Dainty [ed.] Mem- HULBURT, E. M. 1963. The diversity of phytoplanktonic brane transport in plants. Springer Verlag, Berlin. populations in oceanic, coastal, and estuarine regions. 1977. Biochemistry of osmotic regulation. Int. J. Mar. Res. 21: 81-93. Rev. Biochem. (Plant Biochem. II) 13: 119-140. 1970. Competition for nutrients by marine phyto- 1978. Osmotic regulation in algae. Prog. Phyto- plankton in oceanic, coastal, and estuarine regions. chem. 5: 1-27. Ecology 51: 475-484. 1979. Biochemistry of osmotic regulation in 1979. An asymmetric formulation of the distri- Poterioochromonas malhamensis . Ber. Dtsch. Bot. bution characteristics of phytoplankton species: an Ges. 92: 11-22. investigation in interpretation. Mar. Sci. Commun. KIEFER, D. A., AND T. ENNS. 1976. A steady-state . 5: 245-268. model of light-, temperature-, and carbon-limited HULBURT, E. M., AND R. R. L. GUILLARD. 1968. The growth of phytoplankton, p. 319-336. In R. P. Canale relationship of the distribution of the diatom Skele- [cd.] Modeling biochemical processes in aquatic tonema tropicam to temperature. Ecology 49: 337- ecosystems. Ann Arbor Sci. Publ., Ann Harbor, MI. 339. KIRST, G. 0. 1975. Wirkung unterschiedlicher Konzen- HULBURT, E. M., AND J. RODMAN. 1963. Distributions trationen von NaC1 und anderen osmotisch wirksamen of phytoplankton species with respect to salinity Substanzen auf die CO Î Fixierung der einzelligen Alge between the coast of southern New England and Platymonas subconliformis Oecologia 20: 237-254. Bermuda. Limnol. Oceanogr. 8: 263-269. 1979. Osmotic adaptations of the marine plankton HUNTSMAN, S. H., AND R. T. BARBER. 1977. Primary alga Platymonas subcordiformis (Hazen). Ber. Dtsch. production off Northwest Africa: the relationship to Bot. Ges. 92: 31-42. wind and nutrient conditions. Deep-Sea Res. 24: KNOECHEL, R., AND J. KALFF. 1975. Algal sedimen- 7-23. tation: the cause of a diatom-blue-green succession. HUTCHINSON, G. E. 1967. A treatise on limnology. Vol. Verh. Int. Ver. Limnol. 19: 745-754. 2: Introduction to lake biology and the limnoplankton. KREMER, J. N., AND S. W. NIXON. 1978. A coastal Wiley, New York, NY. 1115 p. marine ecosystem. Simulation and analysis. Ecological ICHIMURA, S. E. 1960. Photosynthesis pattern of natural Studies No. 24. Springer-Verlag, Berlin. 217 p. phytoplankton relating to light intensity. Bot. Mag. KREUZER, H. P., AND H. KAUSS. 1980. Role of a galac- Tokyo 73: 458-467. tosidase in the osmotic regulation of Poterioochro- IGNATIADES, L., AND T. J. SMAYDA. 1970. Autoecolo- monas malhamensis . Planta 147: 435-438. gical studies on the marine diatom Rhizosolenia LÂNNERGREN, C. 1979. Buoyancy of natural populations fragilissima Bergon. I. The influence of light, temper- of marine phytoplankton. Mar. Biol. 54: 1-10. ature, and salinity. J. Phycol. 6: 332-339. LEVRING, T. 1966. Submarine light and algal shore zo- JITTS, H. R., C. D. MCALLISTER, K. STEPHENS, AND nation, p. 305-318.In R. E. Bainbridge and E. G. C. J. D. H. STRICKLAND. 1964. The cell division rates Rackham [cd.] Light as an ecological factor. Wiley, of some marine phytoplankters as a function of light New York, NY.

306 Li, W. K. W. 1981. Temperature adaptation in phyto- MORRIS, I., AND K. FARRELL. 1971. Photosynthetic rates, plankton: cellular and photosynthetic characteristics, gross patterns of carbon dioxide assimilation and p. 259-279. In P. G. Falkowski [ed.] Primary pro- activities of ribulose diphosphate carboxylase in ductivity in the sea. Environmental Science Research marine algae grown at different temperatures. Physiol. Series. Vol. 19. Plenum Press, New York, NY. Plant. 25: 372-377. Li, W. K. W., AND I. MORRIS. 1981. Photosynthetic MORRIS, I., AND H. E. GLOVER. 1974. Questions on the adaptation and adjustment to temperature in Phaeo- mechanism of temperature adaptation in marine dactylum tricornutum . J. Exp. Mar. Biol. Ecol. (In phytoplankton. Mar. Biol. 24: 147-154. press) MUNK, W. H., AND G. A. RILEY. 1952. Absorption Lw, M. S., AND J. A. HELLEBUST. I976a. Effects of of nutrients by aquatic plants. J. Mar. Res. 11: 215- salinity changes on growth and metabolism of the 240. MUR, L. R., H. J. GONS, AND L. VAN LIERE. 1977. marine centric diatom Cyclotella cryptica . Can. J. Some experiments on the competition between algae Bot. 54: 930-937. in light-limited environments. 1976b. Effects of salinity and osmolarity of the and blue-green bacteria medium on amino acid metabolism in Cyclotella cryp- FEMS Microb. Lett. 1: 335-338. 1978. Competition of the green algae Scenedes- tica. Can. J. Bot. 54: 938-948. mus and the blue-green alga Oscillatoria . Mitt. Int. LOFIMANN, H. 1908. Untersuchungen zur Feststellung Ver. Limnol. 21: 473-479. des vollstândigen Gehaltes des Meeres an Plankton. NAKANISH I, M., AND M. MONSI. 1965. Effect of variation Wiss. Meeresunters. Abt. Kid. 10: 131-370. in salinity on photosynthesis of phytoplankton growing MACISAAC, J. J. 1978. Diel cycles of inorganic nitrogen in estuaries. J. Fac. Sci. Univ. Tokyo Sect. III Bot. uptake in a natural phytoplankton population dominated 9: 19-42. by Gonyaulax polyedra Limnol. Oceanogr. 23: NELSON, D. M., AND H. L. CONWAY. 1979. Effects 1-9. of the light regime on nutrient assimilation by phyto- MACISAAC, J. J., AND R. C. DUGDALE. 1972. Interactions plankton in the Baja California and northwest Africa of light and inorganic nitrogen in controlling nitrogen upwelling systems. J. Mar. Res. 37: 301-318. uptake in the sea. Deep-Sea Res. 19: 209-232. NELSON, D. M., C. F. D'ELIA, AND R. R. L. GUILLARD. MADDUX, W. S., AND R. F. JONES. 1964. Some inter- 1979. Growth and competition of the marine diatoms actions of temperature, light intensity, and nutrient Phaeodactylum it-icon:lawn and Thalassiosira pseudo- concentration during the continuous culture ofNitzschia nana . II. Light limitation. Mar. Biol. 50: 313-318. closteriwn and Tetraselmis sp. Limnol. Oceanogr. 9: NYHOLM, N. 1978. A mathematical model for growth 79-86. of phytoplankton. Mitt. Int. Ver. Limnol. 21: 193- MAESTRINI, S. Y., AND D. J. BONIN. 1981. Allelopathic 206. relationships between phytoplankton species. Can. O'BRIEN, W. J. 1974. The dynamics of nutrient limitation Bull. Fish. Aquat. Sci. 210: 323-338. of phytoplankton algae: a model reconsidered. Ecology MAHONEY, J. B., AND J. J. A. MCLAUGHLIN. 1979. 55: 135-141. Salinity influence on the ecology of phytoflagellate PAASCHE, E. 1975. The influence of salinity on the growth blooms in lower New York Bay and adjacent waters. of some plankton diatoms from brackish water. Norw. J. Exp. Mar. Biol. Ecol. 37: 213-223. J. Bot. 22: 209-215. MALONE, T. C., C. GARSIDE, K. C. HAINES, AND PACKARD, T. T., AND D. BLASCO. 1974. Nitrate reductase O. A. ROELS. 1975. Nitrate uptake and growth of activity in upwelling regions. 2. Ammonia and light Chaetoceros sp. in large outdoor continuous cultures. dependence. Téthys 6: 269-280. Limnol. Oceanogr. 20: 9-19. PASCIAK, W. J., AND J. GAVIS. 1974. Transport limitation MARRA, J. 1978. Effect of short-term variations in light of nutrient uptake in phytoplankton. Limnol. Oceanogr. intensity on photosynthesis of a marine phytoplankter: 19: 881-888. a laboratory simulation study. Mar. Biol. 46: 191- 1975. Transport limited nutrient uptake rates in 202. Ditylum brightwellii Limnol. Oceanogr. 20: 604- McCommE, A. M. 1960. Actions and interactions of 617. temperature, light intensity and nutrient concentration PAUL, J. S. 1979. Osmoregulation in the marine diatom on the green alga, Chlamydomonas reinhardi . J. Cylindrotheca fusifonnis . J. Phycol. 15: 280-284. Fish. Res. Board Can. 17: 871-894. PLATT, T., K. L. DENMAN, AND A. D. JASSBY. 1977. MCLACHLAN, J. 1960. The culture of Dunaliella tertiolecta Modeling the productivity of phytoplankton, p. 807- Butcher - a euryhaline organism. Can. J. Microbiol. 856. In E. D. Goldberg et al. [ed.] The sea: ideas and 6: 367-379. observations on progress in the study of the seas. 1961. The effect of salinity on growth and chlo- Vol. 6. Wiley, New York, NY. rophyll content in representative classes of unicellular PODAMO, J. 1972. Contribution à l'étude biologique et marine algae. Can. J. Microbiol. 7: 399-406. chimique du port d'Ostende. Ann. Soc. R. Zool. Belg. MIDDLEBROOKS, E. J., AND D. B. PORCELLA. 1971. 102: 105-127. Rational multivariate algal growth kinetics. J. Sanit. PRÉZELIN, B. 1981. The light reaction of photosynthesis. Eng. Div. Proc. Am. Soc. Civ. Eng. 97 (SAI): 135- Can. Bull. Fish. Aquat. Sci. 210: 1-43. 140. PREZELIN, B. B., AND H. A. MATLICK. 1980. Time- MORGAN, K. C., AND J. KALFF. 1979. Effect of light course of photoadaptation in the photosynthetic-ir- and temperature interactions on growth of Coptomonas radiance relationship of a dinoflagellate exhibiting erosa (Cryptophyceae). J. Phycol. 15: 127-134. photosynthetic periodicity. Mar. Biol. 58: 85-96.

307 PRÉZELIN, B. B., AND B. M. SWEENEY. 1979. Photo- II. Der Einfluss des Salzgehaltes. Mar. Biol. 27: adaptation of photosynthesis in two bloom-forming 287-298. dinoflagellates, p. 101-106. /n D. L. Taylor and A. N. SELIGER, H. H., J. H. CARPENTER, M. LOFTUS, AND Seliger [ed.] Toxic dinoflagellate blooms. Proc. 2nd W. D. MCELROY. 1970. Mechanisms for the accu- Int. Conf. Toxic Dinoflagellate Blooms. Elsevier, mulation of high concentrations of dinoflagellates in Amsterdam. a bioluminescent bay. Limnol. Oceanogr. 15: 234- PROVASOLI, L. 1958. Nutrition and ecology of protozoa 245. and algae. Annu. Rev. Microbiol. 12: 279-308. SMAYDA, T. J. 1969. Experimental observations on the RAMUS, J., S. I. BEALE, D. MAUZERALL, AND K. L. influence of temperature, light, and salinity on cell HOWARD. 1976. Changes in photosynthetic pigment division of the marine diatom Detonula confervacea concentration in seaweeds as a function of water depth. (Cleve) Gran. J. Phycol. 5: 150-157. Mar. Biol. 37: 223-229. 1970. The suspension and sinking of phyto- RAMUS, J., F. LEMONS, AND C. ZIMMERMAN. 1977. plankton in the sea. Oceanogr. Mar. Biol. Annu. Rev. Adaptation of light-harvesting pigments to downwel- 8: 353-414. ling light and the consequent photosynthetic perfor- SOEDER, C. J. 1967. Tagesperiodische Vertikalwanderung mance of the eulittoral rockweeds Ascophyllum bei begeisselten Planktonalgen. Umsch. Wiss. Tech. nodosum and Fucus vesiculosus . Mar. Biol. 42: 67: 388. 293-303. SOEDER, C., AND E. STENGEL. 1974. Physico-chemical RAYMONT, J. E. 1963. Plankton and productivity in the factors affecting metabolism and growth rate, p. 714- oceans. Pergamon Press, Oxford. 660 p. 740. In W. D. P. Stewart [ed.] Algal physiology RESHKIN, S. J., AND G. A. KNAUER. 1979. Light stimu- and biochemistry. Blackwell, Oxford. lation of phosphate uptake in natural assemblages of SOURNIA, A. 1981. Morphological basis of competition phytoplankton. Limnol. Oceanogr. 24: 1121-1124. and succession. Can. Bull. Fish. Aquat. Sci. 210: RHODE, W., R. A. VOLLENWEIDER, AND A. NAU- 339-346. WERCK. 1958. The primary production and standing STEEMANN NIELSEN, E., AND V. K. HANSEN. 1959. crop of phytoplankton, p. 299-322. In A. A. Buzzati- Light adaptation in marine phytoplankton populations Travers° [ed.] Perspectives in marine biology. Univ. and its interrelation with temperature. Physiol. Plant. Calif. Press, Berkeley. 12: 353-370. RICE, T. R., AND R. L. FERGUSON. 1975. Response of STEEMANN NIELSEN, E., V. K. HANSEN, AND E. G. estuarine phytoplankton to environmental conditions, JORGENSEN. 1962. The adaptation to different light p. 1-43. In F. J. Vernberg [ed.] Physiological ecology intensities in Chlorella vulgaris and the time de- of estuarine organisms. Univ. S. Carolina Press. pendence on transfer to a new light intensity. Physiol. RILEY, G. A. 1967. The plankton of estuaries, p. 316- Plant. 15: 505-517. 326. In G. H. Lauff [ed.] Estuaries. Am. Assoc. STEEMANN NIELSEN, E., AND E. G. JORGENSEN. 1968a. Adv. Sci. Publ. 83, The adaptation of plankton algae. I. General part. RIPER, D. M., T. G. OWENS, AND P. G. FALKOWSKI. Physiol. Plant. 21: 401-433. 1979. Chlorophyll turnover in Skeletonema costatum , 1968b. The adaptation of plankton algae. Ill. With a marine plankton diatom. Plant Physiol. 64: 49-54. special consideration on the importance in nature. RYTHER, J. H., AND D. W. MENZEL. 1959. Light adap- Physiol. Plant. 21: 647-654. tation by marine phytoplankton. Limnol. Oceanogr. SWIFT, E., AND V. MEUNIER. 1976. ffects of light 4: 492-497. intensity on division rate, stimulable bioluminescence SAUGESTAD, T. 1971. Microflora in supralittoral rockpools and cell size of the oceanic dinoflagellates Dissodinium in a costal area south of Bergen, Norway. Sarsia 46: lunula , Pyrocystis fusiformis and P. noctiluca . J. 79-95. Phycol. 12: 14-22. SCHOBERT, B. 1977. Is there an osmotic regulatory mech- SYRETT, P. J. 1962. Nitrogen assimilation, p. 171-188. anism in algae and higher plants? J. Theor. Biol. 68: In R. A. Lewin [cd.] Physiology and biochemistry 17-26. of algae. Academic Press, New York, NY. 1979. Proline accumulation in Phaeodactylum TAKANO, H. 1963. Diatom culture in artificial sea water. tricornutum and the function of the compatible solutes I. Experiments on five pelagic species. Bull. Tokai in plant cells under water stress. Ber. Dtsch. Bot. Ges. Reg. Fish, Res. Lab. 37: 17-25. 92: 23-30. TALLING, J. F. 1971. The underwater light climate as a 1980a. The importance of water activity and water controlling factor in the production ecology of fresh- structure during hyperosmotic stress in algae and higher water phytoplankton. Verh. Int. Ver. Limnol. 19: plants. Biochem. Physiol. Pflanz. 175: 91-103. 2 14-243 . 1980b. Proline metabolism, relaxation of osmotic THOMAS, W. H., AND A. N. DoosoN. 1974. Effect of strain and membrane permeability in the diatom interactions between temperature and nitrate supply Phaeodactylum tricornutum. Physiol. Plant. 50: on the cell-division rates of two marine phytoflagel- 37-42. lates. Mar. Biol. 24: 213-217. SCHÔNE, H. K. 1972. Experimentelle Untersuchungen TOMAS, C. R. 1978. Olisthodiscus luteus (Chrysophyceae). zur elkologie der marinen Kieselalge Thalassiosira I. Effects of salinity and temperature on growth, rotula . I. Temperatur und Licht. Mar. Biol. 13: motility and survival. J. Phycol. 14: 309-313. 284-291. 1980. Olisthodiscus luteus (Chrysophyceae). V. 1974. Experimentelle Untersuchungen zur Its occurrence, abundance and dynamics in Narra- Ôkologie der marinen Kieselalge Thalassiosira ronda . gansett Bay, Rhode Island. J. Phycol. 16: 157-166.

308 VÀLIKANGAS, I. 1926. Planktologische Untersuchungen analysis and simulation in ecology. Vol. 1. Academic im Hafengebiet von Helsingfors. I. Über das Plankton Press, New York, NY. insbesondere das Netz-Zooplankton des Sommer- WILLIAMS, R. B. 1964. Division rates of salt marsh halbjahres. Acta Zool. Fenn. 1: 1-298. diatoms in relation to salinity and cell size. Ecology WALL, D., AND F. BRIAND. 1979. Response of lake 45: 877-880. phytoplankton communities to in situ manipulations WOOD, E. J. F. 1965. Marine microbial ecology. Chapman of light intensity and colour. J. Plankton Res. 1: 103— & Hall Ltd., London. 243 p. 112. YENTSCH, C. S. 1974. Some aspects of the environmental WALSBY, A. E. 1972. Structure and function of gas vacu- physiology of marine phytoplankton: a second look. oles. Bacteriol. Rev. 36: 1-32. Oceanogr. Mar. Biol. Annu. Rev. 12: 41-75. WETHERELL, D. F. 1963. Osmotic equilibration and YENTSCH, C. S., AND R. W. LEE. 1966. A study of growth of Scenedesmus obliquus in saline media. photosynthetic light relations, and a new interpretation Physiol. Plant. 16: 82-91. of sun and shade phytoplankton. J. Mar. Res. 24: WHITE, A. W. 1978. Salinity effects on growth and toxin 319-337 , content of Gonyaulax excavata , a marine dinoflagellate YoDER, J. A. 1979. Effect of temperature on light-limited causing paralytic shellfish poisoning. J. Phycol. 14: growth and chemical composition of Skeletonema 475-479. costatum (Bacillariophyceae). J. Phycol. 15: 362— WILLIAMS, F. M. 1971. Dynamics of microbial popu- 370. lations, p. 197-267. In B. C. Patten [ed.] Systems

309 The Role of Phytohormones and Vitamins in Species Succession of Phytoplankton

D. J. BONIN AND S. Y. MAESTRINI Station Marine d'Endowne, Chemin de la Batterie des Lions, F 13007 Marseille, France

AND J. W. LEFTLEY Dunstaffnage Marine Research Laboratory, P.O. Box 3, °ban, Argyll PA34 4AD, Scotland

Introduction sion, blossoming of flowers, development of fruit, increase of roots and shoots, and control of dormancy It is very difficult to give a good definition of in seeds. These substances can be separated into for growth-promoting substances. All the metabolites three groups: (1) auxins, activating cell growth, produced by one organism that may enhance the example, indole acetic acid (IAA) and its numerous growth of others, whatever the nature of their action analogues and derivatives; (2) gibberellins, for might be, are included in this family. Nevertheless, example, gibberellic acid (GA) and some analogues it is possible to separate these organic compounds acting principally on cell growth but also on division into three large groups: (1) organic compounds that rate and maturation of organs; (3) kinetins, acting can be used directly by organisms as sources of particularly on cell division. energy, carbon, or other elements when they are As soon as auxins were characterized by lacking in inorganic form in the environment; (2) Thimann (1934) and their action as growth factors compounds that are solubilizers or chelators and make became established in plants, it was tempting to inorganic nutrients more available; (3) organic com- ascribe various positive interactions, observed pounds that participate in metabolic reactions and are between different algae in culture, to a common factor usually essential for algal growth: they are of vitamin of auxinic nature. As a matter of fact, this hypothesis or plant hormone nature. In this review, we shall remained unproven until the 1960s when liberation discuss the role of metabolites belonging only to the of several auxinlike substances by cultured algae was third group. Because the relative need for each nutrient verified, for example, the work of Bentley (1958, is different for each phytoplankter present at any one 1959) on Chlorella pyrenoidosa, Oscillatoria sp., time, all these substances lead to a modification of the and Anabaena cylindrica and that of Tauts and nutritional characteristics of the medium and, there- Semenenko (1971) on Chlorella sp. Simultaneously, fore, contribute to the modification of the growth of a large number of substances of hormonal origin several phytoplankton species. Consequently, , they have been characterized in unicellular and, more may interact with other nutritional and physical factors frequently, , multicellular marine and freshwater in the species succession. algae. It was evident that if these substances exist Growth-promoting substanCes such as hormones in the cytoplasm, they could also exist in that fraction and vitamins are compounds of various origins and of organic matter, sometimes very large, which is are very different in structure and function. Some excreted by algae during growth. The size of this algae (auxotrophs) are absolutely unable to synthesize fraction differs according to various authors and some of their own growth factors and depend wholly the experimental conditions used. For example, it on the medium for these requirements. Auxotrophy can vary from 7 to 50% of the biomass in fresh- does not relate to the sources of energy used by the water (Fogg 1966) and from 35 to 40% of the carbon algae and some obligate photoautotrophic algae, as assimilated in seawater (Antia et al. 1963). well as heterotrophic algae, may need vitamins. In this respect, the first significant work was On the other hand, exogenous plant hormones are that of Van Overbeek (1940a, b) on some 15 algae, never absolutely necessary. However, in some cases, and particularly on a mixture of diatoms (Melosira mainly in seaweeds but sometimes also in unicellular sp. and Biddulphia sp.). Van Overbeek demon- planktonic algae, it appears that they can modify strated that all the algae studied contained more or several aspects of growth. less high quantities of substances producing an auxinlike response with the coleoptile test. Later on, it was verified that some of the substances really Hormones are auxins having biochemical characteristics similar to those of IAA and that they are more abundant In higher plants, hormones are characterized in growing algae than in senescent algae (Schiewer by effecting improvements in growth and cell divi- 1967, on Enteromolpha; Augier 1965, on Botryo- 310 cladia botryoides). Also, the existence of gibberellin- doubling rate of the total biomass of the culture, as like substances has been demonstrated in Tetrasebnis , these hormones act on cell size as demonstrated in Fucus spiralis , and numerous diatoms from natural several experiments (Bentley-Mowat 1967; Bonin plankton samples (Mowat 1963). Likewise, sub- 1969). It is not surprising to find in the literature stances closely related to cytokinins have been found contradictory and even directly opposite conclusions. by Bentley-Mowat and Reid (1967) in Gymnodinhun Thus, working at the same time, but separately, splendens and Phaeodactylum tricomutum . These Brannon and Sell (1945) and Manos (1945) found three chemical families do not appear as frequently exactly conflicting results in their studies on the in algae. Indeed, analyzing the data presented by influence of IAA at a 10 mg • L-1 concentration on Augier (1977) in his review of the subject, and taking Chlorella . When an effect is observed after addition of into account the difficulty of interpreting the indefi- auxin, it is often an inhibitory action for the highest nite results given by different authors, we can state concentrations and no action at all with the lowest that, among 416 assays for determination of growth ones. Augier (1976) gathered results on the influence substances in algae, one finds 127 indoliclike sub- of exogenous phytohormones on the growth of dif- stances (IAA, analogues, derivatives, and precur- ferent algae. The same statement can be made about sors), 18 gibberellinlike compounds, and 12 cyto- these data as that made by Brian (1963), i.e. that kininlike ones. Are these numbers valid? Surely not, responses to phytohormones are less characteristic in and for several reasons. These substances are ex- unicellular algae than in coenocytic algae and in tremely photo- and thermo-labile; their extraction seaweeds. Among 72 IAA positive-reacting strains, can lead to a significant diminution or even to the total 21 are unicellular and 51 multicellular, and among 24 disappearance of activity in the extract. The difficulty IAA nonreactive strains, 17 are unicellular and 7 are of extraction is increased by the fact that hormones multicellular. On the other hand, the different families are distributed diffusely throughout the whole algal of products do not act with the same relative frequency cell and are not concentrated in glands, as in animals. on all the groups of algae: on diatoms, IAA is less often Therefore, it is necessary to use a complete thallus to stimulatory than GA, and on Chlorophyceae, IAA is obtain a small amount of active substance in the final beneficial as frequently as GA; but on these two groups extract; the same applies to unicellular algae. In these kinetins do not usually have a pronounced action. conditions, chemical analysis leads to a precise deter- Are these results understandable, given that they mination of the nature of the molecule but gives un- are obtained in nutrient media that are so varied, reliable information about the real quantity in the but at the same time always excessively rich? Algeus algae. On the other hand, biological tests are very (1946) had already noted that the results could be sensitive for quantitative evaluation but give no in- very different according to the medium used and formation on the chemical nature of the product. It is the culture conditions. On the other hand, are these not surprising to find IAA and its derivatives in diverse unialgal experiments significant for the ecologist? groups of algae as these substances are known to be On this subject, Johnston's (1963) interesting work produced during metabolism of the important aromatic has to be mentioned. He compared growth of the amino acid, tryptophan. If investigation and extraction different components of a natural planktonic sample methods could be improved, indoliclike compounds exposed to various concentrations of gibberellins. probably could be found in many plant cells. He concluded that there is very often "no significant More important for the ecologist is the aim to effect" but in some cases beneficial effects can be define whether an exogenous source of phytohormone discerned on several of the algae studied. After 8 d can effect improvements in growth processes. There of culture a change appears in the dominance of algae; are numerous reports of the action of various concen- for example, GA, at 0.1 mg • L- ' promotes growth trations of auxins and other phytohormon'es on algae of Ditylum brightwelli more than that of Nitzschia in culture, but their interpretation is very difficult. delicatisshna The early studies were too imprecise, both in the Another subject of interest in Johnston's work description of the experiments themselves and in was a simultaneous experiment on 10 replicate sam- the presentation of the results; many of them must ples for the same GA concentration, which produced be disregarded. Most were realized with nonaxenic very scattered results. Are such conclusions valid? cultures. The extracts were used directly such that What can we think of other studies carried out without the measured growth could have resulted from action any statistical basis? Virtually no results are pre- of the solvent (ethanol) itself (Bach and Fellig 1958). sented in kinetic terms for growth measurements, The experimental conditions (nutrition of the algae, so no real conclusion can be given on the exact light, and temperature) are often imprecisely speci- influence of addition of exogenous phytohormones. fied. Results are also given in an ambiguous form; for On the contrary, there are no studies showing that instance, improvement of growth can be expressed a total lack of hormones in the medium stops growth. either as an increase of cell division rate or as a In other words, it means that all the algae studied

311 are able to synthesize their own phytohormones or auxinlike substances found in seawater are unlikely that their metabolic reactions do not require the to have any action on the growth of this alga. In presence of hormones. littoral zones, where plant biomass is composed One more problem has to be discussed; it con- predominantly of seaweeds, growth-promoting sub- cerns the concentration of hormone for which an stances have been detected. Augier (1972) found increase in growth is observed. Most workers have IAA and GA in the midlittoral zone of the Medi- used high concentrations based on those active in terranean shore. Likewise, Pedersen and Fridborg higher plants. Therefore, data obtained with algae may (1972) noted an improvement in growth of seaweeds not reflect the real potential of the hormones as might cultivated in a medium containing seawater from be revealed at lower concentrations. For any one given the Fucus—Ascophyllum zone and Pedersen (1973) substance, the beneficial action appears at different isolated a cytokinin from these waters. The culture concentrations according to various authors and the improvement could really be due to this cytokinin. algae they have studied. The sensitivity of a clone to a Consequently, one may think that regulation by such certain substance can be large. For example, Kim and substances can occur in areas where plant materials Greulach (1961) showed that IAA at 1 mg • L- ' had no are important. But, according to the few results effect on Chlorella pyrenoidosa, at 5 mg • it we collected, this action is less in unicellular algae favored growth, but at 20 mg • ' it was quite toxic. than in seaweeds. Also, when the same species has been studied by other Is it possible to make useful conclusions if the authors, the results have been quite different. On the various extracts have shown positive effects on basis of the data collected by Augier (1976), it is growth, but have not been chemically characterized? possible to find differences ranging from 10 gg • ' Provasoli and Carlucci (1974) noted that there is no (Gray 1962) to 100 mg • L- ' (Bendana and Fried 1967) proof that algae normally employ exogenous higher for concentrations of GA improving growth of Chlo- plant hormones to regulate their biochemical pro- re/la pyrenoidosa. cesses. As a matter of fact, land plants respond It would be interesting to find whether differ- differently to various forms of auxins, gibberellins, ences exist between freshwater and marine algae and cytokinins. It must be the same for algae. The concerning their requirement for hormones. Is this only means to answer these questions would be a requirement dependent on the original biotope? An biochemical analysis of those algal extracts that show answer cannot be given to this question because most a positive effect on growth. It seems that physiologists authors do not specify the ecological characteristics have focussed too much on a systematic search for of the strain they study. Bentley (1958) and Johnston previously known substances associated with the (1959, 1963) admitted that unicellular marine algae phanerogams. are generally very sensitive to auxins and antime- tabolites. As a matter of fact, concentrations of 100 and even 10 gg • L- ' of GA lead to inhibition of the Vitamins growth in several unicellular marine algae although higher concentrations of the same compound favor GENERAL FEATURES growth of Chlorella pymnoidosa; that is what It has been known for many years that vitamins Johnston's (1963) work demonstrated. With Skeleto- play a positive role in the growth of microorganisms in nema costatum, Bentley (1958) found beneficial culture. This idea was first proposed by Lwoff and concentrations of IAA-like substances ranging from Lederer (1935) and it was verified by Lwoff and Dusi 10 to 100 ng • L-1 . Bonin (1969) mentioned that (1937) who demonstrated the pyrimidine and thiazole concentrations of 100 ng •1_,-1 of IAA and 1 gg • L- ' requirements of the flagellate Polytomella caeca. of GA favored growth of Chaetoceros lauded. Thus, Progress in purification and synthesis allowed it seems that planktonic marine algae are sensitive the demonstration that some form of vitamin require- to concentrations ranging from 10 ng • 1_,-1 to ment is widespread among algal taxa (see reviews 1 gg • L- . At lower values, hormones, principally by Droop 1962; Provasoli 1963, 1971; Provasoli indolic ones, have no apparent effect. and Carlucci 1974; Berland et al. 1978; Swift 1981). It is interesting to know the concentrations Hence, we shall not give too many details here of these substances in natural waters. Bentley (1960) concerning the frequency of this requirement, but found some compounds in offshore waters with only some examples: Droop (1962), reviewing 179 characteristics close to those of IAA. They were species mentioned in the literature, noted that 95 are biologically tested with the coleoptile test and they auxotrophic, 80% of which depend on B 12, 53% on appeared to be present at a very low concentration, thiamine, and 10% on biotin. Subsequently, Provasoli around 3 ng • L-1 , i.e. about 10 times less than the and Carlucci (1974) reviewed a larger group of species lowest values known to give a positive effect on (388) and found 203 auxotrophs, 85% of which re- growth of Skeletonema costatum. In such a case, quired 1312, 40% thiamine, and 7% biotin. These

312 results are similar to those of Droop, and all the authors or is required at lower concentrations, if methionine cited note that other water-soluble vitamin require- is added to the culture medium. The explanation of ments are much rarer. If requirements are classified this phenomenon is that B 12 is an essential cofactor according to taxonomy, one also finds significant in transmethylation reactions in the biosynthesis of differences between the algal groups. Numerous cen- certain amino acids, but the B ,o requirement disap- tric diatoms, half of the pennate diatoms, most of pears if such amino acids are supplied directly to Haptophyceae, 90% of Dinophyceae, only 20% of the alga. Rahat and Reich (1963) demonstrated a Cyanophyceae, and just a few Chlorophyceae require related phenomenon in Prymnesium parvum; vitamin B i o. Exogenous thiamine is less frequently methionine could not replace B 12 or reduce the required in these groups, and biotin very rarely, except requirement for the vitamin. However, in the presence in some Dinophyceae and Chrysophyceae. Hitherto, of vitamin B 1 ., , methionine counteracted the inhi- no diatoms or Haptophyceae have seemed to be de- bition of growth caused by some 13 12 analogues pendent on biotin (see Swift 1981). substituted at the benzimidazole part of the molecule, These data have to be considered as approximate but no such effect was observed against inhibition and relative indices. As a matter of fact, in nature by other analogues. a-(5-hydroxybenzimidazoly1)- the percentages of these requirements may be modi- cobamide cyanide (factor III) replaced B 12 in the fied for several reasons. In the earliest experimental presence of methionine and, to a lesser extent, in studies, the workers often made a selection of species the presence of other methyl donors such as betaine. before undertaking the experiments. Also, very rich These workers concluded that factor III was capable media were used that could modify strongly the exact of replacing 13 1 2 in most 13 1 2 -requiring biochemical nutritional status of the species as found in nature. pathways in Prynmesium except for methyl-group Swift (1981) points out that green algae are very synthesis. Although these two examples may not often isolated in media without organic matter such be exceptional, they allow us to suppose the existence that their vitamin requirements tend to be statistically of such phenomena in many other algae. But it is underestimated. On the other hand, Droop (1962) not universal as Van Baalen (1961) demonstrated that, thought that vitamin requirements were overestimated in Synechocystis sp., methionine could not substitute because, when authors gathered data, they did not for B 12 . This could simply mean that the alga is pay enough attention to the possible negative effects impermeable to this amino acid. Hutner et al. (1957) of vitamins, which were not, therefore, always very also mentioned that B , 2 requirements in Ochromonas clearly expressed in the literature. Some data are vary strongly according to temperature, but they could difficult to interpret because they come from experi- not give any explanation for this phenomenon. ments carried out with a mixture of vitamins, some of which are well characterized whereas others are not. Moreover, some strains can require simultaneously SPECIES-SPECIFIC REQUIREMENTS two, three, or possibly more vitamins, a phenomenon Algae do not have the same requirements with that complicates the survey of the requirements of the respect to the molecular structure of the vitamin algae; for example, Prorocentrum micans needs B offered to them. For instance, it has been demon- and biotin, and Amphidinhun carted and Gyinno- strated that, with thiamine, several microorganisms dinium breve require B 1 0, thiamine, and biotin. require only the thiazole moiety of the molecule A pertinent question which may be posed is: whereas others require only the pyrimidine. Some Are vitamin requirements, demonstrated in a given algae need both parts of the molecule even if given species, always constant or do they depend on the separately whereas others require the complete com- experimental conditions? For example, Herdman and pound. Lewin (1961) showed that among 41 algal Carr (1972) had greater success than previous workers strains he studied, only one required the complete in isolating auxotrophic strains of Anacystis nidulans molecule. The exact form of this requirement may be because they used different mutagenic and culture very important ecologically because the quantities techniques. Ideally, research on vitamin requirements needed depend on the kind of requirement. Droop should be done on newly isolated strains in which (1958) noted that algae which require thiazole only natural characteristics still exist. In fact this is need a greater quantity of the vitamin than those which rarely done because, most frequently, scientists try need only the pyrimidine moiety. to improve their knowledge of nutritional require- The growth response towards vitamin B 12 ments of strains about which information already more complex. The () is even exists. cobalamins consist of a cobalt-containing ring Vitamin requirements also vary according to nucleus to which a nucleotide is attached. Specificity other nutritional factors, as demonstrated by Hutner of organisms towards naturally occurring and artificial et al. (1953). They showed that 13 1 2 is necessary variants of the vitamin is determined by the nature for growth of Ochromonas but is no longer necessary, of the nucleotide. Thus 13 1 2 -requiring organisms fall

313 generally into three groups: (1) those like mammals and Cassie (1963) in such a study indicated values which need a benzimidazole type of base in the ranging from 6.5 to 15 molecules of B 12 for a cubic nucleotide; (2) those like Lactobacillus leichmannii micrometre of cell volume. It seems that these values which can also use an adeninelike base; and (3) vary between experiments, but the variations could those like Escherichia coli which require no more be due to experimental artefacts; Droop (1961) had than the nucleotide-free nucleus. B ,0-requiring already raised this point. These results indicate that unicellular algae fall into one of these three categories the growth limit (in terms of cytoplasmic yield) of specificity, but the "mammalian" response is the allowed by a certain concentration of a given vitamin most frequent (Droop et al. 1959; Provasoli 1971). is nearly constant regardless of species. Nevertheless, However, Guillard (1968) claimed that, at least for Guillard and Cassie thought that this conclusion does diatoms, these response categories are somewhat not deny the role of vitamins in the succession of arbitrary. He tested the growth response of 23 isolates populations. As a matter of fact, the same concen- (21 species) of marine diatoms to B12 analogues tration of vitamin may lead to various rates of metab- compared with B 4 O when supplied at the "ecologically olic synthesis, according to the species, and may also significant" concentration of 4 ng • L -1 . The growth be important in certain stages of the reproductive response was not " all or none" but varied continuously cycle, such as auxospore formation in diatoms. depending on the percentage response (B 1 = 100%) The first studies carried out to evaluate the action chosen by the experimenter. At the 10% level of of vitamin concentrations on growth rate using kinetic response, 11 clones had coliform, 4 lactobacillus, and methods gave encouraging results. Droop (1961) 8 mammalian specificity patterns, whereas at the 1% noticed that division rate in Monocluysis hulled response level 14 appeared to have coliform, 5 lacto- does not change when B 12 concentrations vary from bacillus, and 4 mammalian specificities. To the writers 0.1 to 100 ng • L-1 under constant experimental con- this seems to be a somewhat subjective approach. ditions. If concentrations of B 1 0 are to be limiting, At a given concentration of a given analogue according to Monod's law the B l o concentration must of 13 12 , algae respond differently according to the be lower than 0.1 ng • L-1 . Lewin (1954), Droop presence or the absence of other B 1 2 analogues or (1954), and Cowey (1956), in different neritic and of substances acting as antimetabolites (work of oceanic areas, measured B 4 O concentrations ranging Ford 1958 on Ochromonas malhamensis). A very from 0.2 to 20 ng • L-1 . Therefore, in natural complete study carried out on Euglena gracilis with waters (the vitamin-poorest ones excepted) 13 1 0 has 300 different compounds (Epstein and Timmis no regulating effect on the growth rate of Mono- 1963) (purines, pteridines, and nicotinamides) cluysis lutheri, as its concentration is always showed that, for similar or even lower concentra- sufficient for maximum growth. Droop (1961) tions than that of the vitamin, these compounds are compared the B 4 O-dependent growth of illonochowis inhibitory because they interfere with cofactors with that of Ocluvmonas malhamensis, using data closely concerned with the utilization of B ,0. The obtained for the latter alga by Ford (1958). Ochro- analogues show a great variation in their toxicity monas shows a decrease of cell division rate as soon as and in some cases they can even give a positive the 13 ng • L-1 concentration is reached. Although it is response of the lactobacillus type (Jacobsen et al. difficult to compare the growth characteristics of the

1975). two algae, it seems clear that, for B ,. concentrations Taking into account these findings, it is evident ranging from 0.1 to 13 ng • L-1 , Monochtysis would that vitamin requirements can vary for each species have a much higher growth rate than Ochromonas. and even within the same species, depending on Swift and Taylor (1974) also compared growth kine- culture conditions; this has caused many difficulties tics in three marine species according to B 12 concen- in experimental studies. It is nevertheless important tration and observed different characteristics for these to quantify this vitamin requirement for each alga species. The half-saturation constant is 0.39 (ng • L-1 ) as it tells us whether the growth factor may play for Thalassiosim pseudonana, 1.69 for Isocluysis a direct role in the species succession, depending galbana , and 2.77 for Monocluysis lutheri ,i.e. for the on the concentrations found in situ. The subject last one, slightly higher values than those mentioned has been approached indirectly through vitamin by Droop. With respect to B 12 (concentrations of about bioassay. Indeed, such measurements are based on 1 ng• L-1 found in marine waters) Swift and Taylor the fact that experimentally, at least in given limits, thought that, at a certain phase of the life cycle, the one can find a good linear relationship between the development of these three species could be limited by concentration of the growth factor tested and the availability of 13 cells observed at the end of the experiment. number of VITAMIN B. BINDING PROTEIN In this way, it has been possible to establish the relation between the number of molecules of vitamin Many unicellular algae, and other microorgan- used by a single cell of a given species. Guillard isms, produce an extracellular"binding factor" which

314 strongly sequesters vitamin B 2 . The historical back- follows the same pattern, which is similar to that ground to the discovery and characterization of this shown in bacteria. A primary uptake phase, char- factor has been dealt with in detail by Provasoli and acterized by very rapid uptake, is unaffected by Carlucci (1974) and Pintner and Altmeyer (1979). metabolic inhibitors and is somewhat insensitive to Droop (1968) and Pintner and Altmeyer (1979) temperature. Bradbeer (1971) found that, in Ochro- have shown that extracellular B 12 -binder is produced monas, this primary phase showed saturation kinetics, by a wide range of algae (diatoms, cryptophytes, with a K,,, (B ) of 4 nM and a V,„„, of 5 x 10 ' B 12 chrysophytes, dinoflagellates, and green algae) molecules • s - ' • cell' . In contrast, Sarhan et al. including organisms that are not auxotrophic for the (1980) found that although the initial uptake of B 12 vitamin. Culture filtrates containing the factor will by Euglena showed saturation kinetics, these did inhibit the growth of 13 12 -requirers when added to not conform to the Michaelis-Menten equation. The fresh culture medium. The binder produced by each primary phase is followed by a secondary phase class of algae seems to have the same general pro- where the uptake rate is somewhat slower, and the perties: it is heat labile, inhibition of the growth process is sensitive to both metabolic inhibitors and of B ,o-requirers is nonspecific, and the inhibition to temperature. The primary phase has been inter- can be abolished by adding excess B 1 ., . preted as corresponding to the binding of 13 l o to specific receptors, and the secondary To date, only the B r, binding protein from phase as cor- responding to active transport of B12, Euglena gracilis has been biochemically character- and possibly de novo synthesis ized in any detail (Daisley 1970). Binder from cell of new binding sites, both pro- cesses requiring metabolic energy. extracts, and that excreted into the medium, was purified and found to be of a similar nature, a The ability to accumulate B 12 very rapidly from glycoprotein of molecular weight of about 200 000. the medium, even when it is present at very low The extracellular binder had a greater binding capacity concentrations, is obviously advantageous to an alga. than that from cell extracts, but both preparations of For example, Bradbeer (1971) pointed out that, even binder were heterogeneous, tending to disaggregate at concentrations of 13 12 only 0.01 of the measured during purification, which suggested an oligomeric K„, for initial uptake (4 nM), the minimal amount molecule. of vitamin necessary for a new generation of cells could be taken up in only 3 s by Ochromonas; he The metabolic role of B12 binding protein is also noted that this alga, per cell, stores B 1 2 in bound not clear; the function of the analogous mammalian form about 10" times the minimal requirement, which "intrinsic factor," also a glycoprotein, is to facilitate would permit growth for a subsequent 10 generations transport of 13 12 across the intestinal membrane in absence of vitamin. Also, Sarhan et al. (1980) (Glass 1963). Provasoli and Carlucci (1974) sug- found that Euglena could accumulate 600 times the gested that the protein is involved in active uptake minimal cell requirement of B i o, which was sufficient 12 of B by algae and that any excess is excreted to support 10 subsequent generations of algae. These into the medium. Certainly, non-B, 2 -requirers which workers suggested that this "luxury consumption" produce B ,o-binder can actively take up the vitamin might have an ecological significance in that it would (Droop 1968). However, there is presently no data enable the population to survive during periods of as to whether, or how, algal 13 1 2-binder functions vitamin depletion and also to compete with other in the uptake of the vitamin. auxotrophs that did not have such a high affinity Ford (1958) proposed that B 12 was bound at the for the vitamin. surface of Ochromonas cells, and Droop (1968) The importance of B 12 binding factor in natural assumed that this was also the case for Monocluysis. populations is open to debate. Droop (1968) pointed However, Daisley (1970) prepared antiserum against out that inhibitory effect of binding factor was ap- partially purified binder from Euglena but this failed parent in chemostat experiments with cell concen- to agglutinate Euglena cells, indicating that binding trations of Monochrysis as low as 0.4 million •mL does not take place at the surface of this alga. Sarhan and, because this alga can reach a density of 40 et al. (1980) concluded from their data that the million cells • mIL- ' in supralittoral rock pools, the majority of binding sites in Euglena are associated effect of the binder would probably be significant. He with the chloroplasts. Modern affinity cytochemistry went on to suggest that 13 1 2-binder could also be techniques may help to solve this problem. important in the open ocean. Once the producer of the The physiology of uptake of B l o has been studied binder became dominant, the dominance would be in only a few algae: Ochromonas malhamensis effectively maintained by the inhibitor because other (Reeves and Fay 1966; Bradbeer 1971), Euglena B 12 auxotrophs would be unable to take up the gracilis (Varma et al. 1961; Sarhan et al. 1980), sequestered vitamin whereas the dominant species and Monochrysis ( = Pavlova) !tithed (Droop would have already accumulated sufficient B, 2 to be 1968). In all these algae, uptake of the vitamin almost immune from the effect of the binder.

315 A similar hypothesis was proposed by Provasoli significant and sometimes superior to those of the (1971) to explain the continuing abundance of diatoms producing species. But such an experimental demon- for more or less long periods at the end of their stration might hardly be extrapolated to natural blooms. But, of course, algae prototrophic for B,, conditions. Indeed, nutrient concentrations used in would also be unaffected by presence of the binder. the culture media, about 500 p.g—at N • L - ' and 50 Also, if B, 2 was present in sufficient quantities to tg—at P • L- ' , were considerably higher than in saturate the binder, its effect would be neutralized. natural waters. According to Carlucci and Bowes An additional complication is that we have no data (1970a), if we consider nutrient concentrations 10 concerning the specificity and affinity of algal 13 12- times lower than those in their experiment, it is binder for B 12 derivatives that are probably present possible to calculate that Coccolithus Inixleyi releases in seawater; those algae able to utilize B l o derivatives vitamin B ,o at a concentration of 2.5 ng • L- ' at the end not bound, or only loosely bound, would be at an of growth, or about the same order of magnitude as advantage. found in seawater. Therefore, it is permissible to suppose that such a model would be applicable to ALGAE AS VITAMIN PRODUCERS natural waters. Another discovery made by using experimental Another important aspect of these experiments culture methods was the excretion of vitamins by algal is that liberation of vitamin has been shown to be cells themselves. This challenged early concepts a phenomenon which takes place during exponential because, for a very long time, bacteria were assumed growth and not only when the culture reaches the to be the only vitamin producers in natural media. stationary phase, although, at that stage, the dissolved Indeed, many bacteria do excrete vitamins (Ericson vitamin concentrations are considerably increased. and Lewis 1953; Starr et al. 1957; Burkholder 1959). The rate of vitamin excretion is also a function of It appears, furthermore, that in several biotopes, the culture conditions. Thus, in the preceding strains that can provide one of the three essential experiments it was demonstrated that the quantity vitamins are more numerous than those which need of vitamin released by a unit of producing cells them (Berland et al. 1976). In mixed cultures, it had depends on the concentration of another vitamin been already demonstrated that the vitamin B required by the producer. All biological, biochemical, requirement of several marine diatoms could be and physical factors of the environment interfere satisfied totally by heterotrophic marine bacteria with the capability of the alga to excrete vitamin, isolated from the same biotope (Haines and Guillard and that problem is that much more complicated. 1974). Although the available data did not give any information about the vitamin flux between organisms in natural environments, it was tempting to suppose ALGAE AS COMPETITORS FOR that bacteria were the main producers of those vita- VITAMIN PARTITION mins which became available for auxotrophic micro- The fact that prototrophic algae are known to organisms and especially for algae. But it was known take up vitamins may have some ecological signifi- already that vitamins were excreted by various algae: cance. Droop (1968) observed that Phaeodactilum, biotin by Chlore/la pyrenoidosa (Bednar and which does not require B 12 , took up the vitamin in a Holm-Hansen 1964), nicotinic acid by Chlamydo- similar fashion to 13 1 2 auxotrophs; Carlucci and monas (Nakamura and Gowans 1964), and numerous Bowes (1970b) observed a stimulation of growth of growth factors by Ochromonas danica (Aaronson this alga, which was probably due to B 12 . Subse- et al. 1971). However, demonstrating that algae are quently, Swift and Guillard (1978) demonstrated that able to produce vitamins is one thing, but proving the addition of vitamin B 4 O to the culture medium of that the excretion is sufficient to allow growth of 12 clones belonging to the genera Thalassiosim, other plytoplankters from the sanie biotope is quite Porosira, and Chaetoceros, which are not absolutely another, and needs various experiments using mixed 13 12 -requiring, improved growth by increasing the cultures of vitamin-producing and vitamin-requiring growth rate and decreasing the lag period. These organisms. Carlucci and Bowes (1970b) demon- algae are able to use the growth factor when it is strated that such successions can be obtained in present in the medium, but also excrete the vitamin experimental studies. Thus, Dunaliella tertiolecia when they are grown in B 12 -free medium. Swift and Skeletonenta costanun produced thiamine that and Guillard described this characteristic as "facul- Coccolithus Intyleyi could use, Phaeodactylum tative autotrophy" and suggested that it might help tricornuttun and Skeletonema costatum produced such diatoms to compete well with auxotrophs at biotin which was used by Amphidinitun carterae, the beginning of a bloom. However, further work and Coccolithus lutvleyi excreted B ,0 that Cyclo- needs to be done before we can understand fully tella nana could use. In these three cases, the biomass the phenomenon of facultative autotrophy and reached by the vitamin-requiring algae was very whether it has any ecological significance.

316 ROLE OF VITAMINS IN ALGAL SPECIES studies that these organisms are really auxotrophic SUCCESSION for it.

IS it possible from the experimental data already In other cases, such negative correlations are discussed to postulate the possible influence of vita- not evident and the simultaneous evolution of vita- mins on primary production processes and on species mins and species densities is not resolved. Several succession in natural waters? It has been observed algae are able to produce vitamins and, in this case, that vitamins follow annual cycles in many natural we may observe a positive correlation between the biotopes, but it is much more difficult to elucidate vitamin concentration and the abundance of cells their exact role in succession. Very often, authors' producing them. Ohwada and Taga (1972b) observed conclusions are more hypotheses than assertions. such production in Lake Sagami, Japan, where biotin Interpretation of vitamin bioassays is indeed difficult; and thiamine concentrations increased when Cyclo- even if they do permit an appreciation of the real tella sp., Fragilaria crotonensis, and Synedra acus effect of vitamins, analogues, and derivatives on a became dominant in the plankton; simultaneously given species, they only specify the concentrations B 1 2 values fell. The study of such patterns is not of the active substances as a whole for the organism easy because the capacity of algae to release vitamins used in the test. But, typically, this organism is not in the medium is much less understood than their the most representative of the community in the auxotrophic requirements. natural environment. Bioassays carried out with only Interpreting the simultaneous evolution of one species allow comparison between different vitamins and growth of phytoplankton populations waters, but it is usually prohibitive to increase by is a difficult task, and does not always yield clear-cut very much the number of assay specimens. Conse, results and, hence, reliable conclusions. There are quently, it becomes impossible to characterize every many reasons for this; the correlation could be re- area with the organisms of greatest ecological signif- garded as significant only if there is prior evidence icance. from test organisms for which the vitamin response Menzel and Spaeth (1962) were the first to try is very close to that of the algae well represented to establish the role of vitamin B in species succes- in the natural sample. For instance, Bruno and Staker sion. They found that, in the upper layers of the (1978) could not entirely explain the succession of Sargasso Sea during the spring bloom, B r, concen- populations in Block Island Sound, New York, ac- trations increased with abundance of centric diatoms, cording to variations in vitamin B 12 concentrations. some of which were found to be B ,0 dependent They measured the B 1 , concentrations in these waters (Guillard 1968). Later on in the year when B 12 with Thalassiosira pseudonana 3H, which has a decreases, Coccolithus huxleyi, which concentration totally "mammalian" type of response, although in is only thiamine dependent, becomes dominant in these samples one of the most abundant species was the phytoplankton population. At the same time, in Skeletonema costatum, which has a "coliform" type Long Island Sound, Vishniac and Riley (1961) response. Other vitamin B 1 2 derivatives that could ascertained that the maximum B 1 2 concentration be utilized by Skeletonema would not be measured appears in the middle of the winter and remains at by the Thalassiosira bioassay. In the Gulf of Maine, a high level until a bloom of Skeletonetna costatum Swift and Guillard (1977) found that, in a 150-m and other centric diatoms occurs. water column, the ratio of total cobamides versus Negative correlations between vitamin con- true B 1 2 can vary between 3.3 and 1.4. In fresh- centrations and algal abundance are often encoun- water, Kurata et al. (1976) found even higher ratios tered. Thus, Cattell (1973) found one between B 12 at some stations in Lake Mergozzo, Italy. More at different stations and a dinoflagellate population recently, Sharma et al. (1979) compared radioisotope surveyed in the Strait of Georgia, British Columbia, dilution techniques with bioassays for vitamin B 1 2 showing that B 2 might be used by the phytoplankters. seawater. The radioisotope techniques gave results in Carlucci (1970) observed a similar situation with from 4 to 6 times higher than results obtained by Gonyaulax polyedra off La Jolla, California. Parker standard microbiological assays (two strains of (1977) noticed it during the development of Cyano- Thalassiosira pseudonana were used). Radioisotope phyceae belonging to the genus Oscillatoria in Lake dilution assays can be carried out very much quicker Washington. Likewise, Kurata et al. (1976) found than bioassays, but tend to be less specific as they these inverse relationships between B 1 , and thiamine may determine the sum of B 12 and its derivatives concentrations and growth of small planktonic forms, whether the latter are biologically active or not. essentially represented by Tabellaria flocculosa Sharma et al. suggested that only 25-40% of and Mougeotia sp. in Lake Mergozzo, Italy. Many cobalamin molecules measured by the radioassays more examples could be given. Such observations were vitamin B,,, pseudovitamin B 1 0, and their allow us to assert that the algae use the vitamin only analogues, and that the remaining 60-75% were if we have parallel evidence from experimental transformation products formed by the action of

317 light and seawater on cobalamins. However, they reliable explanation for the role of vitamins in the could not exclude the possibility that this large growth of each component of the crop. Furthermore, fraction might include cobalamins that were not it is always necessary in these observations to make a utilized by the bioassay organisms. This work high- distinction between vitamin B 1 9, thiamine, and lights the difficulty of using highly specific bioassays biotin. The experimental method proves that thiamine to measure a wide spectrum of potentially biologically is not as significant ecologically as B 12 as relatively active molecules. These molecules could be meas- fewer organisms require it, and this statement is even ured chemically (Beck 1978; Kolhouse and Allen more applicable to biotin. 1978) but there would still be the problem of assessing On the other hand, it is difficult to understand which of the B 12 derivatives could be used by each the relationships between the different parameters organism in a natural assemblage. when sampling is not frequent enough; the usual Thus, in any attempt to correlate dissolved time interval between two samplings at sea is about vitamin çoncentrations with species composition a month or 2 wk. An example is given by the and succession, ideally it is necessary to know the experiment of Antia et al. (1963) using a large-volume vitamin requirements of each phytoplankter present plastic sphere. They found a slight decrease of vitamin in the sample. But this is not always practicable. ,,, when Skeletonetna costattun dominated the Under such conditions, it is tempting to use relation- phytoplankton. Afterwards, other pelagic species ships between the vitamin concentration and the followed the S. costattun bloom and, at the end of phytoplankton biomass as a whole, by particulate the experiment, the B 1 2 concentration reached values carbon or chlorophyll a measurements. The results higher than those at the beginning (8 ng • L-1 vs. obtained in this way are not always similar. For 3 ng • ). The chlorophyll a concentrations also example, Benoit (1957) noted that, in lakes, the showed the sanie trend. Comparing parameters at proportion of total cobalt tied up in vitamin B 1 2 is the beginning and at the end of the experiment would about 10% in the epilimnion where production is have led to a totally mistaken interpretation of the maximum and only 4% or less in the hypolimnion. relationship between the phytoplankton standing In an oligotrophic lake (Lake Tahoe), Carlucci and crop and the B 12 concentration. Only high frequency Bowes (1972) found that, usually, concentrations of of sampling allows the authors to give a reliable all three vitamins are low or undetectable in deeper account of the biological sequence. waters. The highest concentrations of vitamins were One more difficulty in interpreting the exact observed in the waters from 60 m, and especially influence of vitamins on algal growth and, therefore, in summer at a time when phytoplankton usually on species succession is the interaction between reaches the highest densities. Similar results are requirements for vitamins and other nutrients. given by Natarajan and Dugdale (1966) for surface Obviously, the limiting action of the vitamin is easier seawater in the Gulf of Alaska where, throughout to demonstrate when other nutrients are not limiting. the year, thiamine concentrations are found to range Often there is an intimate relation between the content from undetectable levels up to 490 ng The of nutrient salts and that of vitamins: that is what amounts of dissolved vitamin are usually high in Kashiwada and Kakimoto (1962) showed when coastal regions and low in the open sea; below the comparing Japanese lakes and marshes belonging to euphotic zone, these values are generally very low the eutrophic, merotrophic, and oligotrophic types. or even undetectable. Ohwada and Taga (1972a) Daisley (1969) observed similar results in the English observed in the North Pacific Ocean and in the China Lake District. The primitive rocky lakes with low Sea that the concentrations of dissolved vitamin B 1 ., , productivity are poor in vitamin B12. The evolved thiamine, and biotin generally showed patterns similar silted lakes, richer in sediments and, therefore, in to those of chlorophyll a, with a maximum in July nutrients and organisms, have higher concentrations and a strong decrease with depth. Usually, vitamin of vitamin. Such results are common in the study B 1 2 is less correlated with chlorophyll a than are of well-defined ecosystems. But in the open sea the other vitamins. But, on the other hand, Daisley the observations are less clear. One way to elucidate and Fisher (1958), in the Bay of Biscay, found low the role of vitamins on production is to measure both B 1 2 concentrations in the upper illuminated zone and the soluble and the particulate fractions of the vita- at the greatest depths, with a maximum at intermediate mins. If the soluble fraction is always significant depth. Likewise, Carlucci and Silbernagel (1966) and if the particulate one is small, it is possible to measured higher concentrations from 200 m and assert that the vitamin is not a limiting factor. For below in the Northeast Pacific Ocean, with maximum example, Ohwada and Taga (1972b) showed that at intermediate depth (average of 500 m), and an particulate thiamine and biotin correspond to only inverse chlorophyll a—vitamin B 1 2 relationship in 1% of that of the dissolved form in the surface waters the upper waters. However, these latter results are of the North Pacific Ocean. In coastal waters, on of little interest to the ecologist as they cannot give a the contrary, they were 144 and 54%, respectively.

318 Obviously, factors other than these two vitamins natural environment, and that is what has been were limiting algal growth in the open sea. In the verified by the measurernents of vitamin concen- same way, Natarajan (1970) mentioned that high trations in natural waters and by experimental studies. concentrations of vitamins were usually found in But, because vitamin concentrations sometimes vary low-nutrient areas. Thiamine particularly showed a greatly in natural waters, a lack or an excess of these negative correlation with PO ,—P and NO:1—N whereas natural compounds can enable some species (ob- vitamin B 12 had a less significant correlation with viously not the same ones) to grow faster and then all the other parameters measured. In this case, such can act directly on the succession. Indeed, in peculiar results indicate that the thiamine content of the sea situations it is possible to observe such a direct is sufficient to support a level of production which interaction between species belonging to bacterio- is then limited by other factors, such as availability and phyto-plankton, and so to explain local succes- of other nutrients. In such areas, where vitamins sion. But, more usually, when vitamins and other are quantitatively sufficient with respect to other nutritional elements interact as limiting factors, such nutrients, they have no effect on production and explanations are more difficult to give. The action only a slight effect on species succession. But vita- of vitamins remains very complex and its quantitative mins, especially II, show low values in waters evaluation difficult. where some macronutrients could be limiting and where physical factors, such as temperature and light, are not optimal for growth. An example is the antarctic and subantarctic waters where it has not yet been Acknowledgments possible to find reliable correlations between the chemical and physical parameters and primary We are grateful to Dr Trevor Platt for an initial production (Carlucci and Cuhel 1977; Holm-Hansen critical review of the manuscript. We also wish to et al. 1977; Fiala 1980). express our appreciation to Drs Henry Augier, Michel Fiala, and Dorothy Swift for helpful discussion. COMMENTS AND PROSPECTS The only approach to understanding the precise role of vitamins among the other limiting factors on References each component of the ecosystem would be to know exactly the physiological responses to vitamins of all AARONSON, S., B. DE ANGELIS, O. FRANK, AND H. the species present in the waters, and not just the most BAKER. 1971. Secretion of vitamins and amino acids abundant species in the crop. It would also help to into the environment by Ochromonas danica. J. know (1) the importance of upwelling from deep Phycol. 7: 215-218. waters and the exact nature of all the analogues carried ALGEUS, S. 1946. Untersuchungen über die Embrungs- physiologie der Chlorophyceen. Mit besonderer up with them, (2) the ability of the growth factors to Berücksichtigung von Indolylessigsaüre, Ascorbin- remain in active form in spite of the potentially saüre und Aneurin. Bot. Not. 2: 129-278. destructive action of physical, chemical, and bio- ANTIA, N. J., C. D. MCALLISTER, T. R. PARSONS, logical agents which vary greatly with climate and K. STEPHENS, AND J. D. H. STRICKLAND. 1963. origin of the waters, (3) the turnover time of the Further measurements of primary production using vitamin molecules in the upper layers, and (4) the a large volume plastic sphere. Limnol. Oceanogr. 8: composition and the physiological characteristics of 166-183. bacterial populations that obviously interfere with the AUGIER, H. 1965. Les substances de croissance chez la response of the phytoplankton. Generally, as long as Rhodophycée Botryocladia bonyoides. C.R. Acad. Sci. Paris 260: 2304-2306. we do not know much about these processes, it will be 1972. Contribution à l'étude biochimique et very difficult to understand the exact function of the physiologique des substances de croissance chez les main growth factors and regulators in the equilibrium algues. Thèse Doct. ès Sci. Univ. Aix-Marseille II. of ecosystems. 323 p. Nevertheless, as Swift (1981) says, auxotrophs 1976. Les hormones des algues. État actuel des are usually important in natural waters. This implies connaissances. III. Rôle des hormones dans la crois- that vitamin concentrations are frequently sufficient sance et le développement des algues. Bot. Mar. 19: to allow maximum growth rates for many of them. 351-377. Another argument in support of this statement is that, 1977. Les hormones des algues. État actuel des if it is not so, the autotrophs that do not require connaissances. V. Index alphabétique par espèce des travaux de caractérisation des hormones endogènes. vitamins would always dominate when competition Bot. Mar. 20: 187-203. occurs. When an organism requires a growth factor BACH, M. K., AND J. FELLIG. 1958. Auxins and their in vitro this metabolite, or its physiological equiva- effect on the growth of unicellular algae. Nature lent, should be found in significant amounts in its (London) 182: 1359-1360.

319 BECK, R. A. 1978. Competitive intrinsic factor binding CARLUCCI, A. F., AND P. M. BOWES. 1970a. Production assay technique for cobalamins in natural waters. of vitamin B, 2 , thiamine and biotin by phytoplankton. Anal. Chem. 50: 200-202. J. Phycol. 6: 351-357. BEDNAR, T. W., AND O. HOLM - HANSEN. 1964. Biotin 1970b. Vitamin production and utilization by liberation by the lichen alga Cocconma sp. and by phytoplankton in mixed culture. J. Phycol. 6: 393- Chlorella pyrenoidosa. Plant Cell Physiol. 5: 297- 400. 303. 1972. Determination of vitamin B12, thiamine, BENDANA, F. E., AND M. FRIED. 1967. Stimulatory and biotin in Lake Tahoe waters using modified marine effects of gibberellins on the growth of Chlore/la bioassay techniques. Limnol. Oceanogr. 17: 774- pyrenoidosa. Life Sci. 6: 1023-1033. 777. BENOIT, R. J. 1957. Preliminary observations on cobalt CARLUCCI, A. F., AND R. L. CUHEL. 1977. Vitamins and vitamin B12 in fresh water. Limnol. Oceanogr. in the South Polar Seas: distribution and significance 2: 233-240. of dissolved and particulate vitamin B12, thiamine, BENTLEY, J. A. 1958. Role of plant hormones in algal and biotin in the Southem Indian Ocean, p. 115-128. metabolism and ecology. Nature (London) 181: 1499 In G. A. Llano [cd.] Adaptations within antarctic -1502. ecosystems. Third SCAR Symposium on antarctic 1959. Auxin like substances in algae. Annu. Rep. biology. Smithsonian Institution, Washington, DC. Challenger Soc. 3: 20. CARLUCCI, A. F., AND S. B. SILBERNAGEL. 1966. 1960. Plant hormones in marine phytoplankton, Bioassay of seawater. III. Distribution of vitamin zooplankton and sea water. J. Mar. Biol. Assoc. U.K. B 12 in the northeast Pacific Ocean. Limnol. Oceanogr. 39: 433-444. 11: 642-646. BENTLEY - MOWAT, J. A. 1967. Do plant growth-sub- CATTELL, S. A. 1973. The seasonal cycle of vitamin B 12 stances affect the development and ecology of uni- in the Strait of Georgia, British Columbia. J. Fish. cellular algae? Wiss. Z. Univ. Rostock Math. Natur- Res. Board Can. 30: 215-222. wiss. Reihe 16: 445-449. COWEY, C. B. 1956. A preliminary investigation of the BENTLEY - MOWAT, J. A., AND S. M. REID. 1967. Effect variation of vitamin B 12 in oceanic and coastal waters. of gibberellins, kinetin and other factors on the growth J. Mar. Biol. Assoc. U.K. 35: 609-620. of unicellular marine algae in culture. Bot. Mar. 12: DAISLEY, K. W. 1969. Monthly survey of vitamin B12 185-199. concentrations in some waters of the English Lake BERLAND, B. R., D. J. BONIN, J. P. DURBEC, AND S. Y. District. Limnol. Oceanogr. 14: 224-228. MAESTRINI. 1976. Bactéries hétérotrophes aérobies 1970. The occurrence and nature of Euglena prélevées devant le delta du Rhône. IV. Besoins en gracilis proteins that bind vitamin B 12 . Int. J. Biochem. vitamines et libération de ces substances. Hydro- 1: 561-574. biologia 50: 167-172. DAISLEY, K. W., AND L. R. FISHER. 1958. Vertical BERLAND, B. R., D. I. BONIN, M. FIALA, AND S. Y. distribution of vitamin B12 in the sea. J. Mar. Biol. MAESTRINI. 1978. Importance des vitamines en mer. Assoc. U.K. 37: 683-686. Consommation et production par les algues et les bac- DROOP, M. R. 1954. Cobalamin requirement in Chryso- téries, p. 121-146. In Les substances organiques phyceae. Nature (London) 174: 520. naturelles dissoutes dans l'eau de mer. Colloque 1958. Requirement for thiamine among some GABIM-CNRS, 1976. CNRS, Paris. marine and supralittoral protista. J. Mar. Biol. Assoc. BONIN, D. 1969. Influence de différents facteurs écolo- U.K. 37: 323-329. gigues sur la croissance de la diatomée marine Chaeto- 1961. Vitamin B12 and marine ecology: the ceros affinis Lauder en culture. Tethys I: 173-238. response of Monocluysis lutheri. J. Mar. Biol. Assoc. BRADBEER, C. 1971. Transport of vitamin B 12 in Ochro- U.K. 41: 69-76. monas malhamensis. Arch. Biochem. Biophys. 144: 1962. Organic micronutrients, p. 141-159. In 184-192. R. A. Lewin [ed.] Physiology and biochemistry of BRANNON, M. A., AND H. M. SELL. 1945. The effect algae. Academic Press, New York, NY. of IAA on dry weight of Chlorella pyrenoidosa. Am. 1968. Vitamin B12 and marine ecology. IV. The J. Bot. 32: 257-258. kinetics of uptake, growth and inhibition in Mono- BRIAN, P. W. 1963. Growth regulating hormones in lower chtysis lutheri. J. Mar. Biol. Assoc. U.K. 48: 689- plants. Sci. Prog. 51: 226-238. 733. BRUNO, S. F., AND R. D. STAKER. 1978. Seasonal DROOP, M. R. , J. J. A. MCLAUGHLIN, I. J. PINTNER, vitamin B12 and phytoplankton distribution near AND L. PROVASOLI. 1959. Specificity of some Napeague Bay, New York (Block Island Sound). protophytes toward some vitamin B 12 -like compounds, Limnol. Oceanogr. 23: 1045-1051. p. 916-918. In M. Sears [cd.] International oceano- BURKHOLDER, P. R. 1959. Vitamin producing bacteria graphic congress, preprints. Am. Assoc. Adv. Sci. in the sea, p. 912-913. In M. Sears [cd.] International EPSTEIN, S. S., AND G. M. 'Moms. 1963. Simple oceanographic congress, preprints. Am. Assoc. Adv. antimetabolites of vitamin B 12 . J. Protozool. 10: 63- SCi. 73. CARLUCCI, A. F. 1970. The ecology of the plankton off ERICSON, L. E., AND L. LEWIS. 1953. On the occurrence La Jolla, California, in the period April through of vitamin B 12 -factors in marine algae. Ark. Kemi 6: September, 1967. II. Vitamin B12, thiamine and 427-442. biotin. Bull. Scripps Inst. Oceanogr. 17: 23-31. FIALA, M. 1980. Vitamine B12 et phytoplancton dans

320 diverses régions marines. Thèse Doct. ès Sci. Univ. primary production and phytoplankton population Pierre et Marie Curie, Paris. 128 p. in lake Mergozzo (Northern Italy). Mem. Ist. Ital. FOGG, G. E. 1966. The extracellular products of algae. Idrobiol. 33: 257-284. Oceanogr. Mar. Biol. Annu. Rev. 4: 195-212. LEWIN, R. A. 1954. A marine Stichococcus sp. which FORD, J. E. 1958.13 12 -vitamins and growth of the flagel- requires vitamin B12 (cobalamin). J. Gen. Microbiol. late Ochromonas malhamensis. J. Gen. Microbiol. 10: 93-96. 19: 161-172. 1961. Phytoflagellates and algae, p. 401-417. hi GLASS, G. B. J. 1963. Gastric intrinsic factor and its W. Ruhland [ed.] Handbuch der Pflanzenphysiologie, function in the metabolism of vitamin B12. Physiol. Vol. 14. Springer Verlag, Berlin. Rev. 43: 529-849. LWOFF, A., AND H. Dusi . 1937. La pyrimidine et le GRAY, N. J. P. 1962. The effects of gibberellins on Chlo- thiazol, facteurs de croissance pour le flagellé Poly- rella. Phycol. Soc. Am. News Bull. 15: 34. tomella caeca. C. R. Acad. Sci. Paris 205: 630. GUILLARD, R. R. L. 1968. B 1 2 specificity of marine LWOFF, A., AND E. LEDERER. 1935. Remarques sur centric diatoms. J. Phycol. 4: 59-64. rextrait de terre" envisagé comme facteur de crois- GUILLARD, R. R. L., AND V. CASSIE. 1963. Minimum sance pour les flagellés. C.R. Soc. Biol. 119: 971- cyanocobalamin requirements of some marine centric 973. diatoms. Limnol. Oceanogr. 8: 161-165. MANOS, E. 1945. The effect of heteroauxin on the growth HAINES, K. C., AND R. R. L. GUILLARD. 1974. Growth of Chlorella. Biol. Bull. 7: 11. of vitamin B 12 -requiring marine diatoms in mixed MENZEL, D. W., AND J. P. SPAETH. 1962. Occurrence laboratory cultures with vitamin 13 12 -producing marine of vitamin B 12 in the Sargasso Sea. Limnol. Oceanogr. bacteria. J. Phycol. 10: 245-252. 7: 151-154. HERDMAN, M., AND N. G. CARR. 1972. The isolation MOWAT, J. A. 1963. Gibberellin-like substances in algae. and characterization of mutant strains of the blue-green Nature (London) 200: 453-455. alga, Anacystis nidulans. J. Gen. Microbiol. 70: NAKAMURA, K., AND C. S. GOWANS. 1964. Nicotinic 213-220. acid-excreting mutants of Chlamydomonas. Nature

HOLM -HANSEN, O., S. Z. EL - SAYED, G. A. FRANCES- (London) 202: 826-827. CHINI , AND R. L. CUHEL. 1977. Primary production NATARAJAN, K. V. 1970. Distribution and significance and the factors controlling phytoplankton growth in of vitamin B12 and thiamine in the subarctic Pacific the southern ocean, p. 11-50. In G. A. Llano [ed.] Ocean. Limnol. Oceanogr. 15: 655-659. Adaptations within antarctic ecosystems. Third SCAR NATARAJAN, K. V., AND R. C. DUGDALE. 1966. Bio- Symposium on antarctic biology. Academic Press, assay and distribution of thiamine in the sea. Limnol. New York, NY. Oceanogr. 11: 621-629. HUTNER, S. H., H. BAKER, S. AARONSON, H. A. OHWADA, K., AND N. TAGA. 1972a. Vitamin B12, NATHAN, E. RODRIGUEZ, S. LOCKWOOD, M. thiamine and biotin in Lake Sagami. Limnol. Ocea- SANDERS, AND R. A. PETERSON. 1957. Growing nogr. 27: 315-320. Ochromonas malhamensis above 35°C. J. Protozool. 1972b. Distribution and seasonal variation of 4: 259-269. vitamin B12 thiamine and biotin in the sea. Mar. HUTNER, S. H., L. PROVASOLI, AND J. FILFUS. 1953. Chem. 1: 61-73. Nutrition of some phagotrophic freshwater chryso- PARKER, M. 1977. Vitamin B 12 in Lake Washington, USA: monads. Ann. N.Y. Acad. Sci. 56: 852-862. concentration and rate of uptake. Limnol. Oceanogr. JACOBSEN, D. W., P. M. DIGIROLAMO, AND F. M. 22: 527-538. HUENNEKENS. 1975. analogues PEDERSEN, M. 1973. Identification of a cytokinin, 6- as inhibitors of ribonucleotide reductase and vitamin (3-methy1-2-butenylamino) purine, in sea water and B 12 transport. Mol. Pharmacol. 11: 174-184. the effect of cytokinins on brown algae. Physiol. Plant. JOHNSTON, R. 1959. Antimetabolites and marine algae, 28: 101-105. p. 918-920. In M. Sears [ed.] International oceano- PEDERSEN , M., AND G. FRIDBORG. 1972. Cytokinin- graphic congress, preprints. Am. Assoc. Adv. Sci. like activity in sea water from the Fucus-Ascophyllum 1963. Effects of gibberellins on marine algae in zone. Experientia 28: 111-112. mixed cultures. Limnol. Oceanogr. 8: 270-275. PINTNER, I. J., AND V. L. ALTMEYER. 1979. Vitamin KASHIWADA, K., AND D. KAKimoTo. 1962. Studies 13 1 2 -binder and other algal inhibitors. J. Phycol. 15: on vitamin B 12 in natural water. VII. Relation between 391-398. the fertility and the vitamin B 12 content of lakes. Bull. PROVASOL1, L. 1963. Organic regulation of phytoplankton Jpn. Soc. Sci. Fish. 28: 352-360. fertility, p. 165-219. hi M. H. Hill [ed.] The sea, Kim, W. K., AND V. A. GREULACH. 1961. Promotion Vol. 2. Interscience Publications, New York, NY. of algal growth by IAA, GA and kinetin. Plant. 1971. Nutritional relationships in marine or- Physiol. 36 (suppl.): 14. ganisms, p. 369-382. In J. D. Costlow [ed.] Fertility KOLHOUSE, F. J., AND R. H. ALLEN. 1978. Isolation of the sea, Vol. 2. Gordon and Breach, New York, of cobalamin and cobalamin analogues by reverse NY. affinity chromatography. Anal. Biochem. 84: 486- PROVASOLI , L., AND A. F. CARLUCCI. 1974. Vitamins 490. and growth regulators, p. 741-787. In W. D. P. KURATA, A., C. SARACENI, D. RUGGIU, M. NAKANISHI, Stewart [ed.] Algal physiology and biochemistry. U. MELCHIORRI-SANTOLINI, AND H. KADoTA. Blackwell Scientific Publications, Oxford, U.K. 1976. Relationship between B group vitamins and RAHAT, M., AND K. REICH. 1963. The B 12 vitamins and

321 growth of the flagellate Prymnesium parvum. J. 1978. Unexpected response to vitamin B19 of Gen. Microbiol. 31: 195-202. dominant centric diatoms from the spring bloom in REEVES, B. R., AND S. F. FAY. 1966. Cyanocobalamin the Gulf of Maine (Northeast Atlantic Ocean). J. (vitamin 13 19 ) uptake by Ochromonas malhamensis. Phycol. 14: 377-386. ' Am. J. Physiol. 210: 1273-1278. SWIFT, D. G., AND W. R. TAYLOR. 1974. Growth of SARI1AN, F., M. HOUDE, AND J. P. CHENEVAL. 1980. vitamin B 12 -limited cultures: Thalassiosira pseudo- The role of vitamin B 19 binding in the uptake of the nana, Monocluysis lutheri, and Isocluysis galbana. vitamin by Euglena gracilis. J. Protozool. 27: 235- J. Phycol. 10: 385-391. 238. TAUTS, M. I., AND V. E. SEMENENKO. 1971. Isolation SHARMA, G. M., H. R. DuBois, A. T. PASTORE, AND and identification of physiologically active agents of S. F. BRUNO. 1979. Comparison of the determination indol nature in extracellular Chlore/la metabolites. Dokl. Akad. Nauk. SSSR. 198: 970-973. (In Russian) of cobalamins in ocean waters by radioisotope dilution THIMANN, K. V. 1934. Studies on the growth hormone and bioassay techniques. Anal. Chem. 51: 196-199. of plants. IV. The distribution of the growth substances SCHIEWER, U. 1967. Auxinvorkommen und Auxin- in plants. J. Gen. Physiol. 18: 23-34. stoffwechsel bei mehrzelligen Ostseealgen. I. Zum VAN BAALEN, C. 1961. Vitamin B19 requirement of a Vorkommen von Indo1-3-Essigsaiire. Planta 74: marine blue-green alga. Science (Washington, D.C.) 313-323. 133: 1922-1923. STARR, T. J., M. E. JONES, AND D. MARTINEZ. 1957. VAN OVERBEEK, F. 1940a. Auxin in marine algae. Plant The production of vitamin 13 19 -active substances by Physiol. 15: 291-299. marine bacteria. Limnol. Oceanogr. 114-119. 1940b. Auxin in marine plants. II. Bot. Gaz. 101: Sw tFT, D. G. 1981: Vitamins and phytoplankton growth, 940-947. p. 329-368. /n I. Morris [ed.] Primary productivity VARMA, T. N. S., A. ABRAHAM, AND I. A. HANSEN. of natural waters. Blackwell Scientific Publications, 1961. Accumulation of Co"-vitamin 13 19 by Euglena Oxford, UK. gracilis. J. Protozool. 8: 212-216. SWIFT, D. G., AND R. R. L. GUILLARD. 1977. Diatoms VISHNIAC, H. S., AND G. A. RILEY. 1961. Cobalamin as tools for assay of total B 19 activity and cyano- and thiamine in Long Island Sound: patterns of cobalamin activity in sea water. J. Mar. Res. 35: distribution and ecological significance. Limnol. 309-320. Oceanogr. 6: 36-41.

322 Allelopathic Relationships Between Phytoplankton Species

SERGE Y. MAESTRINI AND DANIEL J. BON IN Station Marine d'Endoume Chemin de la Batterie des Lions F-13007 Marseille

Introduction not be attempted in this paper, as bacteria are beyond the scope of the review. . The first mention of the importance of dissolved Substances released by algae that are toxic to organic matter for aquatic life (i.e. nutrition) goes other algae have been studied for a long time, but back to Thomson (1874); the analysis and experi- not all contributions have been critical. Thus, only ment phase began with Pütter (1907a, b). However, few experimental data are really pertinent and the Akehurst (1931) was probably the first to suggest question of the importance of these substances is the existence between phytoplankton species of non- still open. Reproducing an exhaustive list of published nutritional relationships mediated by organic sub- articles would be useless. The reader can find all stances. From his findings on fluctuations of algal references needed in the reviews of Lefevre (1964) populations in inland ponds, he supposed that most and Tassigny and Lefevre (1971), who were specially algae secrete extracellular organic compounds into dedicated to antialgal—algal ectocrines; the reviews the surrounding water that can inhibit the growth of Lucas (1947, 1955, 1961) and Pourriot (1966), of some algae and favor the growth of others. Lucas who discussed the biochemical interactions between (1938) extended these ideas to marine life and, later aquatic organisms as a whole; and Saunders (1957), (Lucas 1947) developed a broad concept and named Hartman (1960), Fogg (1962), Provasoli (1963), substances produced and released by various living or- Whittaker and Feeny (1971), Hellebust (1974), and ganisms , which can affect other individuals or species Rice and Ferguson (1975), who included this topic at a distance, "external metabolites" or "ectocrines." in their discussions of algal physiology and ecology. These external metabolites can be growth-promoting Only a brief summary of the conceptual progress or growth-inhibiting compounds or micronutrients , before these reviews will be given here. Subsequent and are specially important for microorganisms. papers will be dealt with in more detail. Lucas believed that during the course of evolution many organisms have adapted themselves to tolerate or take advantage of these substances released by ANTIALGAL ECTOCRINES RELEASED their neighbors , and organisms that failed to do so BY ALGAE: HISTORICAL BACKGROUND must have become extinct or have evolved some It is obvious that marine biologists are indebted avoiding mechanisms. This concept is vitally con- to inland aquatic microbiologists for making the cerned with community integration, competition, basic discoveries (see summary in Table 1). and succession of algal species. Among all problems pointed out by Lucas Step one — Pütter's (1907a, b) pioneering ideas (1947), it was contention about the importance of opened a new field in biology (Ranson 1935, 1936; the antibacterial substances released by algae that Johnston 1972), but he obviously overemphasized drew the attention of the scientific community and by one order of magnitude the quantitative importance stimulated most research. The numerous papers on of both algal secretion and seawater content of this field have been frequently reviewed (Lucas 1947, dissolved organic substances. With Krogh (1931), 1955, 1961; Nigrelli 1958; Sieburth 1964, 1968; who was a critical chemist, the dissolved organic Aubert 1971, 1978; Berland et al. 1974) and some- matter content of seawater became accurately known. times overreviewed. On the other hand, it has been From the data tabled in the review of Williams (1975) demonstrated that substances released by bacteria and Whittle (1977) it appears that the approximate can inhibit bacterial growth (Rosenfeld and Zobell mean concentration of dissolved organic carbon 1947; Krasil'Nikova 1964; Burkholder et al. 1966) (DOC) in surface seawaters ranges from 1 mg C •1_," or algal growth (Fitzgerald 1969; Berland et al. to 1.5 mg C • L'. These values will be useful for 1972a, b), but little attention has been paid to growth- later discussion. Recent developments of the anal- limiting bacterial ectocrines, and their importance ytical technology by Lindroth and Mopper (1979) has rarely been discussed. Such a discussion will made available a new method, i.e. the high perfor-

323 TABLE 1. Allelopathic interactions between phytoplankton species: progress of concepts from early contributions to last comprehensive reviews (Hellebust 1974; Rice and Ferguson 1975).

Step 1 Step 2 Step 3 Step 4 Step 5

Harder (1917) records Akehurst (1931) Pratt (1940-43) Rice (1954) does

h accumulation of auto- suggests existence extracts and purifies first fully quantitative inhibiting substances of nontrophic chlorellin, an inimical work; first use of in culture medium relationships ectocrine produced pH buffer and CO2 researc of cyanophyta Nostoc between phyto- by Chlorella vulgaris; supply; uses constant ter punctiforme; in that plankters sustained first use of dialysis mixed cultures in

hwa he followed by chemical sack for alga culturing great extent Woodruff's (1913) substances Fres experiments with the ciliata Paramecium Proctor (1957) extracts an inhibiting sub- stance from the culture filtrate of a green alga

Thomson (1874) Lucas (1947) reviews Pincemin (1971) first suggests importance possible effects of mixed culture which cqti of dissolved organic external metabolites involves a dino- matter for marine and proposes a flagellate life. Pütter (1907a, b) broad concept that . g does first experi- includes physio- ments and emphasizes logical adaptation algal secretion and species succes- sion; proposes the term "ectocrine"

mance liquid chromatography with fluorescence mixed cultures are done with a view to separating derivatization that decreases by 10 times the duration the effects of nutrient competition and "chemical of analysis, 100 times the volume of sample required, war" between the test species (Fig. I). (4) In sub- and increases the sensitivity by 1000 times. culturing experiments, pH varies; no buffer is used; no attention is paid to the decreased CO., content Step 2 — Harder (1917) recorded accumulation, of the subculture. (5) Some experiments are made in the culture medium of the cyanophyte Nostoc by extracting the inimical substances from packed punctiforme , of organic substances that are auto- cells or dried cells; no attempt has been made to inhibiting for the organism that produced them. In extract the active substance from the filtrate. (6) this he emulated Woodruff s (1913) work and results Inhibiting effects are often observed only on mor- with the ciliate Paramecium . Hence, it appears that phological basis. (7) Test algae are cultured in nu- the dissolved organic matter (DOM) can be poisonous trient-rich medium; thus, algal biomass may produce and may have a role other than trophic. Since then, external metabolites at very high levels. Because their similar observations have been made either with ma- activity is growth-promoting at low concentrations rine or freshwater species, but no critical progress was and growth-inhibiting at high concentration, in vitro made as most data published were obtained in such a data are irrelevant to statements on their role at way they are vulnerable to several of the following ecological level. (8) Active algae might have been criticisms: (1) Algal cultures are not bacteria-free; isolated from different locations and may never that is another potential source of production or belong to the same natural community. destruction of growth-inhibiting algal substances. (2) Filtrates are prepared with ordinary paper filters Step 3 — In the 1930s, the aquatic scientific that might contain particles and bacteria. (3) Most community was already preoccupied with changes experiments are made by filtering an algal culture in nutrition resulting from the accumulation of ex- and subculturing another algae in the filtrate; no cretory products in the culture medium of a growing

324 A in fresh B in fresh medium medium / S S BIOMA

ti I V B in filtrate of A e• A in of B

TIME

FIG. 1. Scheme of classic method used to demonstrate allelopathic interactions between two algal species. The following operations are usually done with unialgal or axenic strains: (i) species A and B are separately grown in batch cultures; (ii) species A is subcultured in nutrient-reinriched, cell-free culture medium of species B, species B is similarly subcultured in filtrate medium of species A; (iii) biomass increases are recorded and plotted versus time duration; (iv) growths of species A and B in preconditioned media are compared with respective controls; (v) conclusions are: species B inhibits species A, species A does not inhibit species B. population. However, this "group effect" was hitherto that the autotoxins of the oil group become an acces- studied mainly in vitro with various organisms, sory food for the starch group and vice versa, and mostly animals (see Allee 1934). But at that time he thought this process explains why the growth of the scientific community represented "a few small almost the entire oil group ceases at the same time groups of near-amateurs with enthusiasm and interest and why no recovery is shown until the starch period in a wide range of disciplines" (French 1979), prior has passed through its phase of abundance. to becoming a "very large number of groups all Later work refined the critical observations of intent on a small segment of science and seemingly the 1930s, either in freshwater blooms or marine with little interest in followers of other subjects." species succession (see Lefevre et al. 1952; Pratt Thus, Akehurst (1931) certainly was aware of studies 1965, 1966), but no new concept was proposed. At done in other fields when he tried to extend to field most, autoantagonistic substances were demonstrated conditions some principles elucidated by laboratory to be also inimical for other algae. Thus, in the research. In an inland pond, he observed a roughly ecological perspective, the broad concept of Lucas rhythmical rise and decline in numbers of different (1947) appears to be an extension of Akehurst's species of phytoplankton that he related partly to statement, but Lucas expressed clearly the theory the physical and nutrient conditions, and partly to of the nonpredatory relationships linking marine other factors that he called "toxines." These sub- organisms and showed its peculiar importance for stances were defined by Akehurst as "an excretory marine microorganisms. His discussion was docu- product or products of indefined chemical consti- mented by results which showed that algae excrete tution which may also serve as an accessory food various nutrients, that vitamins are present in waters and may inhibit or stimulate growth." The phyto- and produced by bacteria and algae, and by the iso- plankton was considered to consist of two main lation of a purified algal antibiotic, the "chlorellin," groups, when classified according to cell reserves: by Pratt and Fong (1940, 1944). The rapid develop- the oil group and the starch group. Akehurst wrote ment of antibiotic use, and the recognition of their

325 production by various microorganisms, certainly of nutrient supply, pH buffer, and air bubbling to helped Lucas (1947) to extend to aquatic life some eliminate lack of COo . He found that biological principles well established for land or soil fungi production of poisonous ectocrines is responsible and bacteria (Waksman 1941). for antagonistic growth inhibition whether or not there is nutrient competition, but densities obtained By measuring the variations of yield Step 4 — in experimental pure cultures were much higher concentration of the medium, as a function of nutrient than those ordinarily found in nature. As a matter Fong (1940) demonstrated Pratt (1940) and Pratt and of fact, the greatest inhibition in the growth of N. that the green alga Chiorella vitlgaris produces and frustithitti occurred when 5 x 10" C. vulgaris cells liberates a substance that tends to retard its own per litre were present; only 70% inhibition occurred growth, adding a new species to the list of algae diatom then. Application producing autoinhibiting substances. But later, in the division rate of the to natural populations was uncertain, indeed. The Pratt (1942) cultured, the alga in a collodion dialysis same remark can be made about the work of McVeigh membrane and demonstrated the production of the and Brown (1954) who, nevertheless, introduced a inhibiting substance by healthy growing cells. Then, new refinement. They separated the two algae they he extracted the substance from dried cells using used by enclosing the cells in two dialysis sacks, organic solvents and eliminated the chlorophyll and but allowed the external metabolites to act by im- much of the extraneous organic matter. Thus, he mersion of the two coupled sacks in the same medium could attempt a tentative chemical determination of flask. Unfortunately, their strains were not bacteria- the active substance he named "chlorellin," and study free and contained a strain whose ectocrines inhibited its properties. Logically, he continued by using the or enhanced the growth of algae tested. They also purified substance to run an experiment that allowed did not pay any attention to the pH and CO2 effects versus the number of cells him to express the results in crossed subculturing experiments and, thus, 17 yr to reach the needed to produce enough substance passed before results based on a satisfactory technical threshold concentration for inhibiting effect (Pratt approach were published. 1943). From his data, it appeared that 1.8 x 10' cells • mL-1 are needed for a visible inhibition. He also demonstrated that chlorellin is active against CONTRIBUTIONS PUBLISHED AND PROGRESS bacteria (Pratt and Fong other algae (Pratt 1943) and MADE DURING THE PAST DECADE 1944). Thus, Pratt and co-workers took a critical step by producing the first reliable data for the significance The main feature of recent research seems to be of in vitro results in natural conditions; but, curiously, that research programs have become more compre- Pratt did not make this extension. hensive, involving observations of natural processes The production of hetero-initnical ectocrines and experiments in the laboratory on strains of the being well established, and Proctor (1957) having same origin. Further, the different cell functions such extracted purified substances from cultures filtrates, as nutrition, excretion, mobility, or production of one wonders that few have followed Pratt's example. inimical ectocrine are now not separated. It also It is also surprising that little attention has been paid appears that production of inimical ectocrines is no so far to the discrepancy between the density of cells longer considered a permanent avantage, but a po- involved in laboratory experiments versus those in tential cell weapon, acting in defined situations. natural populations with regard to the real ecological Apparently this marks a welcome shift from the role of the antimetabolites. Apparently the aquatic concept of "chemical warfare" among microor- biologists preferred to avoid time-consuming bio- ganisms of the 1960s (Sieburth 1962). chemical methods in favor of experiments on mixed Most researchers at present pay attention to the cultures. fact that all physiological mechanisms can interact and that "weapons" used in species competition Step 5 — Many aquatic biologists have experi- may vary greatly when environmental conditions mented with simple systems of two or a few species change. Understanding when and why they are acting of microorganisms since Gause (1932) first tried or useful seems to be the aim of the foremost works. to set up a theoretical model from experimental data There is also an impressive development of research obtained with mixed cultures of two species of yeast. done with dinoflagellates (which have been too long However, most experiments with two or more algae ignored); improvement of culture media for growing were made on a morphological basis (see Lefevre these peculiar algae certainly stimulated the growing 1964) and the first fully quantitative study appears interest in them. Since the foremost research involves to be that of Rice (1954). Rice cultured bacteria- different methods of study at one time, or in succes- free strains of the green alga Chlorella vulgaris and sive steps, papers relevant to a particular study may the diatom Nitzschia frustul uni in mixed liquid or appear over an extended period. Thus, chronological agar plate batch cultures, with various combinations priority is sometimes misleading.

326 TABLE 2. Allelopathic interactions between phytoplankton species. Critical progress made during the past decade (for steps 1 to 5, see Table 1).

Step 6 Step 7 Step 8

Kroes (1971, 1972) avoids Bialgal cultures of two dino- Fully integrated field observations and all previous technical short- flagellates demonstrate that experiments done with locally isolated strains comings, devises a filter respective behaviors greatly bialgal culture method, and depend on relative cell pro- Keating (1977, 1978) made observations over extends the inimical ectocrines portions (Elbrâchter 1976) 5 yr and indicated that diatom blooms vary

from cell-free filtrates h inversely with density of preceding cyanophyta Gauthier et al. (1975) c populations; crude lake water contains heat-labile Ranges of concentrations for integrate field observations ear substances which inhibit diatoms; dominant es

some allelopathic effects are and laboratory experiments r species produce only inhibithig or neutral

determined: 10-50 mg • L-1 ke effects on immediate predecessors, and only

I, (Jüttner- 1976-79); 25 mg • Similar approach demonstrates La beneficial or neutral effects on immediate (Mc Cracken et al. 1980) a naturally occurring cyano- successors; strains of same species but from phyte may govern growth of a other locations are less susceptible; bacteria Substances excreted by an toxigenic dinoflagellate act by destroying active substances alga might have different (Martin and co-workers effects with different phases 1974-79) From Pratt (1965) to Tomas (1978-80). of growth (Federov and Blooms of Skeletonetna costatum and Kustenko 1972) Experiments made with Olithodiscus luteus alternate; subculture in natural waters where actual 108 cells • I.» filtrate of O. luteus suppresses The cyanophyta Anaboena dominant species are blooming growth of S. costatum, whereas low densities inhibits growth of its com- and preceding dominant 11.2, stimulate; O. luteus ability to take up nutrients panion species by producing species are becoming scarce is regular, notwithstanding it blooms in a strong iron chelator found (Honjo et al. 1978) nutrient-depleted waters; S. costatum grows « g both in culture and natural well in dialysis sacks incubated when O. luteus lake waters (Murphy et al. blooms; O. luteus escape grazing possibly by 1976) production of an ectocrine repellent

ALLELOPATHIC INTERACTION BY with an exhaustion of selected nutrient elements. CROSSED SUBCULTURING Crossed subculturing of the pool of three species in culture filtrates with or without associated bacteria Harris (1970, 1971) surveyed the auto- and demonstrated that no toxic or allelopathic matter hetero-inhibition of 11 freshwater algae by crossed acts between the cells. However, it should be noted subculturing cell-free culture filtrates (however, not the three strains used were obtained from different reenriched), and found that all species produced culture collections and thus were isolated from dif- inhibiting substances, specially Pandorina morum , ferent origins, so the data are not entirely free from Volvulina pringsheimii , Eudorina cylindrica , and criticism. Schenk and Jüttner (1974), Jiittner (1976), E. illinoisensis . By comparing the growths of the Herrmann and Jüttner (1977), and Jiittner (1979) cyanophyte Microcystis aeruginosa and the green studied the excretion products of freshwater algae algae Monoraphidium minutum and Scenedesmus both from pelleted cells of naturally occurring bloom abundans in monocultures and bicultures, Krzywicka cyanophyte and from cultured cells. By using labo- and Krupa (1975) found that the cyanophyte repro- ratory bioassays, they demonstrated that norca- duces more slowly than the green algae in bicultures. rotenoid excreted by Cyanidiutn caldarium inhibits They suspected that the green algae excrete into the the growths of Anabaena variabilis , Nannochloris environment some toxic substances that inhibit coccoides , and synechococcus sp. within a range development, but they did not try either to isolate of 10-50 mg • . Although these contributions the active substance or to question if nutrient compe- provide some rare quantitative data on inhibiting tition might be responsible for the reduced growth. substances, they help but little the present discussion On the contrary, Lange (1974) showed that in a mixed because their aim is different and their experimental experimental cyanophyte system of Microcystis protocols are questionable. The data of Chan et al. aeruginosa, Nostoc muscorum , and Phortnidium (1980) are equally questionable. These authors foveolarum one species may exclude the other two reported the inhibition of the diatom Cylindrotheca

327 fusifortnis by a dozen other algae belonging to attempts to extend the conclusions to natural condi- several taxa, including diatoms and dinoflagellates, tions. Because, in addition, they do not say how many but they used the plating bioassay technique with cells are required to produce 25 mg of active extract, methanol extracts of packed algal cells or algal- their data contribute little to our discussion. Lam culture filtrates. Hence, they did not get any evidence and Silvester (1979) studied the growth interactions that the zones of inhibition observed around the among the cyanophyte Anabaena oscillarioides and impregnated disks really reflect an in situ occurting Microcystis aeruginosa and the green alga Chlorella process. Chan et al. did not claim such implications sp., by using both standard mixed cultures and cul- and did not ignore the shortcomings of their experi- tures separated by a membrane filter. They demon- ments. They just wanted to survey many species strated that A. oscillarioides and M. aeruginosa in order to get a convenient material for further both inhibit the growth of the Chlorella strain. experiments. However, for such an aim, in our Microcystis aeruginosa probably acts by releasing opinion, the method used by Berland et al. (1973) an inhibiting substance, while the inhibiting effect for surveying alga—bacterium allelopathic inter- of A. oscillarioides on Chlorella sp. is mostly the actions could be a better approach, because it allows consequence of a competition for uptake of phos- both the antagonistic species on the same agar medium phorus. This mechanism can explain why in nature to grow, and this procedure avoids many artefacts M. aeruginosa is one of the most common bloom- and saves time (Fig. 2). forming species, even if its nutritional characteristics Sze and Kingsbury (1974) studied fresh waters leave it vulnerable to domination by other species. heavily polluted by copper but with enough nutrients The demonstration of Lam and Silvester is not, to support the resident phytoplankton growth without however, totally convincing, because they used limitation. This pollution has not resulted in a phyto- isolates from different origins and grew the algae plankton flora of only a few well-adapted species in a very rich nutrient medium, which gave artificially as might be expected; a large number of species high cell densities (i.e. up to 16 x 10" cells ). and a regular succession are still found. Among them, a Chlatnydontonas species appears every year IMPROVED TECHNOLOGIES LEAD TO in several brief but dense peaks; in contrast, another MORE CONVINCING DEMONSTRATIONS alga, Staurastruin paradoxion, is present perma- nently but in low abundance. Sze and Kingsburry Kroes (1971, 1972a, b) developed a filter cul- hypothesized that an allelopathic interaction con- ture method in which two separate cultures of different trolled the observed fluctuations in populations. species are connected via a filtering system, through They did several experiments with mixed cultures which medium is exchanged while the cells them- and crossed subcultures in culture filtrates and found selves are kept separate. The culture medium was that neither the presence of an inhibitor nor nutrient Tris-buffered and air-bubbled. Algal strains were competition could explain the reduced growth of axenic. With such equipment, Kroes demonstrated the planktonic alga, S. paradoxunt, when cultured that Chlorococcum ellipsoideunt inhibits the growth with the Chlatnydomonas strain. They concluded that of Chlantydontonas globosa , but not vice versa. The the interactions reported are similar to the dynamics most important finding was that there are many dif- observed in natural waters and may describe some ferent extracellular compounds released by C. ellip- population fluctuations, even though the experimen- soideum and that each fraction isolated from the tal conditions differed significantly from the natural culture filtrate had its own specific, but small, effect conditions. This is by no means clear, but never- on the growth of other alga. He also pointed out theless, it is almost certain that several Chlatnydo- that the role of pH is more important in interactions monas species are able to produce and release one than extracellular substances. In that, Kroes' state- or several substances inimical to other algae. ment differs from those of previous authors who McCracken et al. (1980) studied the same strain of assumed that in the algal ecosystem strongly inhib- C. reinhardtii used by Proctor (1957), who demon- itory substances were responsible for the observed strated that a substance produced by this alga is toxic phenomena, and paid little, if any, attention to the to another species. The latter is inhibited when grown pH (except Proctor 1957). Then, in a further technical in cell-free filtrate that McCracken et al. demonstrated improvement, Kroes (1973) designed a new apparatus to be a mixture of three fatty acids active against that, for practical purposes, excludes pH as an element six of seven test algae, including two strains (Hama- of the interspecies interaction, and thus let the algae tococcus lacustris and Oocystis sp.) which are killed compete only by competitive nutrient uptake and by by 25 mg'L of crude extract. Unfortunately, production of inimical extocrine. Scott and Ball McCracken et al. did not completely follow Proctor; (1975) built improved equipment derived from they extracted the active substance from the total Kroes' remarks, but no experiment has been reported culture (i.e. including the cells). This does not help yet.

328 Inhibition No inhibition

Possible Growth enhancement inhibition

Possible Divergent strokes with inhibition gradient inhibition (young cultures)

FIG. 2. Scheme of method used by Berland et al. (1973) for screening allelopathic interactions between a gliding alga and a nonmotile microorganism. Experiments are done in Petri dishes and agar medium. The gliding alga is inoculated with a wire loop; a stroke (solid bar) is made at the center of the plate. A stroke (hatched bar) of the potentially antagonistic strain parallels the first. When the nonmotile micro- organism releases an extremely inhibitory substance, the alga moves (black arrows) exclusively in the opposite direction (inhibition); the antagonistic strain stroke may appear like a fence. When inhibition is not so strong, the gliding alga may touch the other strain (possible inhibition). If no inhibiting substance is involved, the gliding alga grows as usual and covers the whole agar medium (no inhibition). Growth-promoting substances may be released by the nonmotile cells; this leads the gliding companion species to grow towards it faster (growth enhancement).

329 Fedorov and Kustenko (1972) studied the notice that allelopathy (inimical effect) may play an competition between the diatoms Thalassiosira important role in large enriched, outdoor cultures. nitzschioides and Skeletonema costatum for uptake The dominant production of the undesired P. tricor- of nitrogen and demonstrated that neither of the minim on Woods Hole mass cultures (Ryther et al. two algae can gain the upper hand over the other 1975) may support this opinion. The similar results in mixed culture with the same initial cell density. of Pintner and Altmeyer (1979) pertain to a similar However, the shape of the growth curves points to technological approach, but their study, which in- the existence of a reciprocal inhibition that was volved 21 algal species belonging to different taxa, ascribed to the production of external metabolites. led them to conclude that heat-stable substances By adding the filtrates from actively dividing, non- excreted by several species are lethal to other species. growing, or autolysing cultures, they demonstrated Their effects are species-specific and each species that actively growing cells of T. nitzschioides do has different patterns and degrees of inhibition de- not substantially affect the growth of S. costatum. pending also on the test organism. The most active On the contrary, the latter species releases substances algae are diatoms, among which P. tricornutum capable of inhibiting active division of T. nitzschioid- appeared to be the most efficient. es, even in the early stage of its own growth, that Altogether, the above-cited contributions left allows it to become dominant in mixed cultures. unclear whether or not the excreted antialgal sub- Filtrate from stationary phase or autolysing cells of stances act at naturally occurring concentrations. both species have no inhibiting effect; on the contrary, Such a lack greatly diminishes most published state- cell-free filtrates of aged S. costatum are beneficial ments or generalizations. However, Honjo et al. to T. nitzschioides, explaining why the latter alga (1978) attempted to use natural water samples and, cannot be displaced by the former in mixed cultures, subsequently, their results could contribute to as- when cultured over a long period. These results are certaining the reality of the allelopathic relationships interesting in that they demonstrate that the sub- among phytoplankters. Daily sampling and counting stances excreted by an alga might have different allowed these authors to observe that, in Hakata effects according to the phase of growth and sub- Bay (Japan), the phytoplankton species association sequent algal physiological status. However, because and dominance changed, at the end of spring, from the experiments were done at high cell densities several diatoms and nanoflagellates to dinofiagellates (up to 10" cells • ) extension of conclusions to and an intensive bloom of Heterosigma sp. These the relations between partners in a natural association observations led the authors to assume that species is hazardous (as Fedorov and Kustenko admit). associated in dominance did not compete with each Kustenko (1979) improved the method used, but other for growth. On the contrary, the species of the does not report any new pertinent data. first group were inhibited by those of the second Another interaction between the marine diatoms group; namely, the blooming of the large diatom Thalassiosira pseudonana and Phaeodactylum Heterosigma sp. temporarily inhibited the growth tricomutum has been studied by Sharp et al. (1979), of the nanoflagellates and completely prevented with the goal of designing a protocol to favor the the growth of most diatoms. Experiments made growth of the first and eliminate the second from with culture filtrates and nutrient-enriched seawater mass cultures used to feed commercial shellfishes. collected during Heterosignia sp. bloom, conducted Phaeodactylum tricomutum, which is considered to with locally isolated strains, demonstrated that be a poor food and, therefore, undesirable, dominates Heterosigma sp. releases an inhibiting ectocrine continuous mass cultures and eliminates other species. acting against most diatoms, especially Skeletonema Experiments done with both crossed subculturing costatum, but has no effect against others, i.e. in filtered reenriched culture media and batch or Nitzschia closterium and Phaeodactylum tricornu- continuous bialgal cultures allowed the observation tum, which are able to remain present during the in laboratory conditions of the inhibition of S. cos- in situ Heterosigma sp. bloom. Unfortunately, the tatum growth by a heat labile substance released authors seem not to have investigated and separated in the medium by P. triconuttum. Because, in addi- the respective effects of the nutrient competition and tion, none of the relative growth constants of the two the production of inimical extocrines on the observed species can explain the dominance of P. tricornutum, species succession. it was concluded by Sharp et al. that an allelopathic effect was responsible for this process. However, DINOFLAGELLATES AS GROWTH-INHIBIT1NG the inimical effects arise from populations that are SUBSTANCE PRODUCERS three orders of magnitude denser than natural phyto- plankton cell densities. Phaeodactylum tricornutum The association of dense populations of marine is always rare in natural population and the authors do dinoflagellates and toxic effects for a large variety not try to extend their results to natural waters, but of organisms has been known for many years (Kofoid

330 1911; Liebert and Deerns 1920). Accumulation of lation (10" - 2 x 10" cells • L- ') might well information on the toxic substances released by these have overtaken the maximum natural population peculiar algae is impressive, indeed. In spite of the (3 x 10" cells • L- ' ) he observed. disappearance of other taxa of algae during the so- Gauthier et al. (1975) reported another study called red-tide blooms of toxic dinoflagellates, in which field observations and laboratory experi- relatively little attention has been paid to the advan- ments are investigated. They surveyed the evolution tage given to them by their external metabolites in of bacterial and algal populations, nutrient concen- competition for nutrients and population dominance. trations, and temperature in shallow coastal Medi- Pincemin (1969) observed the inhibition of terranean waters. They observed that from winter growth of a small Chaetoceros in Mediterranean to summer the densities of diatom and flagellate waters on development of a red tide of Cochlodiniutn populations decrease continuously; the dominant sp. and claimed the presence of a strong inhibiting diatoms were such species as Skeletonema costatum, substance. In further research Pincemin (1971) Asterionella japonica , Phaeodactylum tricornututn , experimented with another diatom, Asterionella and some Navicula species. In warm summer waters, japon ica and another dinoflagellate, Glettodinium blooms of green flagellates may occur, but they end monotis . With cross-subculturings and reenrichments when oxygen content is reduced by bacterial activity. of the cultures he demonstrated the production of The green flagellates are then replaced by some autoinhibiting substances by both species and the dinoflagellates, namely Prorocentrutn micans , toxic effect for the diatom of the substances released Gymnodiniutn sp., and Glenodiniutn sp. Diatoms by the dinoflagellate. Unfortunately, he did not pay and green algae reappear only when the dinoflagel- attention to the pH increase and CO., concentration lates have disappeared and the oxygen content decrease subsequent to the first culture, and used has been restored. To explain this succession, the cultured strains from different origins. Altogether, authors do not invoke nutrient concentration and this work (the first to involve a dinoflagellate in a do not give any information about their seasonal mixed culture) did not avoid the shortcomings of the evolution, but from one of their figures one can older research and, hence, brings little new infor- read that in surface water the nutrient content is mation. In a related work Pincemin and other workers very high (P-PO = 240 ,us-at • L - ' ; N-NO, = of the same laboratory (Aubert et al. 1970; Gauthier 67 ,u,g-at • L - ' ), reflecting a eutrophicated situation. et al. 1978) reported that the dinoflagellate Proro- Gauthier et al. suggested a chemical interaction centrum micans releases some substances which between the above species. To prove their hypothesis, inhibit the production of an antibacterial compound they isolated some of the main species, i.e. P. tnicans , by the diatoms Asterionellajaponica and Chaetoceros S. costatutn , A. japon ica , and Tetraselmis inaculata , lauderi , which remain otherwise capable of growth. and carried out several bicultures and crossed uni- This might be an interesting secondary effect of the cultures in culture filtrates. Having observed some heteroantagonism between algal species. Further inhibition of growth of T. tnaculata when cultured research is needed to confirm this effect and assess in these conditions, they concluded that the observed its importance in situ, because the strain used to algal competition is a result of the effect of external test the antibacterial effect was a Staphylococcus metabolites released by the algae. Their statement aureus , obviously not an aquatic bacterium. Also, seems to be convincing, but the time has come to be the substances were extracted from heavy bulks of cautious with such extension of conclusions from packed cells and were active at concentrations which laboratory experiments to natural conditions. Gau- represent several thousands times the cell densities thier et al. grew their strains in a nutrient-rich medium of the two diatoms in the marine phytoplankton. (Provasoli's medium). Their apparently most con- Production of diatom-inhibitory substance by P. vincing experiment consists of injecting a small micans in nutrient-rich medium was also described inoculum of the second alga into a well-established by Uchida (1977, 1981). In bialgal cultures, the culture of the first alga. For instance, 1840 cells dinoflagellate overcomes Skeletonema costatum of T. maculata are mixed with 5490 cells of P. and Chaetoceros didytnus and, when subcultured tnicans (resident species); because T. maculata di- in cell-free culture filtrate of P. tnicans reenriched vides only 7.6 times (whereas in a fresh medium with fresh nutrients, the two diatoms are also in- control such an inoculum divides 9.9 times), it is hibited. In Uchida's opinion, the production of concluded that T. maculata is inhibited by substances diatom-inhibiting substance might explain the released by P. micans , which divides as much as in bloom of P. micans , which succeeds the diatom unialgal culture. Such a slight difference is not fully flowering in natural Waters of the Japanese coast. convincing because the experimental device does not However, Uchida reports that, in mixed cultures, offset the role of competition for nutrient partition inhibition of diatom growth begins only after 3 d that is related more to the respective specific kinetic incubation; hence, the initial dinoflagellate popu- constants of uptalce than to the cell densities.

331 Elbrâchter (1976) cultivated two dinoflagellates terium inhibits the growth of the dinoflagellate. in bialgal cultures: Prorocentrum micans and Gym- Further research (Kutt and Martin 1975; Martin and nodinium splendens and also did several crossed Martin 1976; McCoy and Martin 1977; McCoy et cultures with reenriched culture filtrates. The two al. 1979; Moon and Martin 1979) demonstrated that species do not show the same growth capacity. It the inhibition did not result from nutrient competition is presumed that G. splendens can excrete an in- but from the production of an antialgal substance hibiting factor affecting the growth of P. oilcans , by the cyanobacterium. This substance, named whereas the effect on G. splendens production by "aponin," was first extracted from packed cells and P. micans is limited to nutrient competition. Thus later from cell-free culture medium, indicating that the behavior of a mixed culture depends greatly on it is primarily an exotoxin. The crude aponin is heat the relative cell proportions; the toxigenic species stable to 110°C, and acid stable. In nutrient-rich can dominate the other one only when enough cells medium the lethal effect is obtained with 6 x 10 6 are present to produce the ectocrine above the acting of G. aponina, whereas the minimum cells • concentration. Elbrâchter points out that because the detectable effect on G. breve after 24 h of incubation cell densities normally involved do not occur in is achieved by only 60 ,u,g • of crude aponin. nature and because the toxic inhibition was not lethal This substance was purified by a broad array of but extended the lag phase, the results do not con- biochemical treatments, including HPLC, and proved stitute a direct proof. However he is convinced that to be a chemically defined sterol. Aponin can reduce the results he obtained in vitro also apply to natural the production of ichthyotoxic substances by G. habitats. In a further development of this research, breve and, hence, may mitigate the effects of the Elbrâchter (1977) ran several combinations of mixed dinoflagellate products on marine life, but it has no cultures of two diatoms — Biddulphia regia and direct effect on fishes, crustaceans, and bacteria. Coscinodiscus concinnus — and two dinoflagellates Aponin is a substance specifically antialgal and — Cerntium horridum and Prorocentrum micans antifungal. Likewise, the toxin produced by G. breve Cultures were initiated with cell densities and nu- h—as no effect on G. aponina. On the other hand, trient levels corresponding to those prevailing in considering the responses to temperature and com- the natural environment from which they were all paring with naturally occurring temperatures in sur- isolated. From data obtained, he could state that face waters of coastal Florida waters, Martin and changes in growth rates, which resulted for each associates postulated that G. breve should have species from association with other species, were optimal proliferation during midspring and late not due to inhibition caused by excreted substances, autumn, and G. aponina may proliferate during but rather by nutrient competition. Unfortunately, late spring and early summer and autumn. As a this conclusion concerned all the nutrients as a whole matter of fact, several red tides have occurred during and was not investigated in more detail. In other late autumn and one during midspring, exactly during words, not enough was done to describe how these the postulated periods, namely when G. aponina species adapt to the nutrient competition. Kayser cannot grow, supporting the opinion that the mech- (1979) used batch cultures, crossed subculturing in anisms postulated from in vitro experiments really cell-free culture filtrates, and continuous cultures occur in situ. Ultimately, these scientists suggest of combinations of three dinoflagellates — Gym- using the aponin or G. aponina freezed-dried cells nodinium splendens, prorocentrum micans, and as a biological tool in red-tide management. Scrippsiella faeroense — also isolated from Hel- goland waters, with a view to separating the phe- FIELD OBSERVATIONS, LABORATORY-STUDIED nomena of nutrient competition and interactions of PROCESSES, AND LOCALLY ISOLATED STRAINS external metabolites. From the data obtained, he INVOLVED IN THE SAME INVESTIGATION expo- could state that in multispecies populations the Keating (1977, 1978) reported that observations nential growth is mainly regulated by nutrient com- made over a 5-yr period in a eutrophicated lalce petition, while inhibiting metabolic products act indicate that diatom blooms vary inversely with the secondarily at maximum cell densities. Whether density of the preceding cyanophyta populations. this inhibiting effect is caused by toxic algal excre- A successsion of dominant cyanophyte species was tions or bacterial decomposition products of dead also recorded. Then, she tried to explain these pro- cells could not be determined. cesses by ecophysiological mechanisms. By using During the course of investigations of a red seven crude (only filtered) lake waters collected tide, Kutt and Martin (1974) isolated from the same over an 18-mo period for bioassays, she demon- sample a toxigenic dinoflagellate (Gymnodinium strated that lake waters contain a heat-labile sub- breve) and a cyanobacterium (Gomphosphaeria stance that inhibits diatom growth. Then, subculturing aponina) which they cultivated in mixed culture. in cell-free culture filtrates confirmed that the pro- They observed that the presence of the cyanobac- ducers of these substances were cyanophyta rather

332 than other organisms. Succession of dominant cyano- the species succession, but, the results demonstrate phyta is also explained by allelopathic effects among that ectocrines are certainly part of the fine control species. By bioassays with predecessors and succes- in determination of bloom sequence (Keating 1977). sors of a dominant species, she demonstrated that Hence, their role is not negligible, even if the coarse dominant species produce only inhibiting or neutral controls, such as light and macronutrients, determine effects on their immediate predecessors, and, in the growth potential of the system and the major contrast, only beneficial or neutral effect on their categories of dominant algae. immediate successors. Hence, both patterns of inter- The physiological mechanism of the algal action enhance the competitive position of the or- inhibition has also been little studied, but Murphy ganism that will dominate an impending bloom. As et al. (1976) remarked that in a eutrophic lake the a matter of fact, in vitro allelopathic effects are well sudden dominance of cyanophytes of genus Anaboena correlated with annual in situ succession patterns. coincided with the production of strong iron chelators Keating's experiments were done with eight cyano- and an active uptake of iron. Several experiments, phytes isolated from the same waters. Diatoms tested ascertained by further research of Bailey and Taub for their susceptibility to crude lake water or to (1980), led them to propose a hypothesis: cyano- cyanophyta external products have also the same phyta have a strong iron uptake system and excrete origin. There were from 7 to 9 species or, excep- hydroxamate siderochromes that act as strong iron tionally, 3, 13, 18, and 29 species. Each cyanophyte binders. Thus, the cyanophytes remain capable of appears to produce a substance inhibiting most and taking up iron while they suppress the availability even all diatoms, but effects against diatoms were of iron to other species, which are inhibited, there- notably decreased when strains isolated from other fore, by a secondary nutrient starvation effect. As lakes were used for comparison. The role of bacteria hydroxamates were found in supernatant of ultra- was also investigated by using axenic or bacterized filtered natural lake waters, this process might be algal strains. They have no qualitative effect but of ecological interest and a promising research di- decrease the inhibition of diatom growth, probably rection, but, of course, it does not exclude the direct by degrading the active substances. This bacterial poisonous effect of inimical ectocrines on one or role might turn out to be critical, because Keating several biochemical pathways. also demonstrated that reducing filtrates to 33% of Another example of antagonism between or- normal strength by dilution eliminates the inhibition, ganisms belonging to different taxonomic groups proving that the substances act like antibacterial was described in marine waters. Weekly observations antibiotics, i.e. they are inhibiting above a threshold by Pratt (1965), over a 7-yr period, showed that concentration and have no action or are enhancing the phytoplankton of the Narragansett Bay is domi- (hormesis) below this concentration. In addition, nated during spring and summer by alternate brief the substances released by the cyanophyta are not blooms of the diatom Skeletonema costatum and all toxic for diatoms. Physically extracted (ultra- the chrysophycean flagellate Olithodiscus luteus . filtration) substances from culture filtrate of Ana- "Environmental factors measured did not provide boena holsaticutn are shown by chemical separation any clue to the sequence and did not appear to favor to contain a diatom growth promoting compound either species over the other, but there were some associated with an inhibiting one. The effect of the indications that each actor awaits the other's exit former compound is masked in full-strength filtrates before coming on stage," said Pratt, who suggested but appears when filtrates are diluted. Keating's that the succession is mediated by an interaction contribution is impressive, indeed. By selecting a between the two algae whose mechanism was not broad array of naturally cooccuring strains and using indicated by the field data. Then, he continued his algal biomass and nutrient concentrations at eco- investigation with in vitro experiments (Pratt 1966). logical levels (e.g. she adds 12.3 ,u,g—at—N • L- ', By the use of bialgal cultures and crossed subculturing 0.57 tg—at—P • L- ', and other nutrients at balanced in nutrient reenriched filtrates, he demonstrated that levels), she overcame the main shortcomings that a concentration of more than 100 x 10" cells • L -1 led to questionable results in the past. Prior to inocu- of O. luteus inhibits the growth of S. costatum by lation of test cells, she allowed CO 2 content and pH secreting a tanninlike substance. A concentration to come to their original values after autoclaving, of less than this value stimulates the growth of the if any, and she duplicated experiments with axenic diatom, with a maximal effect at 20 x 10 cells • L- ', or bacterized cultures, something never done before, but effective up to 4 x 10" cells • Pratt's dem- at least in experhnents designed with all refinements onstration provided a fascinating theory to explain at one time. Unfortunately, the growth characteristics the natural processes he observed: when circum- (K, Vm„, , etc.) of the cyanophyta were not estab- stances combine to give O. luteus a sufficient relative lished, so one cannot separate clearly nutrient com- abundance, it suppresses the growth of S. costatum petition from the role of the external metabolites in by release of large quantities of an ectocrine. When

333 S. costatum outstrips O. luteus in competition, it the dialysis walls allow the passage of the substance does so primarily by virtue of its much greater rate released by O. luteus (M < 2000) (Tomas 1980) of multiplication, thus monopolizing the available the inhibition should have been observed, if it was nutrients. But, further, in small competing popu- there. Thus, if the ectocrine inhibition might occur lations, the slight amount of the presumed tannin for short periods of time in restricted areas where produced by the O. luteus cells favors S. costatum maximum densities are high, no clear explanation growth, perhaps by metal chelation. Unfortunately, of the apparent reciprocal codominance of S. cos- cell concentrations required to produce the inimical maim and O. luteus in Narragansett Bay is yet ectocrine were two orders of magnitude higher than available. In addition, Tomas observed there is no the highest cell density of O. luteus in Narragansett apparent predator on O. luteus , whereas S. costatum Bay. Pratt stated this criticism himself and called is actively grazed on by small zooplankton. During for further work. The challenge was taken up by the bimodal periods of increasing abundance of O. Stuart (1972), who first demonstrated that prevention luteus (May and October), large populations of of S. costatum growth in O. luteus conditioned zooplankton are present and S. costatum is scarce. and nutrient-restored medium was only temporary; During the midsummer decline of the chrysophytes, with time, rapid growth occurred. Hence, final cell copepods are reduced to low levels by ctenophore population, maximal biomass, and growth rate were predation and the diatom blooms. Hence, the decrease unaffected. Because autoclaving removed inhibition of O. luteus and concomitant increase of S. costatum and doubling trace metals caused mortality, he could and vice versa may be events mediated more by suggest that chelation and trace metal availability grazing pressure than through direct physiological might be involved in the inhibition. Then chemical interactions. This raises a new question: why is separations indicated that substances having a molar O. luteus avoided by zooplankters? Observations mass below 2000 cause the lag effect in the growth reported by Tomas (1980) of noxious effects on of S. costatwn. However, conversely, in bialgal zooplankton and even fishes again call attention to culture, S. costatum takes the upper hand over O. the inimical ectocrines released by this species and luteus which grows poorly, having a longer lag the need for more research. phase and a lower growth rate than those of the diatom; thus, there is a discrepancy with Pratt's IMPORTANCE OF GROWTH-INHIBITING (1966) findings. This probably originates in the dif- ECTOCRINES IN ALGAL COMPETITION ferent cell densities involved, wrote Stuart (1972), AND SPECIES SUCCESSION who used 27-190 x 10" cells • I.,' instead of the 400-452 x 10" cells • used by Pratt. But the That some algae in their growth stages are able latter also pointed out that the inhibition begins with to produce and release substances that are inhibiting 190 x 10" cells • of O. luteus . In addition, to other algae is well confirmed. However, three both Pratt and Stuart used nonaxenic strains whose important questions remain: (i) How many species bacteria might have enhanced or suppressed any are capable of it? (ii) Do these substances really effect of extracellular substances. Thus, altogether act in situ? (iii) When do they confer a stronger their results do not unequivocably confirm the inhi- advantage, if at all, to the producing alga than other bition of S. costatum by these substances liberated physiological processes? The answers to these ques- by O. luteus. Further research was still needed. tions will state the real importance of this physio- Tomas's (1978, 1979, 1980) continuing studies of logical adaptation. the autoecology of O. luteus provided physiological Except for a few recent contributions, the strains information for this species that allowed him to used in experiments were choosen mainly for their compare the two competitors on the basis of a broad availability in culture collections. Thus, the total array of physiological patterns. But it appears that number of species known to produce antialgal ecto- this chrysophyte has not any exceptional abilities crines does not represent more than a couple of dozen for the uptake and utilization of nutrients, notwith- for freshwater algae and even less for marine phyto- standing that it is reported to occur in spring blooms plankton. This small array might be readily increased in nutrient-depleted waters. Likewise, cells divide if all or most species could be tested. On the other actively in a broad range of temperature and light hand, the number of species producing ectocrines intensities, but so do those of S. costatum. Hence, relative to the total number of species would be of most mechanisms that act in species competition greater ecological importance, by allowing an esti- seem to be little involved. Tomas (1980) reexamined mate of the average numbers of ectocrine-producing the Pratt hypothesis of biochemical interaction be- cells in a natural population under study; this has tween the two species and stated that S. costatum not been done yet. Collecting strains from different grows well in dialysis sacs incubated in situ during origins may build artificial communities that never periods when O. luteus is naturally abundant. Because occur in situ, but modern researchers usually avoid

334 this problem and study naturally occurring com- From the few quantitative data reported, these thresh- petitors. Nevertheless, only two or a few species old concentrations seem to be high; 10-50 mg • are involved in experiments and, unfortunately, this (Jüttner 1979) and 25 mg • I.» (McCracken et al. is not sufficient. Keating (1978) demonstrated that 1980). Obviously, these are far above the average different isolates belonging to the same species, but concentration of dissolved organic matter in aquatic from different origins, may have different responses, habitats, and it is unreasonable to postulate that all which called in question most published data. In dissolved organic matter could be composed of one addition, the few species involved in experiments substance, namely the inhibiting one. However, are mostly competing dominant species, which can some algae may produce external metabolites acting explain the reciprocal codominance of these species at very low levels, as low as 60 ,ug • as reported through production of inimical ectocrines, but leaves by Martin and associates. Because of the growing the outcome over ôther species unstudied. Only prevalence of eutrophication in fresh waters and Keating tested most naturally companion species coastal marine waters, subsequent algal blooming of several dominant algae-producing ectocrines. With may produce cell densities as high as those that such an approach, she could demonstrate that most cause inhibiting effects in laboratory conditions. cyanophytes of the lake she studied produce sub- The use of batch culturing is probably the main stances that inhibit most, if not all, the diatoms cause of artifacts, in that it allows the substances living in the same waters. By choosing the susceptible excreted by algae to accumulate, while in nature strains, other authors might have overestimated dra- they are immediately diluted. This dispersal is more matically the inhibiting effects of ectocrines, when critical for oceans because the volume/surface ratio considered in the natural population as a whole. is far greater and, therefore, algae are concentrated In addition, for marine algae, almost all the strains in upper layer. The remark of Tomas (1980), who used in experiments were isolated from coastal points out that Skeletonema costatum is strongly waters. Because of the great differences between inhibited in vitro by Olisthodiscus luteus but grows neritic and oceanic waters in nutrient content and well in situ when enclosed in dialysis sacks and phytoplankton composition, extension of published immersed in O. luteus conditioned waters, is of results to the world ocean could be misleading. Thus, primary importance and should prevent premature all told, the relative number of algae-producing extrapolation and imprudent generalization. A con- antialgal external metabolites is still unknown, even tinuous culturing with an apparatus as developed tentatively, either for fresh waters or marine species. by Kroes (1973) is certainly a more pertinent ap- That antialgal ectocrines are released by several proach, but one must remark that data obtained by algae does not necessarily mean the substances are Elbrüchter (1976) and Kayser (1979) diminish the active in situ or provide an obvious advantage to the role of external metabolites, when nutrients are added producers in species competition. Based on labora- at natural levels. The presence of bacteria is a source tory experiments, some previous statements like those of confusion; to prove that the inhibitions observed of Hartman (1960) and Lefevre (1964), to cite only are really caused by algae, axenic strains are needed, the main reviewers, have certainly overestimated but, on the other hand, absence of bacteria is mis- this case by ignoring the criticisms and/or gener- leading because they naturally destroy the dissolved alizing to the global scale a few results belonging organic matter and thus prevent accumulation, if to discrete communities. During the past decade, any. This suggests we use axenic strains only for most experimental shortcomings have been over- preliminary survey of potential ectocrine-producing come. Hence, several contributions have ascertained algae, and naturally bacterized strains for critical ex- the potentiality of the nontrophic relationships among periments. Because demonstration of direct inhibition algal populations, but pertinent demonstrations of by crude natural waters (only filtered) of several inimical ectocrines acting in situ are scarce and isolated algae has been done only once (Keating related to particular conditions. As a matter of fact, 1978), a clear answer cannot be given to the second most laboratory experiments are biased in that they question. Nevertheless, we think that recent contri- involve nutrient-rich culture media and subsequent butions supply growing evidence that aquatic habitats dense algal populations that far exceed those occur- should be classified according to a nutrient criterion. ring in situ. Hence, the substances excreted may Those where naturally occurring conditions make reach very high concentrations that pass the threshold nutrient-rich waters, and those which are eutro- concentration for inhibiting activity. Computations phicated may support a heavy algal biomass that to transfer results to in situ conditions by dividing can produce great quantities of external metabolites; the inhibiting effect by the ratio of in vitro/in situ hence, these substances may really act in situ. Such algal densities or ectocrines concentrations are cer- conditions are certainly more frequent in inland tainly wrong, as the external metabolite is ineffec- waters, because land drainage of natural substances tive or even enhancing below the activity threshold. and farm fertilizers enrich stagnant waters (lake or

335 pond waters where these "water blooms" of a few investigating allelopathy among marine microalgae. species occur impressed Lefevre (1964)). Shallow Mar. Biol. 59: 7-13. coastal waters may be similar; for instance, those ELBRF+CHTER, M. 1976. Population dynamic studies on studied by Gauthier et al. (1975, 1978) were far phytoplankton cultures. Mar. Biol. 35: 201-209. richer in nutrients than the natural Mediterranean 1977. On population dynamics in multi-species cultures of diatoms and dinoflagellates. Helgol. Wiss. waters and almost free from mixing. All things Meeresunters. 30: 192-200. considered, within the red-tide plumes, ectocrine FEDOROV, V. D., AND N. G. KUSTENKO. 1972. Com- inhibition is probably a means of competition for petition between marine planktonic diatoms in mono- the dominant species. On the other hand, in oligo- culture and mixed culture. Oceanology 12: 91-100. trophic waters (e.g. all oceanic waters and mountain FITZGERALD, G. P. 1969. Some factors in the competition lakes) all published data are consistent with the or antagonism among bacteria, algae, and aquatic view that exte rnal metabolites have no role in algal weeds. J. Phycol. 5: 351-359. competition or in species succession, but no direct FOGG, G. E. 1962. Extracellular products, p. 475-489. proof is available. In R. A. Lewin [ed.] Physiology and biochemistry of algae. Academic Press, New York, NY. FRENCH, C. S. 1979. Fifty years of photosynthesis. Annu. Acknowledgments Rev. Plant Physiol. 30: 1-26. GAUSE, G. F. 1932. Experimental studies on the struggle We wish to gratefully thank Dr Trevor Platt for existence. I. Mixed populations of two species for revising the English manuscript. of yeast. J. Exp. Biol. 9: 389-402. GAUTHIER, M. J., P. BERNARD, AND M. AUBERT. 1978. Modification de la fonction antibiotique de deux References diatomées marines, Asterionella japonica (Cleve) et Chaetoceros lauderi (Ralfs) par le dinoflagellé Pro- AKEHURST, S. C. 1931. Observations on pond life, with rocentrum micans (Ehrenberg). J. Exp. Mar. Biol. special reference to the possible causation of swarming Ecol. 33: 37-50. of phytoplankton. J. R. Microsc. Soc. 51: 237-265. GAUTHIER, M. J., J. P. BREITTMAYER, AND M. AUBERT. ALLEE, W. C. 1934. Recent studies in mass physiology. 1975. Étude des facteurs responsables de dérives Biol. Rev. Camb. Philos. Soc. 9: 1-48. écologiques en milieu marin méditerranéen côtier. AUBERT, M. 1971. Télémédiateurs chimiques et équilibre 10th Eur. Symp. Mar. Biol., Ostend, Belgium, Sept. biologique océanique. Première partie. Théorie géné- 17-23. Vol. 2: 271-283. rale. Rev. Int. Oceanogr. Med. 21: 3-16. HARDER, R. 1917. Ernahrungsphysiologische Unter- 1978. Télémédiateurs et rapports inter-espèces suchungen an Cyanophyceen, hauptsâchlich dem dans le domaine des micro-organismes marins, p. endophytischen Nostoc punctiforine. Z. Bot. 9: 145- 179-200. In Actualités de biochimie marine. Colloq. 242. GABIM-CNRS. HARRIS, D. 0. 1970. An autoinhibitory substance pro- AUBERT, M., D. PESANDO, AND J. M. PINCEMIN. 1970. duced by Platydorina caudata Kofoid. Plant Physiol. Médiateur chimique et relations inter-espèces. Mise en 45: 210-214. évidence d'un inhibiteur de synthèse métabolique d'une 1971. Growth inhibitors produced by the green diatomée produit par un péridinien (étude "in vitro"). algae (Volvocaceae). Arch. Mikrobiol. 76: 47-50. Rev. Int. Oceanogr. Med. 17: 5-21. HARTMAN, R. T. 1960. Algae and metabolites of natural BAILEY, K. M., AND F. B. TAUB. 1980. Effects of waters, p. 38-55. In C. A. Tryon Jr. and R. T. hydroxamate siderophores (strong Fe (III) chelators) Hartman [ed.] The ecology of algae. Univ. Pittsburgh on the growth of algae. J. Phycol. 16: 334-339. Press. BERLAND, B. R., D. J. BONIN, AND S. Y. MAESTRINI. HELLEBUST, J. A. 1974. Extracellular products, p. 838- 1972a. Are some bacteria toxic for marine algae? 863. In W. D. P. Stewart [cd.] Algal physiology and Mar. Biol. 12: 189-192. Biochemistry. Blackwell Scientific Publications, 1972b. Étude des relations algues-bactéries du Oxford. milieu marin : possibilités d'inhibition des algues par HERRMANN, V., AND F. JI/TTNER. 1977. Excretion les bactéries. Tethys 4: 339-348. products of algae. Identification of biogenic amines 1973. Study of bacteria inhibiting marine algae: by gas-liquid chromatography and mass spectrometry a method of screening which uses gliding algae. Mem. of their trifluoroacetamides. Anal. Biochem. 78: 365- Biol. Mar. Oceanogr. 3: 1-10. 373. 1974. Importance des substances inhibitrices dans HONJO, T., T. SHIMOUSE, N. UEDA, AND T. HANAOKA. le contrôle des populations d'algues et de bactéries 1978. Changes of phytoplankton composition and its du plancton marin. Mem. Biol. Mar. Oceanogr. 4: characteristics during red tide season. Bull. Plankton 63-97. Soc. Jpn. 25: 13-19. BURKHOLDER, P. R., R. M. PFISTER, AND F. M. LEITZ. JOHNSTON, R. 1972. The theories of August Putter. Proc. 1966. Production of a pyrrole antibiotic by a marine R. Soc. Edinb. Sect. B 72: 401-409.

bacterium. Appl. Microbiol. 14: 649-653. JÜTTNER, F. 1976. p - cyclocitral and alkanes in Micro- CHAN, A. T., R. J. ANDERSEN, M. J. LE BLANC, AND cystis (Cyanophyceae). Z. Naturforsch. Teil C 31: P. J. HARRISON. 1980. Algae plating as a tool for 491-495.

336 1979. Nor-carotenoids as the major volatile LEFEVRE, M., H. JAKOB, AND M. NISBET. 1952. Auto excretion products of Cyanidium • Z. Naturforsch. et hétéroantagonisme chez les algues d'eau in vitro et Teil C 34: 186-191. dans les collections d'eau naturelles. Ann. Stn. Cent. KAYSER, H. 1979. Growth interactions between marine Hydrobiol. Appl. 4: 5-199. dinoflagellates in multispecies culture experiments. LIEBERT, F., AND W. M. DEERNS. 1920. Onderzoek Mar. Biol. 52: 357-369. naar de Oozaak van een Vischsterfte in den polder KEATING, K. I. 1977. Allelopathic influence on blue-green Wookumer-Nieuwland, Nabij Workum. Verh. bloom sequence in a eutrophie lake. Science (Washing- Rijksinst . Visschonderz . 1: 81. ton, D.C.) 196: 885-887. LINDROTH, P., AND K. MOPPER. 1979. High perfor- 1978. Blue-green algal inhibition of diatom mance liquid chromatographic determination of growth: transition from mesotrophic to eutrophic subpicomole amounts of amino acids by precolumn community structure. Science (Washington, D.C.) fluorescence derivatization with 0-phthaldialdehyde. 199: 971-973. Anal. Chem. 51: 1667-1674. LUCAS, C. E. 1938. Some aspects of integration in plankton KOFOID, C. A. 1911. Dinoflagellata of the San Diego communities. J. Cons. Cons. Int. Explor. Mer. 8: region. IV. The genus Gonyaulax . Univ. Calif. Publ. 309-322. Zool. 8: 187-286. 1947. The ecological effects of external metabo- KRASIL'NIKOVA, E. N. 1964. Antagonism in marine micro- lites. Biol. Rev. 22: 270-295. organisms, p. 208-229. In A. E. Kriss, I. E. Mishus- 1955. External metabolites in the sea. Papers in tina, N. Mitskevich, and E. V. Zemtsova [cd.] marine biology and oceanography. Deep Sea Res. 3 population of oceans and seas. Nauka Microbial (suppl.): 139-148. Zhizn. (Transi, from Russian by G. E. Fogg) 1961. On the significance of external metabolites KROES, H. W. 1971. Growth interactions between Chla- in ecology. Symp. Soc. Exp. Biol. 15: 190-206. mydomonas globosa Snow and Chlorococcum ellip- MARTIN, D. F., AND B. B. MARTIN. 1976. Aponin, a soidewn Deason and Bold under different experimental cytolytic factor toward the red tide organism, Gym- conditions, with special attention to the role of pH. nodinium breve . Biological assay and preliminary Limnol. Oceanogr. 16: 869-879. characterization. J. Environ. Sci. Health Part A 11: 1972a. Growth interactions between Chlatnydo- 613-622. monas globosa Snow and Chlorococcum ellipsoideum McCoy, L. F. JR., AND D. F. MARTIN. 1977. The Deason and Bold: the role of extracellular products. influence of Gomphosphaeria aponitta on the growth Limnol. Oceanogr. 17: 423-432. of Gymnodiniwn breve and the effect of aponin on 19726. Extracellular products from Chlorococcum the ichthyotoxicity of Gyttmodinium breve. Chem. ellipsoidewn and Chlatnydomonas globosa . Arch. Biol. Interact. 17: 17-24.

Mikrobiol. 84: 270-274. McCoy, L. F. JR., D. L. ENG - WILMOT, AND D. T. 1973. A spin filter system for the study of algal MARTIN. 1979. Isolation and partial purification of interactions. Oecologia Berl. 11: 93-98. a red tide (Gymnodinium breve) cytolytic factor (s) KROGH, A. 1931. Dissolved substances as food of aquatic from cultures of Gomphosphaeria aponina . Agric. organisms. Rapp. P. V. Reun. Cons. Int. Explor. Food Chem. 27: 69-74. Mer. 75: 7-36. MCCRACKEN, M. D., R. E. MIDDAUGH, AND R. S. KRZYWICKA, A., AND D. KRUPA. 1975. Preliminary MIDDAUGH. 1980. A chemical characterization of investigations on mutual growth relations of the an algal inhibitor obtained from Chlamydomonas • populations of the blue-green alga Microcystis aeru- Hydrobiologia 70 (3): 271-276. gitzosa and green algae Monoraphidiwn minutum and MCVEIGH, I., AND W. H. BROWN. 1954. In vitro growth Scenedesmus abundans in bicultures. Acta Hydrobiol. of Chlamydomonas chlatnydogattla Bold and Haema- 17: 81-88. tococcus pluvialis Flotow Em. Wille in mixed cultures. KUSTENKO, N. G. 1979. Application of factor experiment Bull. Torrey Bot. Club. 81: 218-233. plans in studies of algae relationship in culture. Gidro- MOON, R. E., AND D. F. MARTIN. 1979. Potential biol. Zh. 15: 79-80. (In Russian) management of red tide blooms. Treatment with con- centrated frozen Gomphosphaeria aponina cultures. KUTT, E. C., AND D. F. MARTIN. 1974. Effect of se- lected surfactants on the growth characteristics of J. Environ. Sci. Health Part A 14: 195-199. MURPHY, T. P., D. R. S. LEAN, AND C. NALEWAJKO. Gymodiniutn breve . Mar. Biol. 28: 253-259. 1975. Report on a biochemical red tide repressive 1976. Blue-green algae: their excretion of iron-selective agent. Environ. Lett. 9: 195-208. chelators enables them to dominate other algae. Science (Washington, D.C.) 192: 900-902. LAM, C. W. Y., AND W. B. SILVESTER. 1979. Growth NIGRELLI, R. F. 1958. Dutchman's baccy juice of growth- interactions among blue-green (Anabaena oscilla- promoting and growth-inhibiting substances of marine rioides, Microcystis aeruginosa) and green (Chlorella origin. Trans. N.Y. Acad. Sci. Sen. II(20): 248-262. sp.) algae. Hydrobiologia 63: 135-143. PINCEMIN, J. M. 1969. Apparition d'une eau rouge a LANGE, W. 1974. Competitive exclusion among three Cochlodiniwn sp. devant Juan-les-Pins. Rev. Int. planktonic blue-green algal species. J. Phycol. 10: Oceanogr. Med. 13/14: 205-216. 411-414. 1971. Télémédiateurs chimiques et équilibre LEFEVRE, M. 1964. Extracellular products of algae, p. biologique océanique. Troisième partie. Étude in vitro 337-367. In D. F. Jackson [cd.] Algae and man. de relations entre populations phytoplanctoniques. Plenum Press, New York, NY. Rev. Int. Oceanogr. Med. 22/23: 165-196.

337

_1 PINTNER, I. J., AND V. ALTMEYER. 1979. Vitamin SCOTT, T. G., AND A. J. S. BALL. 1975. An apparatus B 12-binder and other algal inhibitors. J. Phycol. 15: for studying syntrophic interaction between micro- 391-398. organisms. Limnol. Oceanogr. 20: 291-295. POURRIOT, R. 1966. Métabolites externes et interactions SHARP, J. H., P. A. UNDERHILL, AND D. J. HUGHES. biochimiques chez les organismes aquatiques. Ann. 1979. Interaction (allelopathy) between marine dia- Biol. 5: 337-374. toms: Thalassiosira pseudonana and Phaeodactylum PRATT, D. M. 1965. The winter-spring diatom flowering tricornutum. J. Phycol. 15: 353-362. in Narragansett Bay. Limnol. Oceanogr. 10: 173-184. SIEBURTH, J. M. 1962. Biochemical warfare among the 1966. Competition between Skeletonenta costa- microbes of the sea. Honors lecture, Univ. Rhode bun and Olithodiscus luteus in Narragansett Bay and Island, Kingston. 13 p. in culture. Limnol. Oceanogr. 11: 447-455. 1964. Antibacterial substances produced by PRATT, R. 1940. Studies on Chlorella vulgaris . I. Influence marine algae. Dey. Ind. Microbiol. 5: 124-134. of the size of the inoculum on the growth of Chlorella 1968. The influence of algal antibiosis on the vulgaris in freshly prepared culture medium. Am. J. ecology of marine micro-organisms, p. 63-94. In Bot. 27: 52-56. M. R. Droop and E. J. Ferguson Wood [cd.] Advances 1942. Studies on Chlorella vulgaris . V. Some in microbiology of the sea. Vol. 1. Academic Press, properties of the growth-inhibitor formed by Cldorella London. cells. Am. J. Bot. 29: 142-148. STUART, M. 1972. The effects of Olisthodiscus luteus 1943. Studies on Chlorella vulgaris. VI. Re- Carter upon the growth of Skeletonetna costatum tardation of photosynthesis by growth-inhibiting sub- (Grey.) Cleve. M. S. Thesis, Univ. Rhode Island, stance from Chlorella vulgaris. Am. J. Bot. 30: 32-33. Kingston. 82 p. PRATT, R., AND J. FONG. 1940. Studies on Chlorella SZE, P., AND J. M. KINGSBURY. 1974. Interactions of vulgaris . II. Further evidence that Chlore/la cells phytoplankters cultured from a polluted saline lalce, form a growth-inhibiting substance. Am. J. Bot. 27: Onondaga Lake, New York. J. Phycol. 10: 5-8. 431-436. TASSIGNY, M., AND M. LEFEVRE. 1971. Auto., heté- 1944. Chlorellin, an antibacterial substance from roantagonisme et autres conséquences des excrétions Chlorella. Science (Washington, D.C.) 99: 351-352. d'algues d'eau douce ou thermale. Verh. Int. Ver. PROCTOR, V. W. 1957. Studies on algal antibiosis using Limnol. 19: 26-38. Haematococcus and Chlamydomonas Limnol. THOMSON, W. 1874. Les abîmes de la mer. Traduction Oceanogr. 2: 125-139. de 1875. PROVASOLI, L. 1963. Organic regulation of phytoplankton TOMAS, C. 1978. Olisthodiscus luteus (Chrysophyceae). fertility, p. 165-219. In M. N. Hill [ed.] The sea. I. Effects of salinity and temperature on growth, Vol. 2. J. Wiley & Sons, New York, NY. motility and survival. J. Phycol. 14: 309-313. PliTTER, A. 1907a. Die Ernührung der Wassertiere. Z. 1979. Olisthodiscus luteus (Chrysophyceae). III. Allg. Physiol. 7: 283-320. Uptake and utilization of nitrogen and phosphorus. 1907b. Der Stoffhaushalt des Meeres. Z. Allg. J. Phycol. 15: 5-12. Physiol. 7: 321-368. 1980. Olisthodiscus luteus (Chrysophyceae). V. RANSON, G. 1935. Le rôle de la matière organique dissoute Its occurrence, abundance and dynamics in Narra- dans l'eau et les théories de Putter. Bull. Mus. Natl. gansett Bay, Rhode Island. J. Phycol. 16: 157-166. Hist. Nat. 2 (7): 359-366. UCHIDA, T. 1977. Excretion of a diatom-inhibitory sub- 1936. Le rôle de la matière organique dissoute stance by Prorocentrum Mains Ehrenberg. Jpn. J. dans l'eau et les théories de Pütter (suite). Bull. Mus. Ecol. 27: 1-4. Natl. Hist. Nat. (2)8: 160-172. 1981. The relationships between Prorocentrum RICE, T. R. 1954. Biotic influences affecting population micans-growth and its ecological environment. Sci. growth of planktonic algae. Fish. Bull. 54: 227-245. Pap. Inst. Algol. Res. Fac. Sci. Hokkaido Univ. (In RICE, T. R., AND R. L. FERGUSON. 1975. Response of press) estuarine phytoplankton to environmental conditions, WAKSMAN, S. A. 1941. Antagonistic relations of micro- p. 1-43. In F. J. Vernberg [ed.] Physiological ecology organisms. Bacteriol. Rev. 5: 231-291. of estuarine organisms. Univ. South Carolina Press, WHITTAKER, R. H., AND P. P. FEENY. 1971. Allelo- Columbia, SC. chemics: chemical interactions between species. ROSENFELD, D. W., AND C. E. ZOBELL. 1947. Antibiotic Science (Washington, D.C.) 171: 757-770. production by marine micro-organisms. J. Bacteriol. 54: 393-398. WHITTLE, K. J. 1977. Marine organisms and their con- RYTHER, J. H. 1975. Physical models of integrated waste tribution to organic matter in the ocean. Mar. Chem. recycling marine polyculture systems. Aquaculture 5:381-411. 5: 163-177. WILLIAMS, P. J. LE B. 1975. Biological and chemical SAUNDERs, G. W. 1957. Interrelations of dissolved organic aspects of dissolved organic material in sea water, matter and phytoplankton. Bot. Rev. 23: 389-409. p. 301-363. In J. P. Riley and G. Skirrow [ed.] SCHENK, H. E. A., AND F. JI3TTNER. 1974. Nachweis Chemical oceanography, 2 nd Ed. Academic Press, extracellulher Murein-Hydrolase-Aktivitüt beim London. Zusammenbruch einer Bliite von Anabaena solitaria WOODRUFF, L. L. 1913. Effect of excretion products Kleb. Arch. Hydrobiol. 74: 1-7. of Infusoria. J. Exp. Zool. 14: 575-582.

338 Morphological Bases of Competition and Succession

ALAIN SOURNIA Laboratoire d'Ichtyologie, Muséum National d'Histoire Naturelle, 43, rue Cuvier, 75231 Paris 05, France

My idea about this school is that would rather consider them as bacteria. They are it should try to bring everyone to included here for practical reasons. a common level of ignorance. Any review on phytoplankton should include TREVOR PLATT the phototrophic bacteria. In spite of an increasing interest in this category (see e.g. Fujita and Zenitani The above quotation is taken from a letter that the 1975; Sieburth 1979), bacteria will be omitted here Chairman sent to the lecturers during the preparation on "practical grounds." These two words emphasize of this seminar. It reminds me of the philosophy of that we consider all the phototrophic species that cardinal Nicholas of Cusa (1401-64), Looking for are amenable to routine planktological methods. (In truth is tantamount to appraising ignorance, says a this perspective, one flaw is that the smaller Cyano- major theme of his Docta ignorantia . In the following phyceae may escape such methods; but we need a pages, I will strictly obey the principles of Nicholas perspective nevertheless.) and Trevor. The data under examination will be used, Some algal classes in marine plankton are merely hopefully, to fill the gaps in our knowledge. I would incidenal, whereas others predominate in numbers or like also to regain the ingenuousness of the child who biomass. However, all of them will be taken into asks about whatever he may come across: "What is this account with equal interest. Thirteen algal classes are used for?" since the usefulness (in biological terms: of concern here. Their main morphological and cyto- the ecological, physiological, or evolutionary signifi- logical characteristics are summarized in Table 1. cance) of the morphological features of phytoplankton Because the taxonomical literature is extremely cells, frankly speaking, remains practically unknown. profuse and scattered, only a few recent books and At present, phytoplanktologists are divided into reviews are referred to: Dodge (1973, 1979), Stewart several races, or at least they behave as if they were. (1974), Werner (1977), and Sieburth (1979). Although excellent works are produced in one field or another, it would seem that microscopists work SHAPES AND SIZES on morpho-cells, ecologists on eco-cells, physio- logists on physio-cells , and so on. However, in so The sizes of marine phytoplankters range from far as our common aim is to understand the nature a few micrometres (or even 1 ,c(m) to a few milli- of phytoplankton better, the different races of workers metres; hence, there is a range of three orders of should not only treat each other with indulgence magnitude for the average linear dimension on a in a spirit of peaceful coexistence, but they desper- worldwide basis. If cell volumes are compared, the ately need to cooperate. Don't we, all of us, work extreme cases known may differ by a factor of one on the same cells? million or more. Practically, in a given area or study, cell volumes will commonly encompass a span of 4-5 orders of magnitude. Morphological and Cytological Features As a consequence, a number of classifications into size groups have been proposed. They are too According to the current classifications (there numerous and too divergent to be reviewed here. are several of them), algae are separated into 12-16 The more commonly accepted separation is made groups, usually classes. It may then be said that between "microplankton" and "nannoplankton" (or virtually all of them are present in marine plankton, nanoplankton) but the criterion for this varies largely providing that the following restrictions are made: according to authors (the 20-,um size is recommended — Charophycean algae (either as a class or a subclass) as the limit between the two groups). are totally absent of the marine plankton; Algologically speaking, marine plankters may — Phaeophyceae can be encountered there only as be one of the following: (1) unicells, either flagel- spores, since the vegetative forms are benthic macro- lated or not; (2) colonies, usually nonmotile as far phytes. Thus, they will be omitted here; as marine forms are concerned; (3) seldom, filaments; — Cyanophyceae are present, but many authors (4) exceptionally, coenocytic or thallic.

339 Morphodynamically, , the same organisms may last groups, however, are quantitatively unimportant. be ascribed to various geometrical types. The ones Note that flagellated classes may include nonflagel- proposed by Schütt (1892) and Gran (1912) are clas- lated genera, whereas nonflagellated classes may sical but should not be thought of as obsolete, as exhibit flagellated spores and gametes. they represent truly unique features, not to be found There may be up to eight flagella on a cell. in any environment other than planktonic. These are They are equal or subequal or quite different in struc- (1) the bladder type, large, thin-walled cells, possibly ture and function. They may be anchored in a de- including a large vacuole; (2) the ribbon type, flat- pression or gullet of the cell, or not; be partly or tened, thin-walled cells; (3) the hair type, cells entirely lodged in a groove, or completely free; be elongated and/or grouped into elongated chains; inserted apically or laterally; beat forward (thus (4) the branching type, with spiny or lamellar exten- pulling the cell) or backward (pushing it). sions. A fifth, mucous type, was implicitly added Usually, flagella bear secondary branchings (with mucilage sheaths or filaments), which is rather called mastigonemes (or simply, hairs) of various rare in the sea compared with fresh waters. Although kinds. These are inserted in one or two rows or in a such types are obviously arbitrary and somewhat spiral manner, and an apical tuft may be present. The restrictive, their respective values, on adaptative or flagella of some classes or genera bear tiny scales, competitive grounds, have never been assessed either organic or mineralized. In some cases, the experimentally. main axis of the flagellum, or axonema, is lined by Morphodynamically again, one may recognize a sheath, a rod, or a striated strand. several levels of organization in the occupation of Although the internal structure of all flagella space. The following scheme was derived by Lewis presents some univèrsal features, such as the organi- (1976) from the study of a tropical lake; it would zation of the axonema into 9 doublets plus 2 fibrils, merit examination by oceanographers as well. Four the other components of the flagellar apparatus, such levels are distinguished; "(i) primary structure, as the "roots," seem to be highly variable, but they determined by the shape and size of cells comprising have not been studied extensively yet. the biomass unit; (ii) secondary structure, determined The haptonema is a very special kind of fla- by the arrangement of cells with respect to each gellum, only present in the Prymnesiophyceae, in- other as a result of physical connections between serted between the two normal flagella. It is con- them; (iii) tertiary structure, resulting from the tractile or of varying length, has no roots, and its coiling, twisting, or bending of multicellular units; ultrastructure departs from the "9 + 2" model. and (iv) quaternary structure, arising from the com- As far as marine plankton is concerned, the bination of similar multicellular subunits. Less than behavioral and ecological significance of none of half of the phytoplankton species show only primary the above features has ever been investigated (except structure, half or more show secondary structure, the observation that haptonema may serve for attach- and only a few show structure at all levels." ment). It may not be of much use to dwell on the The swimming speed of a relatively minute diversity of shapes among phytoplankters. On one number of species has been measured. This can be hand, every marine biologist is more or less familiar done directly under the microscope or indirectly from with it. On the other hand, no one has ever attempted cell counts along a vertical profile at sea, units being to evaluate its significance. On the contrary, most gm • s-1 or m • cr., respectively (see, for instance, planktologists in their everyday labor tend to cancel Throndsen (1973) and Eppley et al. (1968) for each out the morphological variability, cell counts being respective case). The data are much too scarce to expressed as numbers of individuals or converted answer questions such as which class is the fastest? or into biochemical units; even taxonomical diversity which flagellar type is best?, but an order of magnitude (when this painful task is carried out) does not tell can be set forth. This would be about 150 ,um • s or us anything about morphological diversity because the 15 m • d- ' . species recognized may be quite different (as two species of different genera often are) as well as closely CELL WALLS similar (as congeneric species usually are). Only a very few species of marine phytoplankton Some aspects of the geometry of the cells will may properly be called "naked," in the sense that be discussed later. their cell is not limited by anything other than the FLAGELLA cellular membrane (or plasmalemma). This happens in the chlorophyceanDunaliella , in two minor classes, Most algal classes of marine plankton are flagel- and in various spores and gametes. In all other cases, lated, the exceptions being Bacillariophyceae (dia- the term "naked" is just erroneous. toms), Cyanophyceae, Eustigmatophyceae, Rhodo- Thus, the great majority of organisms are in- phyceae, and Xanthophyceae. The three or four cluded in some kind of wall, or covering, or skeleton 340 TABLE I. Morphological and cytological charactetistics of the various taxonomical classes of marine planktonic algae. (Note: When

Chloromonadophyceae Bacillariophyceae = Rhaphidophyceae Chlorophyceae Chrysop

Cell shape Boxlike -± prominences, Rounded" or flattened Highly variable; marine Highly fairly geometrical plankters often rounded rounded

Grouping of cells Colonies frequent in Solitary All algal types; marine Mostly plankton plankters solitary colony, marine I

Size From a few !£m to 2 mm Rather small All algal sizes; marine Usually (nannoplankton) plankters small (nannop (nannoplankton)

Flagella None Two flagella (one forward 0, 2, 4 or 8; all apparently Usually (gametes may be flagellated) and hair-bearing, the other smooth hair-bea backward and smooth), or and smc no flagella sheath; ■

Cell wall and "Frustule," a siliceous None None; or a cellulosic wall None; o skeleton external skeleton made of "lorica,' two "valves" and a "girdle" exceptic spines, skeleton

Chloroplasts Highly variable in shape and Numerous, in an outer Usually one, cup-shaped One (bi arrangement; usually two region of the cytoplasm to 6 or more

Eyespot None None Comnion at some stages of Comma the life cycle

- Ejectile bodies None Trichocysts; mucocysts None Discob( (except]

Reproduction Division of the frustule into Siliceot (where relevant its 2 halves; auxospores; sexualit to marine statospores; plankton) diatoms are diploid

Additional Several Golgi bodies Peculiar ring of Golgi bodies Mucout characters around the nucleus; vacuole contractile vacuoles (pseudc heterott

World All habitats, often abundant A small class Mostly fresh water Mostly abundance

Abundance in Very common, frequently Minor although locally Minor; higher in Not vet marine plankton dominant dominant in some cases brackish waters ever dc ristics of the marine planktonic representatives differ from those of the class as a whole, an effort has been made to point it out. Photosynthetic pigments and biochemical characteristics are not dealt with.)

Prymnesiophyceae Cryptophyceae Cyanophyceae Dinophyceae Euglenophyceae Eustigmatophyceae Prasinophyceae = Haptophyceae Rhodophyceae Xanthophyceae — Ovoid and flattened with an Variable Variable; many forms with Variable; usually with an Variable Variable, often rounded Variable Highly variable Highly variable antero-lateral gullet a conspicuous equatorial helical or bilateral symme- with an anterior or lateral groove try, depressed anteriorly gullet

or coenocytic or ts; Mostly solitary Solitary or colonial or Mostly solitary Solitary Solitary (always?) Mostly solitary, plus some Mostly solitary; if not: Mostly filaments or thalli; Solitary marine ed filamentous filamentous or colonial colonial, filamentous marine plankters solitary filamentous; stages plankters solitary

Rather small Small (many of bacterial From 10 gm to 1 mm Small (nannoplankton) Small (nannoplankton) Usually small Usually small All algal sizes; marine Usually small; marine (nannoplankton) size) but filaments up to (nannoplankton) (nannoplankton) plankters small plankters small 1 mm (nannoplankton) (nannoplankton)

with 2, and 2, subequal, hair-bearing, None Usually 2 (if not: 0), hair- None to several; usually 2, None (but flagellated 0 to 6, usually 4, equal, Usually 2, perpendicular, None Motile stages unequal flagella lateral inserted in an antero-lateral bearing; usually one unequal, inserted in an spores) inserted in an anterior pit; smooth (unless exceptions?); and gullet equatorial and undulating, anterior reservoir; bear tiny scales in two layers ' plus an haptonema ClIum the other backward hair-bearing in between, sometimes absent

Cs; or "cuticle" = "periplast," A thin though multilayered, "Amphiesma," a complex, "Pellicle," including spiral ? Organic scales and/or Organic scales None; or a cellulosic A 2-layered cellulosic •Inor; including an intemal layer pectidic envelope; 3- or 4-layered wall proteinous strips "lorica," a cellulosic I- calcareous scales envelope or other material envelope, silicified or not; of very small (proteinous?) -± external sheath including or not cellulosic armor (coccoliths) exceptionally: long spines plates plates r up Usually 2; if not, 1; with a None; thylalcoids scattered Variable Variable Usually 1, parietal Usually 1, cup-shaped 1 (bilobed) or 2 or more 1 to several Variable, usually several peculiar ultrastructure in a chromoplasmic region

Not common None Not common; .simple to Common Anterior, not connected Common Exceptional None? Common highly structured with the chloroplast

Ejectosomes, common None Trichocysts and mucocysts, Trichocysts and mucocysts, None? trichocysts None None None common; nematocysts, rare rare?

Sexuality unknown although Cell division of several Peculiar zoospores • Some cases of polymorphic Polymorphic cycles Motile and nonmotile stages genetic recombination types; kysts; a few peculiar cycles, including a common, including stages may altemate in the occurs; spores are unknown cycles; dinoflagellates are nonmotile stage of different morphology plankton; siliceous kysts in marine plankton haploid and/or behaviour; non- motile stages are planktonic or benthic ile "Corps de Maupas," unique Prokaryotic cells! Mesokaryotic nucleus! Peculiar nucleus; Large, crystallike pyrenoid Peculiar Golgi cytoplasmic organelles ±- gaseous pseudovacuoles peculiar mitosis; peculiar mitosis; heterocysts "pusules," permanent contractile vacuoles pulsatile vacuoles

A rather small class All environments Common in many types of Mosdy fresh waters A small class, mostly Not well defined Mosdy marine (or brackish) A dominant class among Mostly fresh water waters fresh water marine macrophytes

Usually negligible y if Rather minor although Local blooms apart, Very common, sometimes Not common, locally Usually negligible Frequent although rarely Common, sometimes Exceptional rather common controversial: negligible dominant abundant dominant dominant or neglected? (none of these expressions has been clearly defined In the few cases that have been studied, the and no attempt will be made here). The different frustule has been shown to be covered by a two- or types of "cell walls" (in the broad sense) can be multi-layered coating, part of which may serve silica briefly described, recalling that the position of the deposition and part of which may control cellular wall with respect to the plasmalemma is controversial exchanges (Hecky et al. 1973). On the other hand, in some classes; thus it cannot be said firmly whether a few, but taxonomically unrelated species can secrete these classes have an external or an internal skeleton. polysaccharidic microfibrils , hardly visible by elec- Periplast, cuticle, and pellicle — The first two tron microscopy, and this has not been investigated terms are generally synonymous and apply mostly to extensively. Should these features prove to be of the Cryptophyceae. Within this class, the periplast general occurrence, the skeleton of diatoms could consists of very small plates, probably proteinaceous , no longer be reasonably called "external." with a larger system of grooves and ridges. Above it is Organic and mineralized scales — Organic the plasmalemma, then an external "fuzz" of granular scales are common in the Prasinophyceae and the or fibrillar material. Prymnesiophyceae, rare among the Chrysophyceae, The pellicle, proper to the Euglenophyceae, is and exceptional in some Dinophyceae. They are made made of proteinaceous bands that encircle the cell of microfibrils arranged in regular and/or irregular (inside of the plasmalemma too) and give it a striated layers. They originate in the Golgi bodies and then appearance. migrate outside the plasmalemma. Two or more dif- Amphiesma — To avoid the ambiguity of the ferent types are commonly present on the same term "theca ," the cell wall of the Dinophyceae is better species. referred to as an amphiesma, from an old name revived In addition, a number of the Prymnesiophyceae, by Loeblich (1970). Even when apparently "naked," forming the Coccolithophorids, bear calcareous scales dinoflagellates bear a multilayered and vesicular wall or coccoliths. A coccolith is made of several elements made, from inside to outside, of (1) a continuous, in an organic matrix and is (always?) surrounded fibrous , noncellulosic layer, (2) a system of vesicles by an organic skin. and membranes, enclosing or not cellulosic plates that Siliceous scales are found in some members of are easily visible in the "armored" species , (3) an the Chrysophyceae. external membrane or the plasmalemma, depending Lorica — Some Prasinophyceae and Chryso- on the author. The plates, when present, are highly phyceae are included in an armor of agglomerated variable with respect to thickness , shape, size, num- grains or fibrils called the lorica. This skeleton is ber, and arrangement; they may or may not possess usually cellulosic and truly external. pores, spines , crests , ridges , and lamellae. Frustule —The skeleton of the Bacillario- phyceae is truly external, in the sense that all authors agree that the plasmalemma lies under it. It is also true CELL ORGANELLES that the new frustule of the dividing cell is formed in The number, shape, size, and arrangement of vesicles (that soon merge into a continuous " silica- chloroplasts in the cells of planktonic algae are highly lemma" ) situated beneath the cell membrane. Thus , variable. Some classes may show a more or less some questions remain about what is lost and what is typical pattern, such as the Prasinophyceae with newly formed when the new frustule resumes an their single, cup-shaped plastid, but most classes, external position. including diatoms and dinoflagellates , display a large The overall organization of the frustule consists range of types under the light microscope. As to the of the well-known boxlike structure (the familiar ultrastructure, most phytoplanktologists ignore it. lid and bottom called the "valves" and joined together Yet, why do most algae have three thylakoids per by a " girdle" ). No detail can be added here , due to lamella, Cryptophyceae only two, and Chlorophyceae the complexity of the subject and the amount of data and Prasinophyceae from one to six? Why do the available. Three remarks will be made, however. chloroplastic envelopes (they are made of 2 to 4 The "glass box" or "Petri dish" commonplaces membranes) and their connection with the endo- are best forgotten, as a significant part of the frustule is plasmic reticulum differ so widely among the classes? organic and its mineralized fraction is made of amor- Similarly, why are some pyrenoids embedded within phous , hydrated silica. the plastid and others stalked on it, and why do some The ultrastructure of the frustule, recently re- genera apparently lack them? vealed by the scanning electron microscope, does Marine phytoplankton offers a diversified choice not lead to a discouraging multiplicity of trifling of cases where the chloroplasts do not belong to details , but rather to a reasonable array of unitary the cell itself, but to an algal symbiont. The best- elements whose disposition and number, but not known examples are found in the dinoflagellates nature, differ widely among genera and species. Ornithocercus or Noctiluca and the ciliate Meso-

341 diniutn. Then, whereas routine counts usually include — Are colonies formed preferentially among small several nonphotosynthetic taxa (mainly dinoflagel- cells or small species, as suggested by Beklemishev lates), the same counts omit various protozoan species (1959)? which are ascribed to microzooplankton but, because — Are all the combinations of characters (e.g. mo- they harbor algal symbionts , should be regarded func- tility, elongation, gigantism, presence of trichocysts, tionally as phytoplankters. and so on) equally met with in nature? If not (as it A granular, carotenoid-containing organelle, seems), which are the preferred combinations, which called the stigma or eyespot, is common in the Chry- others are avoided, and why? sophyceae, Prasinophyceae, and Xanthophyceae, while rare or absent elsewhere. It is either included in the chloroplast or not, and associated with the flagella or not. Its function has presumably . . . Cell Size as an Eco-Physiological Factor something to do with light. Ejectile bodies of various kinds (trichocysts, mucocysts, and others) are known in about one-half of the marine algal groups. Their CELL VOLUME function is also hypothetical. It is surprising that It seems to be a biological law that smaller both series of organelles are generally considered organisms are metabolically more active than larger as animal characters, although so many algae possess ones. In other words, metabolic rates per unit volume them. (One may add motility as another alleged or unit weight decrease with increasing sizes of the "animality" of many phytoplankters.) organisms. In the field of unicellular algae, this is A large vacuole often fills much of the cell substantiated in the following ways: in the centric diatoms. Coccolithophorids seem to contain a large vacuole too. Nothing general can be With increasing size said about the other groups. Contractile vacuoles with an obvious osmoregulatory function are probably Growth rate Decreases common among the Chrysophyceae and rare in two Respiration per unit weight Decreases other classes. A peculiar, permanent osmoregulatory Photosynthesis per unit weight, vacuole, termed the pusule, is specific to the Dino- assimilation number Decrease phyceae where it encompasses Half-saturation constants of a large range of struc- nutrient assimilation Increase tures. Digestive vacuoles are reported in the Dino- (Sinking velocity) (Increases) phyceae and the Prasinophyceae. Gas vacuoles, better to be called pseudovacuoles, are restricted to a few These are general trends that may suffer a few cyanophycean genera, e.g. Oscillatoria (= Tri- exceptions; sinking rate is added here for conven- chodesmium) which they may render positively ience. buoyant. Then, considering that the first three rates all Endosymbiotic bacteria may be common in decrease with increasing size, one may wonder: how dinoflagellate cells and should be looked for in the do they decrease with respect to each other? For other groups. Viruses are but seldom mentioned. instance, how does the ratio "net production:gross production" behave? This remains controversial. Laws (1975) reviewed the available data and con- cluded that respiration drops in such a way that large MORPHOLOGICAL AUTO-CORRELATIONS cells gain advantage over the smaller, but Banse The functional significance of the above fea- (1976), on the account of the same data, reached tures, if we may repeat it, remains largely unknown. the opposite conclusion that "growth efficiency" is A possible way to understanding their signifi- not size dependent. cance would be to investigate the internal correlations As regards physical factors, a large amount of that may exist between such features. To this end, work in the past has been done to correlate cell size intraspecific as well as interspecific and intergeneric with temperature, either on specific, interspecific, variability would be equally relevant. Here are a few or intergeneric bases. The data are so conflicting examples of the questions to be answered: that the hope of finding any simple, universal rela- — Why does the valve of the centric diatoms, in a tionship has been more or less abandoned, the same given species, seem to flatten with increasing valve being true for salinity. This just means that things diameter (personal and subjective observations on are more intricate. Rhizosolenia and Corethron spp.)? Therefore, recent approaches aim to correlate — Do spiny or horny cells tend to be also the larger the size of phytoplankton cells at sea with a number or the smaller cells, and to be thin- or thick-walled of parameters simultaneously. What may be called cells? the "soviet school" (e.g. Semina 1972; Semina et al.

342 1976) faces the problem of cell size as a global, same is true for photosynthetic rates, in the lab average, and adaptative response of phytoplankton (Taguchi 1976) as well as in the sea (Smayda 1965). to its environment. In this way, cell size in the Pacific As for nutrients, one laboratory work at least showed Ocean was shown to be (a) negatively correlated that S/V is increased in deficient media (Harrison with the velocity of vertical water transport, (b) et al. 1977), substantiating the opinion that oligo- positively correlated with the density gradient in the trophic seas favor large S/V (which improve nutrient main pycnocline, and (c) more or less dependent assimilation), that is to say, smaller cells. on nutrient concentrations. Following a different According to a study (already mentioned) by philosophy, Parsons and Takahashi (1973, 1974) Lewis (1976) on a tropical lake, natural populations focus attention on the competition of phytoplankters of phytoplankton keep the range of S/V values which inhabit the same ecological "niche." They within limits that are much narrower than they would developed an equation in which the growth rate of a be if random combinations of sizes and shapes were given cell or species is a function of the six following all "permitted." Although this kind of paradox has factors: (1) incident light, (2) extinction coefficient, not been identified rigorously in marine plankton, (3) depth of the mixed layer, (4) upwelling velocity, we already have some indications: suffice it to ex- (5) rate of nutrient input to the cell, and (6) sinking amine a taxonomical list where both sizes and S /V rate of the cell. Then, comparing the growth rates are reported, for example that of Smayda (1965) for of a small coccolithophorid and a large diatom whose the diatoms of the Gulf of Panama, then it becomes characteristics were already available, they conclude evident, although intuitively, that S/V values vary that "only in a region of high light intensity and high considerably less than random values of the linear nutrient concentration it is possible for the larger dimensions would allow them to do. phytoplankter to grow faster than the small phyto- plankter." As the authors note, this would beautifully explain why nannoplankton quantitatively overwhelm microplankton in most seas of the world. A shrewd Cell Morphology and the Planktonic and fruitful criticism of this model was given by Way of Life Becky and Kilham (1974).

FLOTATION AND SINKING THE SURFACE: VOLUME RATIO According to a myth which originates in the The surface area of a cell is of obvious interest as early times of planktology (V. Hensen, F. Schutt, it permits and limits the exchange of energy (light W. Ostwald, and others) and that still persists in absorption, thermic balance) and matter (osmotic many of the recent textbooks, the morphological equilibration, nutrient assimilation). This is why cell peculiarities of phytoplankton are efficient "flotation surface is a useful index, or the best one in several mechanisms." Because phytoplankton is observed to cases, of phytoplankton activity (Paasche 1960; live in the upper layer of the ocean, it was and still Smayda 1965). It is worth remembering here that is blindly assumed that its curious shapes cannot aim Rubner's "surface rule" (as cited in Bertalanffy 1951), at anything else than ensuring the best flotation as according to which "the metabolic rate per unit weight possible! Breaking off with this obsession, Munk decreases with increasing size, but is constant per and Riley (1952) pointed out that phytoplankton unit surface," has not been evaluated in the case of has not only to keep itself in the top of the ocean, unicellular algae. but also to absorb nutrients and avoid grazing. They Then, a ratio such as "surface area of the cell: made bold to say, and were able to demonstrate, cell volume" (hereafter S/V) is of obvious interest that sinking aids the absorption of nutrients. Some too, as it relates the exchange potential to the biomass years later, a freshwater planktologist wrote: "So present. Two geometrical considerations are essential far we have had as a background the view that perfect in this connection. flotation or a density equal to that of the surrounding 1) For any given shape of cell , S/V will de- medium is a desirable property. Yet it is quite likely crease with increasing size. that it is not the case under most conditions" (Lund 2) At any given size, a spherical shape provides 1959). These new views were then extended by the smallest S/V. Smayda (1970) in a thorough review, the title of Because S/V and size are inversely correlated, which duly associated the two words "suspension" one would expect the growth rate and S/V to be and " sinking ." positively correlated. In effect, this is known from Before examining the various components in- laboratory cultures (e.g. Eppley and Sloan 1966) volved in flotation and sinking, let us remind our- and has been observed at sea (Margalef 1957). The selves that a phytoplankter may behave in three

343 different ways with respect to depth. It may be either involved in flotation to some extent, is an energy- neutrally buoyant (this would be rather common requiring process. according to Eppley et al. 1968), or swimming (if One of the merits of Smayda's review (1970) motile), or sinking down. In the latter case, how has been to emphasize the effects of the physical fast does it sink? A rough average of the available movements of water on the suspension of phyto- data would be 1 m • d', which is rather slow (and plankters. Such movements include turbulence, slower by one order of magnitude than the swimming wind-induced patterns, thermic convection, and velocity, as mentioned in the first section). various water transports. According to Smayda, their Let y stand for the sinking velocity. velocity exceeds y by 2 or 3 orders of magnitude. All things being equal (which never holds true Thanks to this physical aid, plankters would be al- in nature), small cells sink slower. As to the overall lowed to explore the vertical dimension and benefit shape (disc, ribbon, and so on), things are not simple temporarily from the deeper and more nutritive layers . as each shape has its own size/y function. The Morphological features, instead of assisting flota- incidence of cellular appendages (horns, setae, and tion, would thus ensure passive entrainment and so on) is still more complicated as any departure even sinking. Thus the etymology of "plankton" from the spherical or spheroidal shapes will simul- (7r. X cercTbs• = unstable, wandering) is better taneously: understood. In this modern approach as well as in — increase the drag resistance, then lower y; the older one, our ignorance remains complete as — increase the specific weight, then increase y; to the exact function of the morphological features — make cells more subjected to passive entrainment; and the functional differences, if any, among taxo- — modify the orientation of the sinking cell in an nomical classes, genera, and species. unknown way; in this connection, two sorts of effects More details on the problem of suspension can have to be distinguished: (1) the geometrical shape be found in the reviews by Hutchinson (1967), of the cell and the resulting projections, (2) the dis- Smayda (1970), and Walsby and Reynolds (1981). tribution of weight within the volume of the cell. Suffice it here to conclude that floating is not always The role of mucous threads or sheaths remains a requirement for phytoplankters and that, in all unclear or controversial because mucus simulta- cases, an interaction of counteracting factors is neously lessens the total weight and increases the involved. size. In spite of the common preconception, colonies NUTRIENT AND LIGHT REQUIREMENTS sinlc faster than unicells, and long colonies sink The dependence of the absorption of nutrients faster than short ones. The most obvious reason is on cell morphology has been established on theoret- that the total area of a colony is smaller than the ical (or, sometimes, subjective) grounds only. sum of the areas of the isolated cells. If there are good theoretical reasons to believe As to cellular content, it has long been suggested that small and motile cells are best adapted to ab- that lipidic inclusions (the so-called oil droplets) may sorption (Munk and Riley 1952), the other relations compensate for the excess gravity, but precise calcu- are rather hypothetical. For instance, the formation lations as well as experimental measurements make of colonies is thought by Beklemishev (1959) to be this possibility more and more doubtful (see e.g. unfavorable, due to the reduction of the absorbing Anderson and Sweeney 1977). On the other hand, surfaces, whereas Margalef (1978) points out that the ancient theory of a regulation by ionic exchange, colonies are more efficiently tossed about and washed after being severely criticized, has been revived by turbulence, hence an improved absorption. The recently (see for instance the last reference). sanie author estimates that mucilages hinder absorp- It is well known that the viscosity of seawater is tion and serve as a restraint upon excessive growth. inversely correlated with temperature. As a conse- In all respects, experimental data are desperately quence, y is accelerated by about 4% with a rise of 1°C needed, particularly with regard to the incidence and is approximately doubled when temperature rises of the various kinds of cell walls and skeletons. from 0 to 27°C. However, the viscosity of the medium Concerning light, the main cytological differ- differs from that in the immediate environment of the ences among phytoplankters may lie in their respec- cells, as suggested by Margalef (1957) and demon- tive pigmentary equipment, a matter which is beyond strated by Chase (1979). This should stimulate the present subject. Truly morphological adaptations experimental measurements on some selected species. to light have not been studied, except for the "shade A common observation in the laboratory is flora" question. The shade flora consist principally that planktonic algae sink faster when either light of several genera or species of dinoflagellates which or nutrients are deficient, and when cells are aging. occur preferentially at the bottom of the euphotic Reasons for this are not clear. As regards light, the layer, or even somewhat below. Unfortunately, it clue may be that active transport of ions, which is has not been possible so far to identify any morpho-

344 logical character that would be common to such on one hand, ecologists and physiologists on the species. For instance, if one considers the single other hand, the great amount of morphological data genus Ceratium, its half-dozen "shade species" are may well be considered as luxurious and superfluous either flattenned, or delicately elongated, or strongly knowledge. It will remain so until a bridge connects constructed! both sides of planktology, a bridge that may be called eco-physio-morphology, or simply, func- GRAZING PRESSURE tional morphology. Two series of questions may be asked. The Because the grazing of phytoplankton by herbiv- function of the orous zooplankton (and sometimes nekton directly) various geometrical and cytological features is to be defined, and the functional differ- is largely an active and selective process, grazing ences between the taxa are to be investigated. In pressure can be expected to modify the size distri- both ways, the simpler the question will be, the best bution and the taxonomical spectrum of planktonic it will be (for instance, what is an eyespot used for? algae in the sea. Munk and Riley (1952) again may and which flagellates swim faster?). be consulted for a theoretical approach, and ex- Hopefully, this will enlighten the classical perimental evidence may be found in Parsons et al. "plankton paradox" of G. E. Hutchinson. Are those (1967). categories that we call species functionally redundant, It has been frequently reported that the smaller or does each of them occupy a distinct micro-niche forms are less actively preyed upon than larger ones; in the ecosystem? All that can be said today about the same for long-horned organisms as compared with morphological diversity is, plagiarizing E. F. Schu- smooth ones, colonies as compared with unicells, and macher's views in economics, that diverse is beau- so on. Any generalization would be quite hazardous, tiful. This is not a scientific statement, however. however, and no algal category of any kind can be guaranteed against grazing in any case. Obviously, grazing involves reciprocal conditions of suitability between the morphology of the prey and the anatomy Acknowledgments of the grazer, the details of which depend on the specific food chain under consideration. The comments of Prof. P. Bourrelly and Dr M.-J. Again, the suitability of the different groups Chrétiennot-Dinet on Table 1 are gratefully acknowledged. of phytoplankton with respect to grazing remains Dr A. E. Walsby helped to improve both the style and to be evaluated. For instance, there are indications the substance of the manuscript. that trichocysts may be an efficient protection and that bioluminescent species of dinoflagellates dis- courage copepods. References'

THE ECOLOGICAL SUCCESSION ANDERSON, L. W. J., AND B. M. SWEENEY, 1977. According to the synthetic views of R. Mar- Diel changes in sedimentation characteristics of Ditylum brightwellii: changes in cellular lipid and galef on the ecology of phytoplankton, the morpho- effects of respiratory inhibitors and ion-transport logical, cytological, and physiological properties of modifiers. Limnol. Oceanogr. 22: 539-552. the planktonic algae become integral components BANSE, K. 1976. Rates of growth, respiration and photo- of the ecological succession (Margalef 1958, 1978). synthesis of unicellular algae as related to cell size — Thus, cells at the early stage of the succession a review. J. Phycol. 12: 135-140. are typically small and often rounded, have a high BEKLEMISHEV, C. W. 1959. Sur la colonialité des dia- surface: volume ratio, assimilate nutrients intensively, tomées planctoniques. Int. Rev. Gesamten Hydrobiol. and multiply rapidly. As one proceeds in the space- 44: 11-26 . BERTALANFFY, time continuum towards conditions of thermal strati- L. VON. 1951. Metabolic types and growth types. Am. Nat. 85: 111-117. fication and nutrient impoverishment, phytoplankters CHASE, R. R. P. 1979. Settling behavior of natural aquatic tend to exhibit a larger size, a lower S/V ratio, and particulates. Limnol. Oceanogr. 24: 417-426. a reduced rate of cellular exchanges, together with DODGE, J. D. 1973. The fine structure of algal cells. motility, long generation time, and morphological Academic Press, London and New York. 261 p. specialization. At the same time, the taxonomical 1979. The phytoflagellates: fine structure and diversity of the community is considerably increased. phylogeny, p. 7-57. In M. Levandowsky and S. H. Hutner [ed.] Biochemistry and physiology of protozoa. Some Concluding Remarks 'In this paper, the references have been deliberately kept In the present state of the communication of to a minimum. A more extensive bibliography will be given information between microscopists and taxonomists in a later review.

345 2nd ed. Vol. I. Academic Press, New York, London, alternatives in an unstable environment. Oceanol. Acta Toronto, Sydney, and San Francisco. 1: 493-509.

EPPLEY, R. W., O. HOLM - HANSEN, AND J. D. H. MUNK, W. H., AND G. A. RILEY. 1952. Absorption STRICKLAND. 1968. Some observations on the vertical of nutrients by aquatic plants. J. Mar. Res. 11: 215- migration of dinoflagellates. J. Phycol. 4: 333-340. 240. EPPLEY, R. W., AND P. R. SLOAN. 1966. Growth rates PAASCHE, E. 1960. On the relationship between primary of marine phytoplankton: correlation with light ab- production and standing crop of phytoplankton. J. sorption by cell chlorophyll a. Physiol. Plant. 19: Cons. Perm. Int. Explor. Mer 26: 33-48. 47-59. PARSONS, T. R., R. J. LEBRASSEUR, AND J. D. FULTON. FUJITA, Y., AND B. ZENITANI. 1975. Distribution of 1967. Some observations on the dependence of zoo- phototrophic bacteria in Omura Bay during the summer plankton grazing on the cell size and concentration with special reference to brown Chborobitim . J. of phytoplankton blooms. J. Oceanogr. Soc. Jpn. 23: Oceanogr. Soc. Jpn. 31: 124-130. 10-17. GRAN, H. 1912. Pelagic plant life, p. 307-386. In J. PARSONS, T. R., AND M. TAKAHASHI. 1973. Environ- Murray and J. Hjort [ed.] The depths of the ocean. mental control of phytoplankton cell size. Limnol. MacMillan, London. Oceanogr. 18: 511-515. HARRISON, P. J., H. L. CONWAY, R. W. HOLMES, AND 1974. A rebuttal to the comment by Hecky and C. O. DAVIS. 1977. Marine diatoms grown in chem- Kilham. Limnol. Oceanogr. 19: 366-368. ostats under silicate or ammonium limitation. III. SCHÜTT, F. 1892, 1893. Das Pflanzenleben der Hochsee. Cellular chemical composition and morphology of Kiel & Leipzig, 76 p. (1892); Ergebn. Plankton Chaetbceros debilis, Skeletonema costattan, and Exped. Humbold Stift. 1A: 243-314 (1893). Thalassiosira gravida. Mar. Biol. 43: 19-31. SEMINA, H. J. 1972. The size of phytoplankton cells in HECKY, R. E., AND P. KILHAM. 1974. Environmental the Pacific ocean. Int. Rev. Gesamten Hydrobiol. control of phytoplankton cell size. Limnol. Oceanogr. 57: 177-205. 19: 361-366. SEMINA, H. J., I. A. TARKHOVA, AND TRUONG HECKY, R. E., K. MOPPER, P. KILHAM, AND E. T. NGOC AN. 1976. Patterns of phytoplankton distribu- DEGENS. 1973. The aminoacid and sugar composition tion, cell size, species composition and abundance. of diatom cell-walls. Mar. Biol. 19: 323-331. Mar. Biol. 37: 389-395. HUTCHINSON, G. E. 1967. A treatise on limnology. Vol. SIEBURTH, J. McN. 1979. Sea microbes. Oxford Univ. II - Introduction to lake biology and the limno- Press, New York, NY. 491 p. plankton. John Wiley and Sons, New York, London, SMAYDA, T. J. 1965. A quantitative analysis of the phyto- Sydney. 1115 p. [Chap. 20. The hydromechanics plankton of the Gulf of Panama. II. On the relation- of the plankton, p. 245-305.1 ship between C14 assimilation and the diatom standing LAWS, E. A. 1975. The importance of respiration losses crop. Inter-Am. Trop. Tuna Comm. Bull. 9: 465- in controlling the size distribution of marine phyto- 531. plankton. Ecology 56: 419-426. 1970. The suspension and sinking of phyto- LEWIS, W. M. JR. 1976. Surface/volume ratio: implica- plankton in the sea. Oceanogr. Mar. Biol. A. Rev. tions for phytoplankton morphology. Science (Wash- 8: 353-414, 1 pl., I table. ington, D.C.) 192 (4242): 885-887. STEWART, W. D. P. [ed.]. 1974. Algal physiology and LOEBLICH, A. R. III. 1970. The amphiesma or dinoflagel- biochemistry. Blackwell Scient. Publ. Oxford, Lon- late cell covering. Proc. N. Am. Paleontol. Conven- don, Edinburgh, and Melbourne (Bot. Monogr. 10). tion, Chicago 1969. Part G: 867-929. 989 p. LUND, J. W. G. 1959. Buoyancy in relation to the ecology TAGUCHI, S. 1976. Relationship between photosynthesis of the freshwater phytoplankton, Br. Phycol. Bull. and cell size of marine diatoms. J. Phycol. 12: 185- 7: 1-17. 189. MARGALEF, R. 1957. Nuevos aspectos del problema de THRONDSEN, J. 1973. Motility in some marine nano- la suspension en los organismos planctônicos. plankton flagellates. Norw. J. Zool. 21: 193-200. Invest. Pesq. 7: 105-116. WALSBY, A. E., AND C. S. REYNOLDS. 1981. Sinking 1958. Temporal succession and spatial hetero- and floating. In I. Morris [ed.] Physiological ecology geneity in phytoplankton, p. 323-349. In A. A. of phytoplankton. Blackwell Scient. Publ., Oxford. Buzzati-Traverso [ed.] Perspectives in marine biology. (In press) Univ. California Press, Berkeley and Los Angeles; WERNER, D. [ed.]. 1977. The biology of diatoms. Black- Union internationale des sciences biologiques, Paris. well Scient. Publ., Oxford, London, Edinburgh, 1978. Life-forms of phytoplankton as survival and Melbourne (Bot. Monogr. 13). 498 p.

346 DATE DUE DATE DE RETOUR „tin 2 1985

SR Canadian Bulletin of Fishe- 223 ries and Aquatic Sciences. B8213 Bulletin 210 #210

DATE ISSUED TO 14n3A t..Dut R A V R 985 .oc-) - co F

••■••■ / • *k to cn N CO N W 2/ 73 N 0

W C, t o e rn I» CD Cji M H' P. • C N W 0 1 .0 0 , e M C) c-1- a 1-1• et- 0 0 H• O 0 H• t-ftà 11 0 0• H• CD C/1 • CD 0 1

<1 44z W W N 00 N N 0 I-. -4

W C) O P. A). 1•1=9

O 0., Pt) H. 0 •P. n. >> to oo z 0 in Î M• - 0 H. (1- C) H. O 0 o O 0 1-t • ■-f 0 • >XI T( O P. r rn • 0