CYST FORMATION IN THE FRESHWATER DINOFLAGELLATE CERATIUM HIRUNDINELLA (DINOPHYCEAE)1

Cyst formation in Ceratium hirundinella (O. F. Müll.) Bergh was studied by light and electron microscopy, using material from several lakes and reservoirs and also laboratory cultures. Cells preparing to encyst build up large quantities of starch and lipid and at the same time reduce their other cell components. The cyst is released from the theca as a naked cell bounded by a double membrane. The most commonly found cyst deposits a layer of electron-dense granules containing silicon on the outer membrane and lays down a cellulose-like material between the two membranes. Cysts without the electron-dense granules are commonly formed in cultures but rarely found in lakes. These cysts appear less resistant to decay and do not show the reorganization of cell contents for dormancy. It is suggested that C. hirundinella has both a resting cyst, forming part of the life cycle, and a temporary cyst stage.     

COMPARATIVE ANALYSES OF THE DINOFLAGELLATE FLAGELLAR APPARATUS. II. CERATIUM HIRUNDINELLA1

The three-dimensional structure of the flagellar apparatus in the gonyaulacoid dinoflagellate. Ceratium hirundinella var. furcoïdes (Schröder) Hub.-Pest. was determined using serial section electron microscopy. The flagellar apparatus is quite large and consists of several components. The two basal bodies nearly abut at their proximal ends and are separated by an angle of approximately 120° The broad longitudinal microtubular root extends from the cell's left edge of the longitudinal basal body and bends around the sulcal/cingular depression into the cell's left antapical horn. A transverse striated fibrous root is associated with the transverse basal body and a narrow electron dense extension is present along the anterior edge of the transverse basal body. This study revealed severa1 hitherto unreported fibrous components of the flagellar apparatus that link the various microtubular and fibrous components to themselves and to the two striated collars. A large striated fibrous connective links the two striated collars to one another. This fibrous connective is linked to another striated fibrous connective that originates from the longitudinal basal body and lies perpendicular to the longitudinal microtubular root. The readily identifiable and numerous components of the Ceratium flagellar apparatus are comparable to those of other dinoflagellates. The combined presence of well dpveloped striated collars, a striated collar connective, and a basal body angle of approximately 120° indicates that this flagellar apparatus is most like that described for Peridinioid dinoflagellates. Important similarities are also noticeable between this flagellar apparatus and that of Oxyrrhis marina.     

ACTIVE GROWTH OF THE OCEANIC DINOFLAGELLATE CERATIUM TERES IN THE CARIBBEAN AND SARGASSO SEAS ESTIMATED BY CELL CYCLE ANALYSIS1

The growth rate of an oceanic dinoflagellate, Ceratium teres Kofoid, was investigated in the Sargasso and Caribbean Seas from September 1989 to July 1990 using the cell cycle analysis method. Estimated growth rates ranged from 0.29 to 0.58 day^?1 and were 1.5–7.2 times higher than generally accepted rates for oceanic dinoflagellates. The higher rates in this report were mainly due to an improvement in techniques that determine the duration of a terminal cell cycle phase in situ. The day-to-day variation in growth rates was surprisingly small, but, from long-term measurements, a weak correlation was found among temperature, daily irradiance, and seasonal growth rate. The calculated species-specific primary production ranged from 0.5 to 1.8 mg C·m^?2·day^?1, about 1% of the estimated total production. Ceratium teres may be an important carbon source at the base of the grazing food chain.     

BIOGEOGRAPHIC ANALYSIS OF THE ARMORED PLANKTONIC DINOFLAGELLATE CERATIUM IN THE NORTH ATLANTIC AND ADJACENT SEAS1

The distribution of the dinoflagellate genus Ceratium Schrank (Dinophyceae) in the North Atlantic and adjacent seas was studied by a combination of new observations on a large number of plankton samples collected from the northeastern Atlantic and North Sea, data from cruises off the east coast of North America and Caribbean Sea, and reports in the literature of the past 90 years. Seventy species were recorded, and their distribution was examined by several methods. Distribution maps were plotted for all species, and from these the ranges of temperature tolerance were derived. The 240 sets of data, which took the form of lists of species present in 5° latitude / longitude blocks obtained from the new work and the published material, were analyzed by clustering and ordination multivariate techniques using the programs Twinspan and Decorana. Analysis of the individual species showed that surface water temperature is the most important factor determining distribution and the number of species in a particular area. Warm water and /or low latitudes have many more species than cold waters and/or high latitudes. For example, at 5°N there are on average 23 species per block, whereas at 60° N there are only 8 species. On the basis of this work, the Ceratium species are divided into Group 1, Arctic-temperate species normally only found in water of less than 15°C; Group 2, cosmopolitan species, which are found virtually everywhere and are the species most likely to form blooms or “brown water”; Group 3, intermediate species, which extend into neither the coldest nor the warmest water; Group 4, temperate-tropical species, which have a lower temperature boundary of 5°–12° C; Group 5, warm-temperate-tropical species with a lower temperature boundary of 14°–15°C; and Group 6, tropical species, which are rarely found in water of less than 20° C. Analysis of the sample sites also confirmed the predominant influence of temperature, and the Atlantic Ocean was divided into four biogeographical zones of which the boundaries follow isotherms of surface water temperature. Zone 1 consists of the Arctic and subarctic area, with the southern boundary closely following the 10° mean annual temperature (MAT) line. Zone 2 is an intermediate or cold-temperate zone, of which the southern boundary follows the winter 10° C MAT isotherm or the similarly placed summer 15° isotherm. Zone 3 is the warm-temperate or subtropical zone, which is very broad. The southern boundary closely follows the 25°C summer isotherm. Zone 4 is the tropical zone, where water temperature is never likely to be much less than 23°C. These findings are discussed in relation to experimental work and environmental observations. We suggest that the genus Ceratium provides an excellent tool for defining ocean currents and temperature changes and may become of value in studies of global change.     

PHOTOINHIBITION OF MECHANICALLY STIMULABLE BIOLUMINESCENCE IN THE AUTOTROPHIC DINOFLAGELLATE CERATIUM FUSUS (PYRROPHYTA)1

Ceratium fusus (Ehrenb.) Dujardin was exposed to light of different wavelengths and photon flux densities (PFDs) to examine their effects on mechanically stimulable bioluminescence (MSL). Photoinhibition of MSL was proportional to the logarithm of PFD. Exposure to I μmol photons·m^?2s^?1 of broadband blue light (ca. 400–500 nm) produced near-complete photoinhibition (≥90% reduction in MSL) with a threshold at ca. 0.01 μmol photons·m^?2·s^?1. The threshold of photoinhibition was ca. an order of magnitude greater for both broadband green (ca. 500–580 nm) and red light (ca. 660–700 nm). Exposure to narrow spectral bands (ca. 10 nm half bandwidth) from 400 and 700 nm at a PFD of 0.1 μmol photons·m^?2·s^?1 produced a maximal response of photoinhibition in the blue wavelengths (peak ca. 490 nm). A photoinhibition response (≥ 10%) in the green (ca. 500–540 nm) and red wavelengths (ca. 680 nm) occurred only at higher PFDs (1 and 10 μmol photons·m^?2·s^?1). The spectral response is similar to that reported for Gonyaulax polyedra Stein and Pyrocystis lunula Schütt and unlike that of Alexandrium tamarense (Lebour) Balech et Tangen. The dinoflagellate's own bioluminescence is two orders of magnitude too low to result in self-photoinhibition. The quantitative relationships developed in the laboratory predict photoinhibition of bioluminescence in populations of C. fusus in the North Atlantic Ocean.     

THE TOXIC DINOFLAGELLATE GYMNODINIUM CATENATUM (DINOPHYCEAE) REQUIRES MARINE BACTERIA FOR GROWTH1

Interactions with the bacterial community are increasingly considered to have a significant influence on marine phytoplankton populations. Here we used a simplified dinoflagellate‐bacterium experimental culture model to conclusively demonstrate that the toxic dinoflagellate Gymnodinium catenatum H. W. Graham requires growth‐stimulatory marine bacteria for postgermination survival and growth, from the point of resting cyst germination through to vegetative growth at bloom concentrations (10³ cells · mL^?1). Cysts of G. catenatum were germinated and grown in unibacterial coculture with antibiotic‐resistant or antibiotic‐sensitive Marinobacter sp. DG879 or Brachybacterium sp., and with mixtures of these two bacteria. Addition of antibiotics to cultures grown with antibiotic‐sensitive strains of bacteria resulted in death of the dinoflagellate culture, whereas cultures grown with antibiotic‐resistant bacteria survived antibiotic addition and continued to grow beyond the 21 d experiment. Removal of either bacterial type from mixed‐bacterial dinoflagellate cultures (using an antibiotic) resulted in cessation of dinoflagellate growth until bacterial concentration recovered to preaddition concentrations, suggesting that the bacterial growth factors are used for dinoflagellate growth or are labile. Examination of published reports of axenic dinoflagellate culture indicate that a requirement for bacteria is not universal among dinoflagellates, but rather that species may vary in their relative reliance on, and relationship with, the bacterial community. The experimental model approach described here solves a number of inherent and logical problems plaguing studies of algal‐bacterium interactions and provides a flexible and tractable tool that can be extended to examine bacterial interactions with other phytoplankton species.     

A UNIQUE MITOTIC VARIATION IN THE MARINE DINOFLAGELLATE OXYRRHIS MARINA (PYRROPHYTA)1

Mitosis is described in the flagellate Oxyrrhis marina Dujardin and is compared in related genera. Dense plaques develop in the nuclear envelope at prophase and give rise to an intranuclear spindle. Some of the microtubules associate with the chromosomes while others extend across the nucleus. The basal bodies migrate toward the poles early in division and retain a position lateral to the nuclear poles throughout mitosis. Microtubules are not present between the nucleus and the basal bodies. The nucleolus is persistent and elongates throughout anaphase and telophase. Chromosomal separation is accomplished by sliding of non-chromosomal microtubules and by elongation of the nuclear envelope rather than by shortening of the spindle microtubules. The nuclear envelope begins to constrict in the center early in anaphase. Continued constriction of the envelope and elongation of the nucleus leads to the formation of a dumbbell-shaped nucleus by late telophase. Mitosis culminates by the constriction of the nucleus into two daughter nuclei. The taxonomic position of Oxyrrhis marina is discussed in light of these findings.     

EFFECTS OF METALS AND ORGANIC CONTAMINANTS ON THE RECOVERY OF BIOLUMINESCENCE IN THE MARINE DINOFLAGELLATE PYROCYSTIS LUNULA (DINOPHYCEAE)1

Pyrocystis lunula Schütt is a unicellular photoautotrophic dinoflagellate, commonly found in marine environments, displaying circadian‐controlled bioluminescence. Because of this species' characteristics, effects of pollutants on bioluminescence in P. lunula may make for an easy and simple bioassay that would be valuable for toxicity testing and the protection of coastal resources. This study therefore investigated the short‐term effects of metals and organic pollutants on the recovery of the bioluminescent potential in P. lunula. Recovery of bioluminescence was strongly inhibited in a dose‐dependent manner by all reference contaminants tested, the system being most sensitive to copper and cadmium (4‐h IC₅₀s 0.96 and 1.18 μM, respectively), followed by phenanthrene, lead, SDS, and nickel (4‐h IC₅₀s 1.64, 12.8, 15.6, and 73.1 μM, respectively), whereas relatively high concentrations of phenol were needed to elicit a response (4‐h IC₅₀ 1.64 mM). Except for exposure to lead and nickel, the inhibitory effects of cadmium, copper, and all organic pollutants were reversible, with P. lunula recovering 80%–100% of its bioluminescence potential after a period of 72 h in uncontaminated medium. Our results show that the restoration of bioluminescence in P. lunula is sensitive to the reference contaminants tested and obtains highly reproducible results.     

EVIDENCE FOR ASEXUAL RESTING CYSTS IN THE LIFE CYCLE OF THE MARINE PERIDINOID DINOFLAGELLATE,SCRIPPSIELLA HANGOEI1

Scrippsiella hangoei (Schiller) Larsen is a peridinoid dinoflagellate that grows during winter and spring in the Baltic Sea. In culture this species formed round, smooth cysts when strains were mixed, indicating heterothallic sexuality and hypnozygote production. However, cysts of the same morphology were also formed in clonal strains exposed to slightly elevated temperature. To better understand the role of cysts in the life cycle of S. hangoei, cyst formation and dormancy were examined in culture experiments and the cellular DNA content of flagellate cells and cysts was compared in clonal and mixed strains using flow cytometry. S. hangoei exhibited a high rate of cyst formation in culture. Cysts produced in both clonal and mixed strain cultures were thick‐walled and underwent a dormancy period of 4 months before germinating. The S. hangoei flagellate cell population DNA distributions consisted of 1C, intermediate, and 2C DNA, indicative of respective eukaryotic cell cycle phases G1, S, and G2M. The majority (>95%) of cysts had a measured DNA content equivalent to the lower 1C DNA value, indicating a haploid nuclear phase and an asexual mode of cyst formation. A small percentage (<5%) of cysts produced in the mixed strain culture had 2C DNA, and thus could have been diploid zygotes. These findings represent the first measurements of dinoflagellate resting cyst DNA content, and provide the first quantitative evidence for dinoflagellate asexual resting cysts. Asexual resting cysts may be a more common feature of dinoflagellate life cycles than previously thought.     

PIGMENTS OF THE DINOFLAGELLATE PERIDINIUM BALTICUM AND ITS PHOTOSYNTHETIC ENDOSYMBIONT1

An examination of the pigments of the binucleate dinoflagellate Peridinium balticum (Levander) Lemmerman revealed the presence of chlorophylls a, c₁ and c₂ and the carotenoids: fucoxanthin (most abundant), diadinoxanthin, diatoxanthin, an unidentified fucoxanthin-like xanthophyll, β-carotene, γ-carotene and astaxanthin. A comparison of the pigments of P. balticum and P. foliaceum (Stein) Biecheler, also a binucleate dinoflagellate, demonstrated similar compositions. However P. balticum lacked the β-carotene precursors (e.g. phytoene) which accumulated outside the chloroplast in P. foliaceum. This study indicates that P. balticum and P. foliaceum are closely related; each species is a heterotrophic dinoflagellate with a photosynthetic endosymbiont taxonomically affiliated with the Chrysophyta (Chrysophyceae or Bacillariophyceae).     

SEXUAL REPRODUCTION IN THE FRESHWATER DINOFLAGELLATE GLOEODINIUM MONTANUM1

The sexual life cycle of Gloeodinium montanum Klebs was examined with light and scanning electron microscopy. In unialgal cultures G. montanum divided predominantly by simple division, giving rise to two nonmotile cells. When placed in fresh medium, 2–4 biflagellated swarmers were formed from the vegetative cells. Swarmers developed directly into vegetative cells or acted as gametes. Both isogamy and anisogamy were observed. Gloeodinium montanum is homothallic. Fusion occurred in the non-motile state producing a large aplanozygote, which germinated after approximately two months to a year or more. Zygote germination liberated four aplanospores. Budding of the zygote, resulting from unequal division of the protoplast and multiple fusion attempts also were observed.     

PHENOTYPIC VARIATION AND GENOTYPIC DIVERSITY IN A PLANKTONIC POPULATION OF THE TOXIGENIC MARINE DINOFLAGELLATE ALEXANDRIUM TAMARENSE (DINOPHYCEAE)1

Multiple clonal isolates from a geographic population of Alexandrium tamarense (M. Lebour) Balech from the North Sea exhibited high genotypic and phenotypic variation. Genetic heterogeneity was such that no clonal lineage was repeatedly sampled according to genotypic markers specified by amplified fragment length polymorphism (AFLP) and microsatellites. Subsampling of genotypic data from both markers showed that ordination of individuals by pair‐wise genetic dissimilarity indices was more reliable by AFLP (482 biallelic loci) than by microsatellites (18 loci). However, resulting patterns of pair‐wise genetic similarities from both markers were significantly correlated (Mantel test P < 0.005). The composition of neurotoxins associated with paralytic shellfish poisoning (PSP) was also highly diverse among these isolates and allowed clustering of toxin phenotypes based on prevalence of individual toxins. Correlation analysis of pair‐wise relatedness of individual clones according to PSP‐toxin profiles and both genotypic characters failed to yield close associations. The expression of allelochemical properties against the cryptophyte Rhodomonas salina (Wis?ouch) D. R. A. Hill et Wetherbee and the predatory dinoflagellate Oxyrrhis marina Dujard. manifested population‐wide variation of responses in the target species, from no visible effect to complete lysis of target cells. Whereas the high genotypic variation indicates high potential for adaptability of the population, we interpret the wide phenotypic variation as evidence for lack of strong selective pressure on respective phenotypic traits at the time the population was sampled. Population markers as applied here may elucidate the ecological significance of respective traits when followed under variable environmental conditions, thereby revealing how variation is maintained within populations.     

MORPHOLOGY AND ECOLOGY OF THE MARINE BENTHIC DINOFLAGELLATE SCRIPPSIELLA SUBSALSA (DINOPHYCEAE)1

The thecal surface morphology of Scrippsiella subsalsa (Ostenfeld) Steidinger et Balech was examined using the scanning electron microscope. This species is distinguished by a number of morphological characteristics. Apical plate 1′ is wide, asymmetric, and pentagonal, and it ends at the anterior margin of the cingulum. Intercalary plates 2a and 3a are separated by apical plate 3′. The apical pore complex includes a large P_o plate with a raised dome at the center and a deep canal plate with thickened margins at plates 2′, 3′, and 4′. The intercalary bands are wide and deeply striated. The cingulum is deep, formed by six cingular plates; its surface is transversely striated and aligned with a row of minute pores. The cingular list continues around postcingular plate 1′” to form a sulcal list. The sulcal list is a flexible ribbon with a rounded tip that protrudes posteriorly, partially covering the sulcal plates. The hypotheca is lobed, and the antapical plates are irregularly shaped and wide in antapical view. The thecal surface is vermiculate to reticulate. A comparison in morphology and ecology is presented between S. subsalsa and other known Scrippsiella species.     

SEXUAL REPRODUCTION IN THE TOXIC DINOFLAGELLATE GONYAULAX MONILATA1

The sexual cycle of Gonyaulax monilata Howell was observed in stationary cultures and in nitrogen-deficient medium. The armored, isogamous gametes fuse in a characteristic manner with cingula at oblique angles. Nuclear fusion lags slightly behind cytoplasmic fusion. The zygote enlarges for several days. The dark, double-flagellated planozygote encysts within 1–3 wk. Early hypnozygotes are round to ovoid and contain lipid and one or two large golden-yellow globules. As the hypnozygote matures, the globules become smaller and the cytoplasm darkens and pulls from the wall. All cysts examined contained only one nucleus. A very dark, uninucleate post-hypnozygotic cell escapes through an archeopyle and within 24 h divides into daughter cells which divide in 24–48 h forming a small chain. The production of thick walled zygates in culture implies that such resting stages in marine sediments could serve as a source stock for blooms. This species causes toxic red tides and the existence of benthic “seed beds” consisting of hypnozygotes is now plausible.     

SALINITY EFFECTS ON GROWTH AND TOXIN CONTENT OF GONYAULAX EXCAVATA. A MARINE DINOFLAGELLATE CAUSING PARALYTIC SHELLFISH POISONING1

The optimum salinity for growth of Gonyaulax excavata (Braarud) Balech from Cape Ann, Massachusetts, is 30.5%o, and it grows well over a range of 20–40%o.It tolerates salinities of 11–43%o. Growth rates at 24 and 20%o are only ca. 10 and 20% less, respectively, than the maximum, 036 divisions/day. It is considered unlikely that salinity fluctuations in the coastal areas where this organism occurs significantly influence its growth rate. The paralytic toxin content in G. excavata increases with increasing salinity, up to 37%o. Therefore, the degree of toxicity of the organism in nature may be influenced by this environmental factor.     

PHAGOTROPHIC FEEDING AND ITS IMPORTANCE TO THE LIFE CYCLE OF THE HOLOZOIC DINOFLAGELLATE,GYMNODINIUM FUNGIFORME1

The holozoic dinoflagellate, Gymnodinium fungiforme Anissimova, has been observed in both asexually and sexually reproducing cultures. Asexual reproduction is characterized by zoosporangium formation and subsequent new cell release. Sexuality is gametic, and planozygotes and hypnozygotes are present. The life cycle is highly dependent on feeding, and in food-depleted cultures the swimming cells rapidly disappear. These are replaced with resistant long-term resting cysts. Despite its small size (8.5–19 μm), G. fungiforme can feed on prey as large as the ciliated protozoan, Condylostoma magnum Spiegel (600–1000 μm in length), or small injured metazoans, and has been cultured phagotrophically with the chlorophyte, Dunaliella salina Teodoresco as a food source. Eleven additional species of algae including 1 chlorophyte, 7 chrysophytes and 3 rhodophytes, however, were not suitable as food sources. Feeding is characterized by the formation of ‘dynamic aggregations’ of hundreds of dinoflagellates that attach to the surface of a prey organism by a peduncle. G. fungiforme ingests the cytoplasm or body fluids of its prey and a feeding aggregation can ingest a C. magnum in 20–30 minutes.     

INTRACELLULAR LOCALIZATION OF SAXITOXINS IN THE DINOFLAGELLATE GONYAULAX TAMARENSIS1

The potent neurotoxin saxitoxin, and possibly several of its derivatives, are localized in two types of sites within the marine dinoflagellate Gonyaulax tamarensis Lebour. Immunocytochemical techniques using a polyclonal antibody and epifluorescence microscopy demonstrate toxin localization within the nucleus as well as on the periphery of small granules thought to be starch grains. In the nuclear region, the labelling occurred on or close to the permanently-condensed chromosomes as well as in an area within the two arms of the nucleus in the vicinity of the nucleous. No binding was observed in a closely-related, non-toxic dinoflagellate. Different binding affinities were observed between the nucleus and the grains at high and low antibody dilutions. This may relate to the polyclonal nature of the antiserum and to the presence of multiple toxins within the G. tamarensis isolate studied. Mechanistic interpretations of these labelling patterns remain speculative, especially the localization of the antigen at the outer edge of starch grains, but the distinct labelling in the nuclear region suggests that saxitoxin, with its two positively charged guanidinium groups, may bind to nucleic acids or nuclear proteins in a manner analogous to the polyamines and other cations. The labelling patterns reported here suggest that the saxitoxins may not simply be secondary metabolites but instead could be important compounds involved in the structure and function of the G. tamarensis genome.     

ADAPTATION OF CERATIUM FURCA AND GONYAULAX POLYEDRA (DINOPHYCEACE) TO DIFFERENT TEMPERATURES AND IRRADIANCES: GROWTH RATES AND CELL VOLUMES1

Growth rates and cell volumes of Ceratium furca Ehrenberg and Gonyaulax polyedra Stein were determined during the log phase of growth in cultures which had been extensively adapted to one of three temperatures and five irradiances. At each temperature, curves for the growth rate vs. irradiance for both species had light-limited and light-saturated regions. Three properties of these curves characterized the response of each species to temperature: the light-saturated growth rate, the irradiance at which growth became light-saturated and the compensation irradiance for growth. For both species, the first two properties generally decreased with declining growth temperature, while the compensation irradiance declined for Ceratium but had a V-shaped response pattern for Gonyaulax. The light-saturated growth rates were generally higher for Ceratium than for Gonyaulax, while the irradiance at which growth became saturated and the compensation irradiance were lower for Ceratium. The changes in cell volume associated with the irradiance and temperature of growth were very different for Ceratium and Gonyaulax. The cell size of Gonyaulax increased as irradiance and temperature decreased, while cell volumes of Ceratium did not change with temperature but were smallest at the highest and lowest growth irradiances. In general, the growth rate patterns were similar for Ceratium and Gonyaulax, while those for cell size were different. The maximum growth rate, the irradiance at which growth became saturated, the compensation irradiance, and the cell volume all showed that Ceratium grew at the same rate or faster than Gonyaulax over the entire range of irradiances and temperatures examined.     

PHAGOTROPHY IN THE FRESHWATER,PHOTOSYNTHETIC DINOFLAGELLATE AMPHIDINIUM CRYOPHILUM1

The cold-water, photosynthetic dinoflagellate Amphidinium cryophilum Wedemayer, Wilcox & Graham feeds phagotrophically on other dinoflagellate species. Food is ingested through a feeding tube, termed here the “phagopod,” which extends from the antapex. The peduncle of this organism plays no observable role in the feeding process. The phagopod is essentially a hollow cylinder composed electron-opaque material that is possibly deposited on a membrane. No Amphidinium cytoplasmic components, including microtubules or other cytoskeletal elements, were observed in the phagopod. Prefeeding cells aggregate, in small clumps near prey organisms with their phagopods extended. Eventually some cells commence feeding, first inserting the phagopod through the prey cell-covering and then slowly, over a period of 10 min or more, drawing cytoplasm through the phagopod and into a nascent food vacuole. Both light and electron microscopy suggest that one or more prey cell amphiesmal membranes remain intact during the feeding process. Upon completion of feeding, the Amphidinium cell swims off with a prominent food vacuole in the hypocone, leaving at least part of the phagopod attached to the prey cell. Phagotrophy in A. cryophilum seems to vary with light intensity. At low light intensities, cells feed phagotrophically and are nearly colorless, whereas at high light levels they feed much less frequently, if at all, and are brightly pigmented.