Food selection by calanoid copepods in response to between-lake variation in food abundance

1. Food selection experiments were conducted by acclimating calanoid copepods (Eudiaptomus spp.) in suspensions of natural seston and then adding pairs of dual-labelled (¹⁴C/³²P) algae. Each feeding trial measured selectivity between a small, high-quality reference alga, Chlamydomonas reinhardii, and a test alga that differed in size and/or food quality. The influence of food concentration on food selection was tested by using seston from two lakes with contrasting food abundance and by including treatments with filtered lake water ('starved’) and seston diluted with filtered water or enriched with cultured algae. 2. Copepods that had been starved or acclimated to natural seston with low food abundance preferred the larger of two labelled algae, regardless of the nutritional quality of the algae. In agreement with the predictions of an optimal diet model, however, copepods that were acclimated to high food conditions discriminated against low-quality foods, including digestion-resistant algae and dead algae. 3. Selectivity coefficients showed excellent agreement with a previous study involving the same taxa of copepods and labelled algae but in which the copepods had been acclimated to pairs of cultured algae rather than natural seston. Thus, these comparisons emphasize the importance of food availability in modulating copepod selectivity for foods that differ in nutritional quality and suggest that such behaviour occurs in nature.     

CULTURE OF THE HETEROTROPHIC DINOFLAGELLATE PROTOPERIDINIUM CRASSIPES (DINOPHYCEAE) WITH NONCELLULAR FOOD ITEMS1

The genus Protoperidinium is an assemblage of heterotrophic dinoflagellates, several species of which have been successfully cultured in the past using various photosynthetic algae as a food source. We succeeded in culturing Protoperidinium crassipes (Kof.) Balech on three separate occasions for periods ranging from 2 to 21 months using rice flour as a food source. In these cultures, unusual small types of cells that were never observed to actively feed sometimes appeared. We confirmed that P. crassipes in culture exhibited bioluminescence.     

Does the nutrient stoichiometry of primary producers affect the secondary consumer Pleurobrachia pileus?

We investigated whether phosphorus limitations of primary producers propagate upwards through the food web, not only to the primary consumer level but also onto the secondary consumers’ level. A tri-trophic food chain was used to assess the effects of phosphorus-limited phytoplankton (the cryptophyte Rhodomonas salina) on herbivorous zooplankters (the copepod Acartia tonsa) and finally on zooplanktivores (the ctenophore Pleurobrachia pileus). The algae were cultured in phosphorus-replete and phosphorus-limited media before being fed to two groups of copepods. The copepods in turn were fed to the top predator, P. pileus, in a mixture resulting in a phosphorus-gradient, ranging from copepods having received only phosphorus-replete algae to copepods reared solely on phosphorus-limited algae. The C:P ratio of the algae varied significantly between the two treatments, resulting in higher C:P ratios for those copepods feeding on phosphorus-limited algae, albeit with a significance of 0.07. The differences in the feeding environment of the copepods could be followed to Pleurobrachia pileus. Contrary to our expectations, we found that phosphorus-limited copepods represented a higher quality food source for P. pileus, as shown by the better condition (expressed as nucleic acid content) of the ctenophore. This could possibly be explained by the rather high C:P ratios of ctenophores, their resulting low phosphorus demand and relative insensitivity to P deficiency. This might potentially be an additional explanation for the observed increasing abundances of gelatinous zooplankton in our increasingly phosphorus-limited coastal seas.     

Dead bleached coral provides new surfaces for dinoflagellates implicated in ciguatera fish poisonings

Synopsis Dead sections of bleached corals in the United States Virgin Islands and British Virgin Islands were found to be colonized by filamentous algae harboring epiphytic dinoflagellates implicated in ciguatera fish poisonings. The dinoflagellates Ostreopsis lenticularis, Prorocentrum concavum, and P. lima were found in association with filamentous algae growing on dead sections of bleached Montastrea annularis and Acropora cervicornis. Several fish species from the families Acanthuridae, Pomacentridae, and Scaridae were observed to readily consume this filamentous algae/epiphytic dinoflagellate food source. Such fishes are common prey for large piscivores inhabiting tropical reefs. An increased incidence of ciguatoxic fishes may occur on reefs where bleaching events have caused significant coral mortality.     

Algae or Protozoa: Phylogenetic Position of Euglenophytes and Dinoflagellates as Inferred from Mitochondrial Sequences

The chloroplasts of euglenophytes and dinoflagellates have been suggested to be the vestiges of endosymbiotic algae acquired during the process of evolution. However, the evolutionary positions of these organisms are still inconclusive, and they have been tentatively classified as both algae and protozoa. A representative gene of the mitochondrial genome, cytochrome oxidase subunit I (coxI), was chosen and sequenced to clarify the phylogenetic positions of four dinoflagellates, two euglenophytes and one apicomplexan protist. This is the first report of mitochondrial DNA sequences for dinoflagellates and euglenophytes. Our COXI tree shows clearly that dinoflagellates are closely linked to apicomplexan parasites but not with algae. Euglenophytes and algae appear to be only remotely related, with euglenophytes sharing a possible evolutionary link with kinetoplastids. The COXI tree is in general agreement with the tree based on the nuclear encoded small subunit of ribosomal RNA (SSU rRNA) genes, but conflicts with that based on plastid genes. These results support the interpretation that chloroplasts present in euglenophytes and dinoflagellates were captured from algae through endosymbioses, while their mitochondria were inherited from the host cell. We suggest that dinoflagellates and euglenophytes were originally heterotrophic protists and that their chloroplasts are remnants of endosymbiotic algae. Received: 24 March 1997 / Accepted: 21 April 1997     

Effects of poor food quality on copepod growth are dose dependent and non‐reversible

Herbivores regularly face imbalanced diets on a variety of timescales. They respond to such diets with reductions in growth and reproduction. The effect of exposure time to poor quality diets and recovery potential in herbivores has yet not been intensively studied. In order to investigate the response of herbivores growth and nutritional condition to phosphorus limited algal food, we fed the copepod Acartia tonsa with the autotrophic flagellate Rhodomonas salina, which was cultured towards high and low phosphorus content. To test the effect of the duration of the exposure to low and high P food on copepod growth, we also switched a subset of the copepods from high P to low P food for two to 12 days before putting them back on their original diet. Phosphorus limited prey clearly reduced copepod nutritional condition, expressed as their RNA:DNA ratio and consequently their growth rates. Several days after re‐feeding on high P algae, the copepods were able to reach growth rates comparable to the control groups, but they were in no case able to over shoot the growth rates of the control groups. A dose of two days feeding on low P algae resulted in slightly less than one stage growth reduction compared to copepods exclusively reared on high P algae. Our results show that phosphorus limited food drastically reduces secondary production, and no compensatory growth occurs in the investigated copepod species. Hence, even short term exposure of herbivores to poor food has a lasting effect on secondary production.     

Vectorial transport of toxins from the dinoflagellate Gymnodinium breve through copepods to fish

Toxins from the dinoflagellate Gymnodinium breve were tracedthrough experimental food chains from dinoflagellates, throughcopepod grazers, to juvenile fish. The generality of this foodweb transfer was demonstrated using three different combinationsof copepods and juvenile fish during different seasons. Fishwere not exposed directly to the toxic dinoflagellates but werefed toxin-laden copepods in order to examine sublethal vectorialintoxication. Toxins were shown to move from fish viscera tomuscle tissue within periods of 2–6 h to 25 h. A new toxindetection method was used in this first stepwise demonstrationof multi-trophic-level intoxication of a planktonic food chainby G.breve. Micellar electrokinetic capillary electrophoresiswith laser-induced fluorescence allowed measurements of toxinsat trace levels and nanoliter-sized volumes critical for planktonicfood web transfer studies.     

Copepods act as a switch between alternative trophic cascades in marine pelagic food webs

A recent meta‐analysis indicates that trophic cascades (indirect effects of predators on plants via herbivores) are weak in marine plankton in striking contrast to freshwater plankton ( Shurin et al. 2002 , Ecol. Lett., 5, 785–791). Here we show that in a marine plankton community consisting of jellyfish, calanoid copepods and algae, jellyfish predation consistently reduced copepods but produced two distinct, opposite responses of algal biomass. Calanoid copepods act as a switch between alternative trophic cascades along food chains of different length and with counteracting effects on algal biomass. Copepods reduced large algae but simultaneously promoted small algae by feeding on ciliates. The net effect of jellyfish on total algal biomass was positive when large algae were initially abundant in the phytoplankton, negative when small algae were dominant, but zero when experiments were analysed in combination. In contrast to marine systems, major pathways of energy flow in Daphnia‐dominated freshwater systems are of similar chain length. Thus, differences in the length of alternative, parallel food chains may explain the apparent discrepancy in trophic cascade strength between freshwater and marine planktonic systems.     

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.     

Influence of temperature and body weight on mosquito predation by the dragonfly nymph Mesogomphus lineatus

Population densities, migrations and food composition of Echinogammarus veneris in Lake Kinneret were studied during march–September 1976. Migrations are affected by changes of water level and littoral currents. Peridinium was abundant in the gut content during its blooming period in the lake while cladocerans, benthic copepods, rotifers and other algae were abundantly found in other months. Metabolic parameters that were calculated for the gammarid population indicated that its contribution to the carbon-flow system in Lake Kinneret is negligible in comparison with nematodes, molluscans or planktonic copepods and cladocerans.     

Structural Confirmation of a Unique Carotenoid Lactoside,P457, in Symbiodinium sp. Strain nbrc 104787 Isolated from a Sea Anemone and its Distribution in Dinoflagellates and Various Marine Organisms

The molecular structure of the carotenoid lactoside P457, (3S,5R,6R,3′S,5′R,6′S)‐13′‐cis‐5,6‐epoxy‐3′,5′‐dihydroxy‐3‐(β‐d ‐galactosyl‐(1→4)‐β‐d ‐glucosyl)oxy‐6′,7′‐didehydro‐5,6,7,8,5′,6′‐hexahydro‐β,β‐caroten‐20‐al, was confirmed by spectroscopic methods using Symbiodinium sp. strain NBRC 104787 cells isolated from a sea anemone. Among various algae, cyanobacteria, land plants, and marine invertebrates, the distribution of this unique diglycosyl carotenoid was restricted to free‐living peridinin‐containing dinoflagellates and marine invertebrates that harbor peridinin‐containing zooxanthellae. Neoxanthin appeared to be a common precursor for biosynthesis of peridinin and P457, although neoxanthin was not found in peridinin‐containing dinoflagellates. Fucoxanthin‐containing dinoflagellates did not possess peridinin or P457; green dinoflagellates, which contain chlorophyll a and b, did not contain peridinin, fucoxanthin, or P457; and no unicellular algae containing both peridinin and P457, other than peridinin‐containing dinoflagellates, have been observed. Therefore, the biosynthetic pathways for peridinin and P457 may have been coestablished during the evolution of dinoflagellates after the host heterotrophic eukaryotic microorganism formed a symbiotic association with red alga that does not contain peridinin or P457.     

Macroalgal palatability and the flux of ciguatera toxins through marine food webs

The benthic dinoflagellate Gambierdiscus toxicus produces polyether toxins that cause ciguatera fish poisoning in humans. The toxins initially enter food webs when fish forage on macroalgae, or other substrates, hosting this epiphytic dinoflagellate. Population studies of G. toxicus and risk assessments in ciguatera-prone regions often rely on quantifying dinoflagellates on macroalgae. Underlying these studies is the assumption that the algae sampled represent a readily consumable resource equally available for benthic grazers. However, many algal hosts of G. toxicus possess a variety of defenses against grazing, and host–dinoflagellate associations may act as toxin sources or sinks depending on their palatability. Marine macroalgae may tolerate or avoid herbivory by exhibiting fast growth, by having poor nutritional quality, by utilizing spatial or temporal escapes or by using chemical or structural defenses. Thus, rapidly consumed algae that cope with herbivores by growing fast, such as many filamentous turfs, could be responsible for a high toxin flux even at low dinoflagellate densities. In contrast, ubiquitous unpalatable algae with much higher dinoflagellate densities might contribute little to toxin flux, and effectively act as refuges for G. toxicus. To date, G. toxicus has been reported from 56 algal genera, two cyanobacteria, one diatom, and one seagrass; 63% of these contain species that are defended from fish grazing and other grazers via chemical, morphological or structural defenses, by low nutritional quality, or by a combination of defensive strategies. High dinoflagellate densities on unpalatable macroalgae could indicate passive accumulation of cells on undisturbed hosts, rather than population explosions or active toxin sources for food webs. Understanding the flow of ciguatoxins in nature requires consideration of the ecology of both G. toxicus and its algal hosts. The complexity of marine algal–herbivore interactions also has consequences for other benthic dinoflagellates that produce toxins, which accumulate in consumers.     

A SURVEY OF THE STEROL COMPOSITION OF THE MARINE DINOFLAGELLATES KARENIA BREVIS,KARENIA MIKIMOTOI,AND KARLODINIUM MICRUM: DISTRIBUTION OF STEROLS WITHIN OTHER MEMBERS OF THE CLASS DINOPHYCEAE1

The sterol composition of different marine microalgae has been examined to determine the utility of sterols as biomarkers to distinguish members of various algal classes. For example, members of the class Dinophyceae possess certain 4‐methyl sterols, such as dinosterol, which are rarely found in other classes of algae. The ability to use sterol biomarkers to distinguish certain dinoflagellates such as the toxic species Karenia brevis Hansen and Moestrup, responsible for red tide events in the Gulf of Mexico, from other species within the same class would be of considerable scientific and economic value. Karenia brevis has been shown by others to possess two major sterols, (24S)‐4α‐methyl‐5α‐ergosta‐8(14),22‐dien‐3β‐ol (ED) and its 27‐nor derivative (NED), having novel structures not previously known to be present in other dinoflagellates. This prompted the present study of the sterol signatures of more than 40 dinoflagellates. In this survey, sterols with the properties of ED and NED were found in cultures of K. brevis and shown also to be the principal sterols of Karenia mikimotoi Hansen and Moestrup and Karlodinium micrum Larsen, two dinoflagellates closely related to K. brevis. They are also found as minor components of the more complex sterol profiles of other members of the Gymnodinium/Peridinium/Prorocentrum (GPP) taxonomic group. The distribution of these sterols is consistent with the known close relationship between K. brevis, K. mikimotoi, and K. micrum and serves to limit the use of these sterols as lipid biomarkers to a few related species of dinoflagellates.     

Growth of dominant zooplankton species feeding on plankton microflora in Lake Shira

Batch cultures and continuous flow cultures were used to study the growth rates of zooplankton species from Shira lake, the rotifer Brachionus plicatilis Muller and calanoid copepod Arctodiaptomus salinus Daday, which were fed on phytoplankton and bacterioplankton from the lake. Analyses of the birth and survival rates were used to demonstrate that the lake phytoplankton, consisting mostly of cyanobacteria and diatomaceous algae, is inadequotes for optimal realisation of the reproductive potential of B. plicatilis when compared with the bacterial diet. The study revealed that the kinetic growth characteristics of the two zooplankters were similar: B. plicatilis r _max, 0.120 d^–1; S ₀, 0.253; and K _s, 0.114 mg dry mass l^–1; and for A. salinus r _max, 0.129 d^–1; S ₀, 0.240; and K _s, 0.171 mg dry mass l^–1. Fluctuations in natural food concentration reduced the growth rate of both species. Even though the threshold concentration of food for B. plicatilis and A. salinus were quite similar, the copepods were less sensitive to food limitation.     

Diapause of Cyclops vicinus (Uljanin) in Lake Søbygård: Indication of a Risk-spreading Strategy

Cyclops vicinus is reported to enter summer diapause triggered by day length in order to survive food scarcity and fish predation. Development ceases and the fourth or fifth copepodid stages persist in the sediment for several weeks. In Lake Søbygård, however, a small eutrophic lake in Denmark, C. vicinus is found in the sediment as well as in the open water throughout the year. We performed laboratory experiments to elucidate the life cycle strategy of this population. In order to find the factors that induce diapause we tested the influence of food composition and light regime on the proportion of copepods entering diapause. Further we examined the diapause pattern of the offspring from diapausing and non-diapausing parents to see whether the co-occurrence of diapausing and non-diapausing copepods results from genetic differentiation within the population. In all experiments some of the copepods developed directly into adults, while others remained at the fourth or fifth copepodid stage and displayed diapause features. The proportion of copepods that developed directly into adults was influenced by food supply. Fewer individuals entered diapause when they received a mixed diet of algae and ciliates or algae, ciliates and seston compared with a pure algal diet. The response to light was different than described in literature: about 30% of the copepods entered diapause under dark conditions, and the diapause frequency was not higher when copepods had been exposed to continuous light. The offspring from diapausing parents showed no higher tendency to enter diapause than the offspring of non-diapausing parents, indicating no genetic differentiation. We conclude that diapause in C. vicinus represents a risk-spreading strategy, modified by food. The simultaneous production of diapausing and non-diapausing offspring ensures survival under harsh conditions; the influence of the food supply on the proportion of individuals entering diapause may adjust the population to the actual environmental condition. This flexible life cycle strategy might contribute to the exclusive dominance of C. vicinus in Lake Søbygård.     

Diel changes in vertical distribution of Oithona similis (Cyclopoida) and Oncaea curvata (Poecilostomatoida) under sea ice in mid-summer near Syowa Station, Antarctica

The diel vertical distributions of two small copepods, Oithona similis and Oncaea curvata, were investigated at 4-h intervals over a 24-h period under fast ice near Syowa Station during continuous daylight conditions in the Antarctic mid-summer, December 1993. Oithona similis and O. curvata exhibited small-scale diel vertical migrations during the study period, in a way opposite to what is expected, i.e., remaining mostly in the upper layer during the day and moving into deeper layers at night. The nighttime descent of both species coincided with the time of disappearance of a high algal concentration at the ice–water interface during the day and an increase of the algal concentration in the mid-water layer at night. This suggests the migration behavior of the copepods was responsible for the change of food availability. The daily grazing impact of these copepods was estimated to remove one-third of the algae daily released from ice during mid-summer at Syowa Station.     

Quantification of invertebrate predation and herbivory in food chains of low complexity

Zooplankton grazing impact on algae, heterotrophic flagellates and bacteria, as well as invertebrate predation on herbivorous zooplankton, were investigated in two sub-Antarctic lakes with extremely simple food chains. The two species of herbivorous zooplankton present in the lakes (the copepods boeckella michaelseni and Pseudoboeckella poppei) exerted substantial grazing pressure on algae. However, the dominant algal species exhibited properties that enabled them to avoid (large size or extruding spines, e.g. Staurastrum sp., Tribonema sp.) or compensate (recruitment from the sediment, Mallomonas sp.) grazing. There are only two potential invertebrate predators on the herbivorous copepods in the two lakes: the copepod Parabroteas sarsi and the diving beetle Lancetes claussi. Vertebrate predators are entirely abscent from sub-Antarctic lakes. Based on our experiments, we estimated that the predators would remove at most about 0.4% of the herbivorous copepods per day, whereas planktivorous fish, if present in the lakes, would have removed 5–17% of the zooplankton each day. Consequently, the invertebrate predators in these high-latitude lakes had only a marginal predation impact compared to the predation pressure on zooplankton in the presence of vertebrate predators in temperate lakes. The study of these simple systems with only two quantitatively functionally important trophic links, suggests that high grazing pressure foreces the algal community towards forms with grazer resistant adaptations such as large size, recruitment from another habitat, and grazer avoidance spines. We propose that due to such adaptations, predictions from food web theory are only partly corroborated, i.e. algal biomass actually increases with increasing productivity, although the grazer community is released from predation. In more species-rich and complex systems, e.g temperate lakes with three functionally important links, such adaptations are likely to be even more important, and, consequently, the observable effects of trophic interactions from top predators on lower trophic levels even more obscured.     

Fitness consequences for copepods feeding on a red tide dinoflagellate: deciphering the effects of nutritional value, toxicity, and feeding behavior

Phytoplankton exhibit a diversity of morphologies, nutritional values, and potential chemical defenses that could affect the feeding and fitness of zooplankton consumers. However, how phytoplankton traits shape plant–herbivore interactions in the marine plankton is not as well understood as for terrestrial or marine macrophytes and their grazers. The occurrence of blooms of marine dinoflagellates such as Karenia brevis suggests that, for uncertain reasons, grazers are unable to capitalize on, or control, this phytoplankton growth—making these systems appealing for testing mechanisms of grazing deterrence. Using the sympatric copepod Acartia tonsa, we conducted a mixed diet feeding experiment to test whether K. brevis is beneficial, toxic, nutritionally inadequate, or behaviorally rejected as food relative to the palatable and nutritionally adequate phytoplankter Rhodomonas lens. On diets rich in K. brevis, copepods experienced decreased survivorship and decreased egg production per female, but the percentage of eggs that hatched was unaffected. Although copepods showed a 6–17% preference for R. lens over K. brevis on some mixed diets, overall high ingestion rates eliminated the possibility that reduced copepod fitness was caused by copepods avoiding K. brevis, leaving nutritional inadequacy and toxicity as remaining hypotheses. Because egg production was dependent on the amount of R. lens consumed regardless of the amount of K. brevis eaten, there was no evidence that fitness costs were caused by K. brevis toxicity. Copepods limited to K. brevis ate 480% as much as those fed only R. lens, suggesting that copepods attempted to compensate for low food quality with increased quantity ingested. Our results indicate that K. brevis is a poor food for A. tonsa, probably due to nutritional inadequacy rather than toxicity, which could affect bloom dynamics in the Gulf of Mexico where these species co-occur.     

Omnivory in cyclopoid copepods: comparisons of algae and invertebrates as food for three, differenfly sized species

We examined the importance of algal versus invertebrate preyfor three cyclopoid copepods by comparing reproductive successfor females fed on three diets: exclusively algal food, exclusivelyinvertebrate prey and a combination of algal plus invertebrateprey. The three cyclopoid species represent a distinct gradientin body size: Tropocyclops prasinus mexicanus (small body size),Diacyclops thomasi (medium body size) and Mesocyclops edax (largebody size). Our results provide strong evidence for the potentialimportance of algae in the diet of adult cyclopoid copepods.However, the role of herbivory is dependent on species bodysize. We found that algae are the key dietary factor for reproductivesuccess of the small T.p.mexicanus. In contrast, the availabilityof invertebrate prey had a greater influence on reproductionby the two larger species. Overall, cyclopoid copepods shouldbe generally considered as having the potential to derive theirenergy from either animal or algal sources. This has importantconsequences for interpretations of the forces influencing zooplanktoncommunity diversity as well as for general food web theory.     

Marine microalgae attack and feed on metazoans

Free-living microalgae from the dinoflagellate genus Karlodinium are known to form massive blooms in eutrophic coastal waters worldwide and are often associated with fish kills. Natural bloom populations, recently shown to consist of the two mixotrophic and toxic species Karlodinium armiger and Karlodinium veneficum have caused fast paralysis and mortality of finfish and copepods in the laboratory, and have been associated with reduced metazooplankton biomass in-situ. Here we show that a strain of K. armiger (K-0688) immobilises the common marine copepod Acartia tonsa in a density-dependent manner and collectively ingests the grazer to promote its own growth rate. In contrast, four strains of K. veneficum did not attack or affect the motility and survival of the copepods. Copepod immobilisation by the K. armiger strain was fast (within 15 min) and caused by attacks of swarming cells, likely through the transfer and action of a highly potent but uncharacterised neurotoxin. The copepods grazed and reproduced on a diet of K. armiger at densities below 1000, cells ml⁻¹, but above 3500 cells ml⁻¹ the mixotrophic dinoflagellates immobilised, fed on and killed the copepods. Switching the trophic role of the microalgae from prey to predator of copepods couples population growth to reduced grazing pressure, promoting the persistence of blooms at high densities. K. armiger also fed on three other metazoan organisms offered, suggesting that active predation by mixotrophic dinoflagellates may be directly involved in causing mortalities at several trophic levels in the marine food web.