sign in sign up
<<
Home , Aquatic science, Chemosynthesis, Limnology, Phytoplankton, Plankton, Primary production

METHODS 2277 area of research. The Rrst computation of global Falkowski PG and Woodhead AD (eds) (1992) Primary oceanic primary production using the remote- and Biogeochemical Cycles in the . sensing approach appeared in the literature in 1995 : Plenum Press. (Figure 1). Other, similar computations have Geider RJ and Osborne BA (1992). Algal . since appeared in the literature. It is a method that New York: Chapman & Hall. will continue to improve, with improvements in Li WKW and Maestrini SY (eds) (1993) Measurement of Primary Production from the Molecular to the Global technology as well as in the techniques for Scale, ICES Marine Science Symposia, vol. 197. extrapolation of local biological measurements to Copenhagen: International Council for the Exploration large scales. of the Sea. Longhurst A (1998) Ecological of the Sea. San Diego: Academic Press. See also Longhurst A, Sathyendranath S, Platt T and Caverhill C (1995) An estimate of global primary production in Microbial Loops. Network Analysis of Food Webs. the from satellite radiometer data. Journal of . Pelagic . Research 17: 1245}1271. Primary Production Processes. Primary Produc- Mann KH and Lazier JRN (1991) Dynamics of Marine tion Methods. Ecosystems. Biological}Physical Interactions in the . Cambridge, USA: Blackwell Science. Platt T and Sathyendranath S (1993) Estimators of pri- Further Reading mary production for interpretation of remotely sensed data on . Journal of Geophysical Research Chisholm SW and Morel FMM (eds) (1991) What Con- 98: 14561}14576. trols Production in -Rich Areas Platt T, Harrison WG, Lewis MR et al. (1989) Biological of the Open Sea? vol. 36. Lawrence, KS: American production of the oceans: the case for a consensus. Society of and . Marine Progress Series 52: 77}88.

PRIMARY PRODUCTION METHODS

J. J. Cullen, Department of Oceanography, measurement programs. However, details of these Halifax, Canada patterns can depend on methodology, so it is

Copyright ^ 2001 Academic Press important to appreciate the uncertainties and built- in biases associated with different methods for doi:10.1006/rwos.2001.0203 measuring primary production.

Introduction De\nitions Primary production is the synthesis of organic ma- Primary production is centrally important to eco- terial from inorganic compounds, such as CO2 and logical processes and biogeochemical cycling in . The synthesis of organic from CO2 is marine systems. It is thus surprising, if not discon- R R commonly called carbon xation: CO2 is xed by certing, that (as discussed by Williams in 1993), both photosynthesis and . By far, there is no consensus on a deRnition of planktonic photosynthesis by phytoplankton accounts for most primary productivity, or its major components, net marine primary production. Carbon Rxation by and gross primary production. One major reason macroalgae, microphytobenthos, chemosynthetic for the problem is that descriptions of ecosystems microbes, and symbiotic associations can be locally require clear conceptual deRnitions for processes important. (e.g., net daily production of organic material by Only the measurement of marine planktonic phytoplankton), whereas the interpretation of primary production will be discussed here. These measurements requires precise operational deRni- measurements have been made for many decades tions, for example, net accumulation of radiolabeled using a variety of approaches. It has long been CO2 in particulate matter during a 24 h incubation. recognized that different methods yield different Conceptual and operational deRnitions can be rec- results, yet it is equally clear that the variability of onciled for particular approaches, but no one set of primary productivity, with depth, time of day, deRnitions is sufRciently general, yet detailed, to , and region, has been well described by most serve as a framework both for measuring planktonic 2278 PRIMARY PRODUCTION METHODS

primary production with a broad variety of methods ed as CH2O. in sea water is found and for interpreting the measurements in a range of in several chemical forms which exchange quickly R R scienti c contexts. It is nonetheless useful to de ne enough to be considered in aggregate as total CO2 three components of primary production that can be (TCO2). In principle, photosynthesis can be quanti- estimated from measurements in closed systems: Red by measuring any of three -dependent pro- cesses: (1) the increase in organic carbon; (2) the E Gross primary production (Pg ) is the rate of decrease of TCO2 ; or (3) the increase of O2. How- photosynthesis, not reduced for losses to excre- ever, growth of phytoplankton is not so simple: tion or to respiration in its various forms since phytoplankton are composed of , E Net primary production (Pn) is gross primary , nucleic acids, and other compounds production less losses to respiration by phyto- besides , both photosynthesis and plankton the assimilation of are required. Conse- E Net production (Pnc) is net primary quently, many chemical transformations are asso- production less losses to respiration by hetero- ciated with primary production, and eqn [1] does trophic and metazoans. not accurately describe the process of light-depen- dent growth. Other components of primary production, such as It is therefore useful to describe the growth of , regenerated production, and export phytoplankton (i.e., net primary production) with production, must be characterized to describe food- a more general reaction that describes how trans- web dynamics and biogeochemical cycling. As formations of carbon and depend on the pointed out by Platt and Sathyendranath in 1993, in source of nutrients (particularly ) and on any such analysis, great care must be taken to the chemical composition of phytoplankton. For reconcile the temporal and spatial scales of both growth on : the measurements and the processes they describe. \# # Marine primary production is commonly ex- 1.0NO3 5.7CO2 5.4H2O pressed as grams or moles of carbon Rxed per unit P(C H O N)#8.25O #1.0OH\ [2] volume, or pet unit area, of sea water per unit time. 5.7 9.8 2.3 2 The timescale of interest is generally 1 day or The idealized organic product, C H O N, rep- 1 year. Rates are characterized for the euphotic 5.7 9.8 2.3 resents the elemental composition of phytoplankton. zone, commonly deRned as extending to the depth Ammonium is more reduced than nitrate, so less of 1% of the surface level of photosynthetically water is required to satisfy the demand for reduc- active radiation (PAR: 400}700nm). This conve- tant: nient deRnition of euphotic depth (sometimes sim- pliRed further to three times the depth at which 1.0NH>#5.7CO #3.4H O a Secchi disk disappears) is a crude and often inac- 4 2 2 P # # > curate approximation of where gross primary pro- (C5.7H9.8 O2.3N) 6.25O2 1.0H [3] duction over 24 h matches losses to respiration and excretion by phytoplankton. Regardless, rates of The photosynthetic quotient (PQ; mol mol\1) is the R photosynthesis are generally insigni cant below the ratio of O2 evolved to inorganic C assimilated. It depth of 0.1% surface PAR. must be speciRed to convert increases of oxygen to the synthesis of organic carbon. For growth on ni- \1 Photosynthesis and Growth of trate as described by eqn [2], PQ is 1.45 mol mol ; with ammonium as the source of N, PQ is 1.10. The Phytoplankton photosynthetic quotient also reSects the end prod- Primary production is generally measured by quan- ucts of photosynthesis, the mixture of which varies tifying light-dependent synthesis of organic carbon according to environmental conditions and the spe- from CO2 or of O2 consistent with the cies composition of phytoplankton. For example, if simpliRed description of photosynthesis as the the synthesis of carbohydrate is favored, as can reaction: occur in high light or low nutrient conditions, PQ is lower because the reaction described in eqn [1] be- &8hJ comes more important. Uncertainty in PQ is often CO #2H O &P (CH O)#H O#O [1] 2 2 2 2 2 ignored. This can be justiRed when the synthesis of organic carbon is measured directly, but large errors Absorbed photons are signiRed by hl and the carbo- can be introduced when attempts are made to hydrates generated by photosynthesis are represent- infer carbon Rxation from the dynamics of oxygen. PRIMARY PRODUCTION METHODS 2279

Excretion of organic material would have a small sult from that for the light bottle. It is thus assumed inSuence on PQ and is not considered here. that respiration in the light equals that in the dark. As documented by Geider and Osborne in their Approaches 1992 monograph, this assumption does not gener- ally hold, so errors in estimation of the respiratory

Primary production can be estimated from chloro- component of Pg must be tolerated unless isotopi- phyll (from satellite color or in situ Suorescence) if cally labelled oxygen is used (see below). carbon uptake per unit of is known. Methods based on the direct measurement of Therefore, ‘global’ estimates of primary production oxygen are less sensitive than techniques using the depend on direct measurements by incubation. The isotopic tracer 14C. However, careful implementa- technical objectives are to obtain a representative tion of procedures using automated titration or sample of sea water, contain it so that no signiRcant pulsed oxygen electrodes can yield useful and exchange of materials occurs, and to measure light- reliable data, even from oligotrophic of the dependent changes in carbon or oxygen during incu- open ocean. Interpretation of results is complicated bations that simulate the . by containment effects common to all methods for Methods vary widely, and each approach involves direct measurement of primary production (see be- compromises between needs for logistical conveni- low). Also, a value for photosynthetic quotient must ence, precision, and the simulation of natural condi- be assumed in order to infer carbon Rxation from tions. Each program of measurement involves many oxygen production. Abiotic consumption of oxygen decisions, each of which has consequences for the through photochemical reactions with dissolved resulting measurements. Several options are listed in can also contribute to the measure- Tables 1 and 2 and discussed below. ment, primarily near the surface, where the effective wavelengths penetrate. Light-dependent Change in Dissolved Oxygen Light-dependent Change in Dissolved Inorganic The light-dark oxygen method is a standard ap- Carbon proach for measuring photosynthesis in aquatic sys- tems, and it was the principal method for measuring Changes in TCO2 during incubations of sea water marine primary production until it was supplanted can be measured by several methods. Uncertainties 14 by the C method, which is described below. Accu- related to biological effects on pH-alkalinity-TCO2 mulation of oxygen in a clear container (light bottle) relationships are avoided through the use of represents net production by the enclosed commun- coulometric titration or infrared gas analysis after ity, and the consumption of oxygen in a dark bottle acidiRcation. Measurement of gross primary pro- is a measure of respiration. Gross primary produc- duction and net production of the enclosed com- tion is estimated by subtracting the dark bottle re- munity is like that for the light-dark oxygen

Table 1 Measurements that can be related to primary production

Measurement Advantages Disadvantages Comments

Change in TCO2 Direct measure of net Relatively insensitive: small Not generally practical for inorganic C fixation change relative to large open-ocean work background Change in oxygen Direct measures of Small change relative to large Very useful if applied with

concentration (high O2 dynamics can yield background great care. precision titration) estimates of net and gross Interpretation of light-dark Requires knowledge of PQ production incubations is not simple to convert to C-fixation Incorporation of Very sensitive and relatively Tracer dynamics complicate The most commonly used 14C-bicarbonate into organic easy. interpretations method in oceanography material (radioactive Small volumes can be used Radioactive I requires special isotope) and many samples can be precautions and permission processed Incorporation of No problems with radioactivity Less sensitive and more work A common choice when 13C-bicarbonate into organic than 14C method 14C method is material (stable isotope) Larger volumes required impractical Measurement of Measures photosynthesis Requires special equipment A powerful research tool, 18 18 O2 production from H2 O without interference from not generally used for respiration routine measurements 2280 PRIMARY PRODUCTION METHODS method, but there is no need to assume a photo- The 18O Method synthetic quotient. However, precision of the Gross photosynthesis can be measured as the pro- analyses is not quite as good as for bulk oxygen duction of 18O-labeled O from water labeled with methods. Extra procedures, such as Rltration, would 2 this heavy isotope of oxygen (see eqn [1]). Detection be required to assess of calcium car- is carried out by mass spectrometry. Net primary bonate (e.g., by ) and photochemi- production of the enclosed community is measured cal production of CO . These processes cause 2 as the increase of oxygen in the light bottle, and changes in TCO that are not due to primary pro- 2 respiration is estimated by difference. In principle, duction. The TCO method is not used routinely for 2 the difference between gross production measured measurement of primary production in the ocean. with 18O and gross production from light-dark The 14C Method oxygen changes is due to light-dependent changes in respiration and photochemical consumption of Marine primary production is most commonly mea- oxygen. Respiration can also be measured directly sured by the 14C method, which was introduced by by tracking the production of H 18O from 18O . Steemann Nielsen in 1952. Samples are collected 2 2 The 18O method is sufRciently sensitive to yield and the dissolved inorganic carbon pool is labeled useful results even in oligotrophic waters. It is not with a known amount of radioactive 14C-bicarbon- commonly used, but when the measurements have ate. After incubation in clear containers, carbon been made and compared to other measures of pro- Rxation is quantiRed by liquid scintillation counting ductivity, important insights have been developed. to detect the appearance of 14C in organic form. Generally, organic carbon is collected as particles on a Rlter. Both dissolved and particulate organic Methodological Considerations carbon can be quantiRed by analyzing whole water after acidiRcation to purge the inorganic carbon. It Many choices are involved in the measurement of S is prudent to correct measurements for the amount primary production. Most in uence the results, of label incorporated during incubations in the dark. some more predictably than others. A brief review 14 of methodological choices, with an emphasis on the The C method can be very sensitive, and good 14 precision can be obtained through replication and C method, reveals that the measurement of pri- adequate time for scintillation counting. The mary production is not an exact science. method has drawbacks, however. Use of radioiso- Sampling topes requires special procedures for handling and disposal that can greatly complicate or preclude Every effort should be made to avoid contamination some Reld operations. Also, because 14C is added as of samples obtained for the measurement of primary dissolved inorganic carbon and gradually enters production. Concerns about toxic trace elements pools of particulate and dissolved matter, the are especially important in oceanic waters. Trace dynamics of the labeled carbon cannot accurately metal-clean procedures, including the use of represent all relevant transformations between specially cleaned GO-FLO sampling bottles sus- organic and inorganic carbon pools. For example, pended from Kevlar2+ line, prevent the toxic respiration cannot be quantiRed directly. The inter- contamination associated with other samplers, parti- pretation of 14C uptake (discussed below) is thus cularly those with neoprene closure mechanisms. Fre- anything but straightforward. quently, facilitates for trace metal-clean sampling are unavailable. Through careful choice of materials The 13C Method and procedures, it is possible to minimize toxic The 13C method is similar to the 14C method in that contamination, but enrichment with trace a carbon tracer is used. Bicarbonate enriched with nutrients such as is probably unavoidable. the stable isotope 13C is added to sea water and the Such enrichment could stimulate the photosynthesis incorporation of CO2 into particulate matter is fol- of phytoplankton, but only after several hours or lowed by measuring changes in the 13C:12C ratio of longer. particles relative to that in the TCO2 pool. Isotope Exposure of samples to during samp- ratios are measured by mass spectrometry or emis- ling can damage the phytoplankton and other sion spectrometry. Problems associated with radio- microbes, altering measured rates. Also, signiRcant isotopes are avoided, but the method can be more inhibition of photosynthesis can occur when deep cumbersome than the 14C method (e.g., larger samples acclimated to low irradiance are exposed volumes are generally needed) and some sensitivity to bright light, even for brief periods, during is lost. sampling. PRIMARY PRODUCTION METHODS 2281

Method of Incubation This system has many advantages, including im- proved security of samples compared with in situ Samples of can be incubated in situ, under deployment, convenient access to incubations for simulated in situ (SIS) conditions, or in incubators time-course measurements, and freedom of illuminated by lamps. Each method has advantages movement after sampling. Because the spectrally and disadvantages ( 2). neutral of by screens does not Incubation in situ ensures the best possible simu- mimic the ocean, signiRcant errors can be intro- lation of natural conditions at the depths of samp- duced for samples from the lower where ling. Ideally, samples are collected, prepared, and the percentage of surface PAR imposed by a screen deployed before dawn in a drifting array. Samples will not match the percentage of photosynthetically are retrieved and processed after dusk or before the utilizable radiation (PUR, spectrally weighted for next sunrise. If deployment or retrieval occur during photosynthetic absorption) at the sampling depth. daylight, deep samples can be exposed to unnatural- Incubators can be Rtted with colored Rlters to simu- ly high irradiance during transit, which can lead to late subsurface irradiance for particular water types. artifactually high photosynthesis and perhaps to Also, chillers can be used to match subsurface counteracting inhibitory damage. Incubation of , avoiding artifactual warming of deep samples in situ limits the number of stations that samples. can be visited during a survey, because the ship ArtiRcial incubators are used to measure photo- must stay near the station in order to retrieve the synthesis as a function of irradiance (P versus E). samples. Specialized systems both capture and Illumination is produced by lamps, and a variety of inoculate samples in situ, thereby avoiding some methods are used to provide a range of light levels logistical problems. to as many as 24 or more subsamples. Ship operations can be much more Sexible if pri- is controlled by a water bath. The duration of mary productivity is measured using SIS incuba- incubation generally ranges from about 20 min to tions. Water can be collected at any time of day and several hours, and results are Rtted statistically to incubated for 24 h on deck in transparent incubators a P versus E curve. If P versus E is determined for to measure daily rates. The incubators, or bottles in samples at two or more depths (to account for the incubators, are commonly screened with neutral physiological differences), results can be used to density Rlters (mesh or perforated metal screen) to describe photosynthesis in the as reproduce Rxed percentages of PAR at the surface. a function of irradiance. Such a calculation requires Light penetration at the station must be estimated to measurement of light penetration in the water and choose the sampling depths corresponding to these consideration of spectral differences between the in- light levels. Cooling comes from surface sea water. cubator and natural waters. Because many samples,

Table 2 Approaches for incubating samples for the measurement of primary production

Incubation system Advantages Disadvantages Comments

Incubation in situ Best simulation of the natural Limits mobility of the ship Not perfect, but a good field of light and temperature Vertical mixing is not simulated standard method if Artifacts possible if deployed or a station can be recovered in the light occupied all day Simulated in situ Many stations can be surveyed Special measures must be Commonly used when Easy to conduct time-courses taken to stimulate spectral many stations must be and experimental irradiance and temperature sampled. manipulations Vertical mixing not simulated Significant errors possible if incubated samples are exposed to unnatural irradiance and temperature Photosynthesis versus Data can be used to model Extra expenses and A powerful approach when irradiance (P versus E) photosynthesis in the water precautions are required applied with caution incubator (14C) column Spectral irradiance is not With care, vertical mixing can matched to be addressed Results depend on timescale of measurement Analysis can be tricky 2282 PRIMARY PRODUCTION METHODS usually of small volume, must be processed quickly, they transmit both visible and UV (280d400 nm) only the 14C method is appropriate for most P radiation. When the primary emphasis is an assess- versus E measurements in the ocean. ing effects of UV radiation, incubations are conduc- ted in polyethylene bags or in bottles made of Containers quartz or TeSon2+. Ideally, containers for the measurement of primary The size of the container is an important consid- production should be transparent to ultraviolet and eration. Small containers (450 ml) are needed when visible solar radiation, completely clean, and inert many samples must be processed (e.g. for P versus (Table 3). Years ago, soft glass bottles were used. E) or when not much water is available. However, Now it is recognized that they can contaminate small samples cannot represent the planktonic as- samples with trace elements and exclude naturally semblage accurately when large, rare or occurring ultraviolet radiation. Glass scintillation colonies are in the water. Smaller containers have vials are still used for some P versus E measure- greater surface-to-volume ratios, and thus small ments of short duration; checks for effects of con- samples have greater susceptibility to contamina- taminants are warranted. Compared with soft glass, tion. If it is practical, larger samples should be used -grade borosilicate glass bottles (e.g., for the measurement of primary production. The Pyrex2+) have better optical properties, excluding problems with large samples are mostly logistical. only UV-B (280d320 nm) radiation. Also, they con- More water, time, and materials are needed, more taminate less. Laboratory-grade glass bottles are radioactive waste is generated, and some measure- commonly used for oxygen measurements. Polycar- ments can be compromised if handling times are too bonate bottles are favored in many studies because long. they are relatively inexpensive, unbreakable, and Duration of Incubation can be cleaned meticulously. Polycarbonate absorbs UV-B and some UV-A (320d400 nm) radiation, so Conditions in containers differ from those in open near-surface inhibition of photosynthesis can be water, and the physiological and chemical differ- underestimated. The error can be signiRcant very ences between samples and nature increase as the close to the surface, but not when the entire water incubations proceed. Unnatural changes during column is considered. TeSon2+ bottles, more expen- incubation include: extra accumulation of phyto- sive than polycarbonate, are noncontaminating and plankton due to exclusion of grazers; enhanced inhi-

Table 3 Containers for incubations

Container Advantages Disadvantages Comments

Polycarbonate bottle Good for minimizing trace Excludes UV radiation Many advantages for element contamination Compressible, leading to gas routine and specialized Nearly unbreakable dissolution and filtration measurements at sea Affordable problems for deep samples Laboratory grade borosilicate More transparent to UV More trace element A reasonable choice if glass (e.g., PyrexTM) Incompressible contamination compromises are Breakable evaluated Borosilicate glass scintillation Inexpensive Contaminate samples with Can be used with caution vials Practical choice for P versus trace elements and Si for short-term P versus E Exclude UV radiation E measurements Polyethylene bag Inexpensive More difficult to handle Used for special projects, Compact Requires caution with respect e.g., effect of UV UV-transparent to contamination Quartz, TeflonTM UV-transparent Relatively expensive Used for work assessing TeflonTM does not effects of UV contaminate Small volume (1}25 ml) Good for P versus E Cannot sample large, rare Used for P versus E with Samples can be processed by phytoplankton evenly many replicates acidification (no filtration) Containment effects more likely Large volume (1}20 l) Some containment artifacts More work Required for some types of are minimized Longer filtration times with analysis, e.g., 13C Potential for time-course possible artifacts measurements PRIMARY PRODUCTION METHODS 2283 bition of photosynthesis in samples collected from through, especially when vacuum is applied. The mixed layers and incubated at near-surface irra- Rlters are also subject to clogging, leading to reten- diance; stimulation of growth due to contamination tion of small particles. with a limiting trace nutrient such as iron; and Labeled , including ex- poisoning of phytoplankton with a contaminant, creted photosynthate and cell contents released such as copper. When photosynthesis is measured through ‘sloppy feeding’ of grazers, is not collected with a tracer, the distribution of the tracer among on Rlters. These losses are generally several percent pools changes with time, depending on the rates of of total or less, but under some conditions, excre- photosynthesis, respiration, and grazing. All of these tion can be much more. When 14C samples are effects, except possibly toxicity, are minimized by processed with a more cumbersome acidiRcation restricting the time of incubation, so a succession of and bubbling technique, both dissolved and partic- short incubations, or P versus E measurements, can ulate organic carbon is measured. in principle yield more accurate data than a day- Interpretation of Carbon Uptake long incubation. This requires much effort, how- ever, and extrapolation of results to daily productiv- Because the labeled carbon is initially only in the ity is still uncertain. The routine use of dawn-to- inorganic pool, short incubations with 14C(41h) dusk or 24 h incubations may be subject to artifacts characterize something close to gross production. of containment, but it has the advantage of being As incubations proceed, cellular pools of organic much easier to standardize. carbon are labeled, and some 14C is respired. Also, some excreted 14C organic carbon is assimilated by Filtration or Acidi\cation heterotrophic microbes, and some of the phyto- Generally, an incubation with 14Cor13C is termin- plankton are consumed by grazers. So, with time, ated by Rltration. Labeled particles are collected on the measurement comes closer to an estimate of the a Rlter for subsequent analysis. Residual dissolved net primary production of the enclosed community inorganic carbon can be removed by careful rinsing (Table 4). However, many factors, including the with Rltered sea water; exposure of the Rlter to acid ratio of photosynthesis to respiration, inSuence the purges both dissolved inorganic carbon and precipi- degree to which 14C uptake resembles gross versus tated carbonate. The choice of Rlter can inSuence net production. Consequently, critical interpretation the result. Whatman GF/F glass-Rber Rlters, with of 14C primary production measurements requires nominal pore size 0.7 lm, are commonly used and reference to models of carbon Sow in the system. widely (although not universally) considered to cap- ture all sizes of phytoplankton quantitatively. Per- Conclusions forated Rlters with uniform pore sizes ranging from 0.2 to 5 lm or more can be used for size-fractiona- Primary production is not like temperature, tion. Particles larger than the pores can squeeze or the concentration of nitrate, which can in

Table 4 Incubation times for the measurement of primary production

Incubation time Advantages Disadvantages Comments

Short (41 h) Little time for unnatural Usually requires artificial illumination Closer to Pg physiological changes Uncertain extrapolation to daily rates in nature 1}6 h Convenient Uncertain extrapolation to daily rates Used for P versus E, especially Appropriate for some process in nature with larger samples studies Dawn}dusk Good for standardization of Limits the number of stations that can A good choice for standard methodology be sampled method using in situ Containment effects incubation

Vertical mixing is not simulated, Closer to Pnc near the surface;

leading to artifacts close to Pg deep in the photic zone 24 h Good for standardization of Results may vary depending on start A good standard for SIS methodology time. incubations.

Longer time for containment effects to Close to Pnc near the surface;

act closer to Pg deep in the photic zone 2284 PRIMARY PRODUCTION PROCESSES principle be measured exactly. It is a biological lites. Primary Production Distribution. Primary process that cannot proceed unaltered when phytop- Production Processes. Tracers of Ocean Produc- lankton are removed from their natural surround- tivity. ings. Artifacts are unavoidable, but many insults to the sampled plankton can be minimized through the Further Reading exercise of caution and skill. Still, the observed rates will be inSuenced by the methods chosen for mak- Geider RJ and Osborne BA (1992) Algal Photosynthesis: ing the measurements. Interpretation is also uncer- The Measurement of Algal Gas Exchange. New York: Chapman and Hall. tain: the 14C method is the standard operational Morris I (1981) Photosynthetic products, physiological technique for measuring marine primary produc- state, and phytoplankton growth. In: Platt T (ed.) tion, yet there are no generally applicable rules for Physiological Bases of Phytoplankton Ecology. Cana- 14 relating C measurements to either gross or net dian Bulletin of and 210: primary production. 83d102. Fortunately, uncertainties in the measurements Peterson BJ (1980) Aquatic primary productivity and the 14 and their interpretation, although signiRcant, are C}CO2 method: a history of the productivity prob- not large enough to mask important patterns of lem. Annual Review of Ecology and Systematics 11: primary productivity in nature. Years of data on 359}385. marine primary production have yielded informa- Platt T and Sathyendranath S (1993) Fundamental issues tion that has been centrally important to our under- in measurement of primary production. ICES Marine Science Symposium 197: 3}8. standing of marine ecology and biogeochemical Sakshaug E, Bricaud A, Dandonneau Y et al. (1997) cycling. Clearly, measurements of marine primary Parameters of photosynthesis: deRnitions, and production are useful and important for understand- interpretation of results. Journal of Plankton Research ing the ocean. It is nonetheless prudent to recognize 19: 1637}1670. that the measurements themselves require circum- Steemann Nielsen E (1963) Productivity, deRnition and spect interpretation. measurement. In: Hill MW (ed.) The Sea, vol. 1, pp. 129}164. New York: John Wiley. Williams PJL (1993a) Chemical and tracer methods of See also measuring plankton production. ICES Marine Science Symposium 197: 20}36. . Fluorometry for Biological Sensing. Williams PJL (1993b) On the deRnition of plankton pro- Ocean Carbon System, Modelling of. Network duction terms. ICES Marine Science Symposium 197: Analysis of Food Webs. Ocean Color from Satel- 9}19.

PRIMARY PRODUCTION PROCESSES

J. A. Raven, Biological Sciences, University of Primary producers are organisms that rely Dundee, Dundee, UK on external sources such as light energy Copyright ^ 2001 Academic Press (photolithotrophs) or inorganic chemical reactions (chemolithotrophs). These organisms are further doi:10.1006/rwos.2001.0202 characterized by obtaining their elemental require- ments from inorganic sources, e.g. carbon from Introduction inorganic carbon such as carbon dioxide and bicar- bonate, nitrogen from nitrate and ammonium (and, This article summarizes the information available for some, dinitrogen), and from inorganic on the magnitude of and the spatial and temporal phosphate. These organisms form the basis of food variations in, marine plankton primary productivity. webs, supporting all organisms at higher trophic The causes of these variations are discussed in levels. While chemolithotrophy may well have had a terms of the biological processes involved, the vital role in the origin and early evolution of , the organisms which bring them about, and the rela- role of chemolithotrophs in the present ocean is mi- tionships to oceanic and . The nor in energy and carbon terms (Table 1), but is very discussion begins with a deRnition of primary important in biogeochemical element cycling, for production. example in the conversion of ammonium to nitrate.