sign in sign up
<<
Home , Haptophyte, Prymnesiophyceae, Prymnesium, Prymnesium parvum

AQUATIC MICROBIAL ECOLOGY Vol. 14: 155-160, 1998 Published February 13 Aquat Microb Ecol

Phagotrophy by a plastidic , pa tellifer urn

Urban Tillmann*

Forschungs- und Technologiezentrum Westkiiste, Universitat Kiel, D-25761 Biisum, Germany

ABSTRACT:Observations and photographs of the previously unsuspected ingestion capacity of the phytoflagellate Prymnesiurn patelliferurnare presented. P. patelliferurntook up prey cells of different sizes by an elongated, U-shaped pseudopodium formed at the posterior part of the cells. While feeding, aggregates of several P. patelliferurncells were formed in which their come in close con- tact. Motile prey cells were immobilised or probably killed before ingestion, but not all protistan of ingestible size were immobilised or attacked. It is suggested that P. patelliferurnmay secrete a toxin that serves as a killing agent for prey organisms.

KEYWORDS: Prymnesiurn patelliferurn. Mixotrophy . Phagotrophy . Haptophyta

INTRODUCTION 1936, Dietrich & Hesse 1990), Bulgaria (Valkanov 1964) and Norwegian coastal waters (Kaardtvedt et Phagotrophy by photosynthetic is a well- al. 1991). Because of its economic importance as an known phenomenon described in the phycological lit- ichthyotoxin-producing genus, Prymnesjumis an often- erature as early as the late 19th century. This process cultivated and well-studied organism. Nevertheless, has recently been 'rediscovered' and its potential eco- for a long time, the only documentation of phagotrophy logical significance addressed as part of the rapid within this genus was an anecdotal description by evolution in microbial food web ecology. Many higher Conrad (1941), who described particle ingestion by algal taxa have species with both photosynthetic and pseudopodia in amoeboid cells of P. saltans Massart. phagotrophic nutrition (mixotrophy). This is most Little detail was given in that report and confirmatory common in the (Schnepf & Elbrachter observations were lacking (Green 1991). However, 1992), chrysophytes (Sanders & Porter 1988) and prym- Nygaard & Tobiesen (1993) recently showed that P. nesiophytes (Green 1991, Jones et al. 1994). In the parvum, among several other species of flagellates, latter group, descriptions of mixotrophic species are was able to ingest both in the field and in the nearly exclusively restricted to the genus Chrysochro- laboratory. In the present study, I show that particle mulina.In only 1 case has phagotrophy been recorded ingestion by Prymnesium is not restricted to bacteria in the coccolithophorids (Parke & Adams 1960). but that at least P. patelliferum is able to incorporate Prymnesium Conrad, the type genus of Prymnesio- different sized particles, sometimes even larger than phyceae, has a number of toxic bloom-forming species itself. separated morphologically only by the fine-structural details of body scales (Green et al. 1982) and is well known to cause extensive fish kills in brackish ponds MATERIALS AND METHODS in Israel (Reich & Aschner 1947), Denmark (Otterstrsm & Steemann Nielsen 1940), Germany (Lenz 1933, Hickel In 1992, Prymnesium patelliferumwas isolated from a brackish pond near Biisum, northern Germany. From this isolate, unialgal cultures were obtained by the 'E-mail: [email protected] dilution technique and were grown in natural seawater

0 Inter-Research 1998 156 Aquat Microb Ecol14: 155-160, 1998

(S = 28%) enriched with nutrients (Guillard & Ryther Table 1 List of species tested for their suitability as prey for 1962). Species determination, based on TEM examina- Pryrnnesium patelliferum tion of the scales, was carried out by J. Green (Ply- mouth Marine Laboratories, UK).Cultures were main- Species Diameter Ingestion tained at 20°C under natural light in the laboratory. To (pm) observed study phagotrophy, senescent cultures were mixed Unknown Chlorococcales 5-10 with different species of other (Table 1). Live DunalieUa sp. 7 observations were made using an inverted microscope Cryptornonas sp. 10 (Zeiss Axiovert 35) equipped with a Nikon photomicro- Skeletonema costaturn (Greville) Cleve 10 graphy system and documented with a Panasonic Thalassiosira rotula Meunier 3 0 Oxyrrhis manna Dujardin 3 0 video system. From video tapes, single frames were Gyrodinium sp. 45 printed by a Sony video-printer. Unknown 12

RESULTS pseudopodium (Fig. 1). Presumably, the formation of the pseudopodium is chemically stimulated and not In exponentially growing cultures, Prymnesium patel- triggered by direct contact with the prey, because it liferum is elongate and slender with 2 was often observed in cells not surrounded by algal covering the whole cell in the longitudinal axis. In food. If P. patelliferum with a pseudopodium en- dense and senescent cultures, the cell shape changes countered a food particle, the particle was completely to a more rounded form. The chloroplasts are smaller enclosed by the pseudopodium and taken up into a and the posterior part of the cell is often filled with a food located at the posterior of the cell within large body of reserve metabolite, probably chrysolam- about 1 min (Figs. 2 & 3). inarin. The cells are often filled with small oil droplets. In cultures of motile protists mixed with Prymnesium In addition, cells with a large brownish pellet in the patelliferurn, a varying number of potential prey cells posterior occurred and sometimes excretion of this lost their motility. This effect was not due to observation body by the cells was observed. These cellular fea- handling (e.g. microscopic illumination) as determined tures, with hindsight, possibly resulted from cannibal- by prey-alone control cultures under the microscope. ism, i.e. the brown bodies strongly resembled 'faecal For example, Fig. 4a shows the immobilised and pellets', suggesting the occurrence of phagotrophy. rounded form of an amoeba. This species can then eas- After mixing senescent Prymnesium patelliferum ily be ingested by P. patelliferum, as shown in Fig. 4b, c. cultures with other cultures, some P. patel- Single prey cells were often attacked by several liferurn cells showed a swollen and transparent poste- Prymnesium patelliferum cells simultaneously. Some- rior, which afterwards formed an elongated U-shaped times, when multiple P. patelliferurn encountered a single prey, the membranous pseudopodia of P. patelliferum cells appeared to merge to enclose the prey between them. This process is illustrated with a series of sequential images taken from video (Fig. 5). Resulting feeding aggregates of the 2-cell type are most common (Figs. 4c & 6a-e), but 3 or more P. patelliferurn cells may be involved (Fig. 6f). P. patelliferum also aggregated around relatively large prey. Both the heterotrophic (Fig. 7) and the phototrophic Gyro- dinium sp. (Fig. 8) were attacked. The pre- dator cells formed a more or less complete coat over the prey. In such cases, the shape of a single P. patelliferurn cell could be extremely deformed. Differences in suitability of protistan prey for Prymnesiurn patelliferurn were observed. Fig. 1. Prymnesium patelliferum forming elongated pseudopodia While the Chlorococcales and an amoeba (arrows) at the posterior part of the cell pnor to feedng activity. Scale were ingested (Figs. 2 to 4)e the slmilarl~sized bar = 10 p Cryptomonas sp. and Skeletonerna costaturn Tillmann: Phagotrophy by a plastidic haptophyte 157

Figs. 2 & 3. Time course of prey ingestion by Pryrnnesiurn patelliferurn. The drawings in the lower part give a schematic simpli- fied view of pseudopod action. Scale bars = 10 pm. Fig. 2. P. patelliferurn ingesting an unidentified chlorococcalean alga. Fig. P. patelliferurn ingesting Dunaljella sp.

Fig. 4. Prymnesium patelliferurn ingesting an amoeba. (a) After addition to a P. patemerum culture, the amoebae lost their motility and became rounded. (b) After ingestion, the amoeba can be seen in a large food vacuole. (c) Two cells in close connec- tion with the food between them. Scale bars = 10 pm 158 Aquat Microb Ecol 14: 155-160, 1998

U .- 2 - II 2 2 -2C 2 9 2; c 0 aJ 2 al W 2 5 9" -g 6 g L 2 U c 0) 0 8 '= Fig. 6. Pryrnnesium patelliferurn. Examples of communal phagotrophy by U U0) (a-e) pairs or (f)a triplet of P. patellderum cells feeding on (a,b) unknown , c c chlorococcalean or (c-f) DunalieUa sp. Scale bars = 10 pm aJ c 8 5 $ 5 0 were not captured (Table 1). Although suitable prey 'S v) : items were apparently immobilised, Cryptornonas sp. = motility was unaffected by the presence of l? patel- cE L g 2 liferum.

.-0 DISCUSSION &V)U W =W c u The mode of particle ingestion of Prymnesiumpatel- 2 CY uc liferum by means of pseudopodial development at aJ 0 5 5 the non-flagellar pole resembled that of Chrysochro- U n) mulina kappa Parke & Manton and C. ericina Parke h.4 -1 0 2 & Manton. Parke et al. (1955, 1956) described how a c 5 '30 V colourless, slightly granular substance seemed to " E, flow out from the non-flagellar pole of C. kappa and g U 52 enveloped the prey particles, which were then drawn aJ5 5 ;;; quickly inside the cell. I did not observe involvement of z Q the haptonema in the feeding process of P. patel- a! a liferum.This is in contrast to C, hirta Manton, in which $22 * the haptonema plays an important role in catching and tr U taking up particles (Kawachi et al. 1991) aJ"'U a, 0 c The observations from this study suggest that prey a S 5 g are immobilised or even killed by Prymnesium pa- .2 telliferurnbefore being ingested. Cells being ingested .2 were, in every case, immobilised and rounded. Al- 3- though the toxicity of the P. patelliferurn strain used Q, here has not been tested with standard toxicity tests, -0 the rapid immobilisation of prey suggests toxin pro- duction, especially since the toxicity of 2 strains of P. patelliferurn was demonstrated in a recent study by Larsen et al. (1993). Thus, it may be speculated that Tillmann: Phagotrophy by a plastidic haptophyte 159

Fig. 7. Prymnesiurn patelliferum attacking a cell of the heterotrophic dinoflagellate Oxyrrhis manna. Scale bars = 10 pm. (a) The dinoflagellate cell, coloured due to ingested Dunaliella cells, is completely rounded off. Nearby, P. patelliferum cells 'ready to feed' with the typical swollen hyaline posterior part (arrows) can be seen. (b) Several P. patelliferum cells have settled on the surface of a smaller individual 0. marina Prymnesium toxin may be used to kill potential prey Not all potential prey were ingested by Prymnesium organisms prior to ingestion. Toxins of are patelliferum. Cryptomonas sp. retained their motility in well known to affect organisms of their own size the presence of P. patelliferum and thus are probably , like amoeba, flagellates and (Valkanov insensitive to P. patelliferum toxin. The failure of 1964) and bacteria, and zooplankton P. patelliferum to ingest the 2 tested diatom species (Moestrup 1994). Estep & McIntyre (1989) hypothe- may be due to chain formation (Skeletonema costa- sised that toxin released by spp. turn), spines (Thalassiosira rotula) or in general to simply punches holes in the cell membrane of other surface properties of silica shells. Handling problems organisms. According to this hypothesis of auxotrophy seem unlikely, however, since P. patelliferum could following induced osmosis (dasmotrophy), the prey is ingest other species in the same size range of the not killed, but an outflow of nutrients is induced, which diatoms (Table 1).Field studies have demonstrated that Chrysochrornulina spp, can take up and absorb. The diatoms are not affected by either P. parvum (Otter- advantage to species producing toxins is still under strsm & Steemann Nielsen 1940) or Chrysochromulina discussion. leadbeateri Estep et al. (Johnsen & Lsmsland 1991).

Fig. 8. Pymnesium patelliferum attacking the autotrophic dinoflagellate Gyrodinium sp. Scale bars = 10 pm. (a) lmmediately after addition to the P. patelliferurn culture, the dinoflagellate loses its motility, but the cell shape is still recognisable. Nearby, P. patelliferum cells 'ready to feed' with the typical swollen hyaline posterior part (arrow) can be seen. (b) In a later stage, several P. pateUiferum cells form a dense layer around Gyrodinium sp. P. patelhferum cells are extremely deformed 160 Aquat Mlcrob Ecol 14: 155-160, 1998

When more than 1 Prymnesium patelliferum cell Guillard RRL, Ryther JH (1962) Studies of marine planktonic attacked a prey item, the membrane material of the diatoms. I. Cyclotella nana Hustedt and Detonula confer- predatory cells often appeared to fuse and form a com- vacea (Cleve) Gran. Can J Microbial 8:229-239 Hickel B (1976) Fischsterben in einem Karpientelch bei einer mon food vacuole (e.g. Fig. 6). Such feeding commu- Massenentwicklung des toxischen Phytoflagellaten Prym- nities with concomitant cell fusion and a subsequent nesium parvum Carter (Haptophyceae). Arch FischWiss formation of common food are known from 27(2):143-148 the freshwater heliozoan Actinophrys sol (Patterson & Johnsen TM, Lomsland ER (1991) Growth experiments with Chrysochromulina leadbeateri. In: Rey F (ed) The Chryso- Hausmann 1981). For P. patelliferum a cell fusion pro- chromulina leadbeateri bloom in Vesffjorden, North Nor- cess remains speculative, since it was not possible to way, May-June 1991. Fisken og Havet No. 3, Havforsk- identify a continuous membrane unambiguously in the ningsinstituttet, Bergen, p 85-88 light microscope. Refined cytological techniques are Jones HLJ, Leadbeater BSc, Green JC (1994) Mixotrophy in needed to clarify this problem. However, formation of a haptophytes. In: Green JC, Leadbeater BSc (eds) The haptophyte algae. The Systematics Association, Spec Vol fused common food vacuole will not be required in the No. 51. Clarendon Press, Oxford, p 247-263 case of pinocytosis of leaked dissolved nutrients or for Kaardtvedt S, Johnsen TM, Aksnes DL, Lie U (1991) Occur- phagocytosis of small cell particles which may eventu- rence of the toxic phytoflagellate ally be released after disruption of the prey's . and associated fish mortality in a Norwegian fjord system. Can J Fish Aquat Sci 48:2316-2323 It is unknown whether the formation of cell aggre- Kawachi M, Inouye I, Maeda 0, Chihara M (1991) The hapto- gates and feeding by Prymnesium patelliferum cells nema as a food-capturing device: observations on Chryso- occur under natural conditions. The current observa- chromulina hirta (). Phycologia 30(6): tions of phagotrophy by senescent-stage cells suggest 563-573 that the phenomenon may be most common towards Larsen A, Eikrem W, Paasche E (1993) Growth and toxlcity in Prymnesium patelliferum (Prymnesiophyceae) isolated the end of Prymnesium blooms. Further detailed micro- from Norwegian waters. Can J Bot 71:1357-1362 scopical inspection of natural Prymnesium blooms for Lenz F (1933) Untersuchungen zur Limnologie von Strand- uncommon cell aggregates will give more insight into seen. Verh Int Verein Theor Angew Limnol6:166-177 these interesting questions. Moestrup 0 (1994) Economic aspects: 'blooms', nuisance species, and toxins. In: Green JC, Leadbeater BSc (eds) The haptophyte algae. The Systematics Association, Spec Acknowledgements. This study was supported by a grant Vol No. 51. Clarendon Press, Oxford, p 265-285 from the German Ministry for Research and Technology Nygaard K, Tobiesen A (1993) Bacterivory in algae: a sur- (03F0130) within the framework of the TRANSWATT project. vival strategy during nutrient limitation. Lirnnol Oceanogr Thanks to Anette Mayer-Brinkmann for her excellent techni- 38(2):273-279 cal assistance and to John Green for species determination. Otterstrem CV, Steemann Nielsen E (1940) Two cases of 1 greatly appreciate valuable comments on the manuscript extensive mortality in fishes caused by the from John Green (Plymouth Marine Laboratories), Malte Prymnesium parvum. Rep Dan Biol Stn 44:l-24 Elbrachter (Biologische Anstalt Helgoland) and Franciscus Parke M, Adams I(1960) The motile (CrystalloLithus hyalinus Colijn (Forschungs- und Technologiezentrum Biisum). Gaarder & Markali) and non-motile phases in the history of pelagicus (Wallich) Schiller. J Mar Biol Assoc UK 39:263-274 LITERATURE CITED Parke M, mant tonI, Clarke B (1955)Studies on marine flagel- lates. 11. Three new species of Chrysochromulina. J Mar Conrad W (1941) Sur les Chrysomonadines a trois fouets. Biol Assoc UK 34:579-609 Aper~usynoptique. Bull Mus R d'Hist Nat Belgique 17:l-16 Parke M, Manton I,Clarke B (1956) Studies on marine flagel- Dietrich W, Hesse KJ (1990) Local in a pond at the lates. 111. Three further species of Chl-ysochromulina. German North Sea coast associated with a mass develop- J Mar Biol Assoc UK 35387-414 ment of Prymnesium sp. Meeresforsch 33:104-106 Patterson DJ, Hausmann K (1981) Feeding by Actinophrys Estep KW, Maclntyre F (1989) , life cycle, distri- sol (Protista, ): 1 Light microscopy. Microbios 31. but~onand dasrnotrophy of Chrysochromulma a theory 39-55 accounting for scales, haptonema, muciferous bodies and Reich K, Aschner M (1947) Mass development and control of toxicity. Mar Ecol Prog Ser 57:ll-21 the phytoflagellate Prymnesium parvum in fish ponds in Green JC (1991) Phagotrophy in prymnesiophyte flagellates. Palestine. Palest J Bot 4:14-23 In: Patterson DJ, Larson J (eds) The biology of free-living Sanders RW, Porter KC (1988) Phagotrophic phytoflagellates. heterotrophic flagellates. The Systematics Association. In: Marshall KC (ed) Advances in microbial ecology, Spec Vol No. 45. Clarendon Press, Oxford, p 401-414 Vol 10. Plenum, New York, p 167-192 Green JC. tiibberd DJ, Pienaar RN (1982) The taxonomy of Schnepf E, Elbrachter M (1992) Nutritional strategies in dino- Prymnesium (Prymnesiophyceae) lncludiny a description flage1,lates. Eur J Protistol 28:3-24 of a new cosmopolitan species, P. patellitera sp. nov., and Valkanov A (1964) Untersuchungen uber Prymnesium parvum further observation on P. parvum N. Carter. Br Phycol J Carter und seine toxische Eintvirkung auf Wasserorganis- 17:363-382 men. Kieler Meeresforsch 20:65-81

Editorial responsibility: Robert Sanders, Submitted: October 29, 1996; Accepted: October 22, 1997 Philadelphia, Pennsylvania. USA Proofs received from author(s): January 28, 1998