LIFE CYCLE OF THE HETEROTROPHIC DINOFLAGELLATE PFIESTERIA PISCICIDA (DINOPHYCEAE)1

The putatively toxic dinoflagellate Pfiesteria piscicida (Steidinger et Burkholder) has been reported to have an unusual life cycle for a free‐living marine dinoflagellate. As many as 24 life cycle stages were originally described for this species. During a recent phylogenetic study in which we used clonal cultures of P. piscicida, we were unable to confirm many reported life cycle stages. To resolve this discrepancy, we undertook a rigorous examination of the life cycle of P. piscicida using nuclear staining techniques combined with traditional light microscopy, high‐resolution video microscopy, EM, and in situ hybridization with a suite of fluorescently labeled peptide nucleic acid (PNA) probes. The results showed that P. piscicida had a typical haplontic dinoflagellate life cycle. Asexual division occurred within a division cyst and not by binary fission of motile cells. Sexual reproduction of this homothallic species occurred via the fusion of isogamous gametes. Examination of tanks where P. piscicida was actively feeding on fish showed that amoebae were present; however, they were contaminants introduced with the fish. Whole cell probing using in situ hybridization techniques confirmed that these amoebae were hybridization negative for a P. piscicida‐specific PNA probe. Direct observations of clonal P. piscicida cultures revealed no unusual life cycle stages. Furthermore, the results of this study provided no evidence for transformations to amoebae. We therefore conclude that P. piscicida has a life cycle typical of free‐living marine dinoflagellates and lacks any amoeboid or other specious stages.     

THE LIFE CYCLE AND TOXICITY OF PFIESTERIA PISCICIDA REVISITED1

Despite use of excellent molecular techniques, Litaker et al. (2002) cannot provide insights about the life history of toxic Pfiesteria piscicida because they showed no data in support of having used toxic strains; rather they presented evidence that they used non‐inducible strains. Litaker et al. did not find amoeboid stages or a chrysophyte‐like cyst stage in several cultures and unequivocally concluded that the stages do not exist in all P. piscicida strains. Thus, they did not consider the tenet that absence of evidence does not constitute proof of absence. Apparent discrepancies between the research by Litaker et al. and previous research on Pfiesteria can be resolved as follows: First, Litaker et al. did not use toxic strains. We have reported findings (similar to Litaker et al.) showing few amoeboid transformations in non‐inducible strains, which manifest some but not all of the forms that have been documented in some toxic strains. We, and others, have documented active toxicity to fish, transformations to amoebae, and chrysophyte‐like cysts in some clonal toxic strains. Second, the data from several recent publications, which were available but not mentioned by Litaker et al. or by Coats (2002) in accompanying commentary, have verified P. piscicida amoebae, chrysophyte‐like cysts, and other stages in some toxic strains through a combination of approaches including PCR data from clonal cultures.     

PFIESTERIA PISCICIDA GEN. ET SP. NOV. (PFIESTERIACEAE FAM. NOV.), A NEW TOXIC DINOFLAGELLATE WITH A COMPLEX LIFE CYCLE AND BEHAVIOR1

The newly described toxic dinoflagellate Pfiesteria piscicida is a polymorphic and multiphasic species with flagellated, amoeboid, and cyst stages. The species is structurally a heterotroph; however, the flagellated stages can have cleptochloroplasts in large food vacuoles and can temporarily function as mixotrophs. The flagellated stage has a typical mesokaryotic nucleus, and the theca is composed of four membranes, two of which are vesicular and contain thin plates arranged in a Kofoidian series of Po, cp, X, 4′, 1a, 5″, 6c, 4s, 5″′, and 2″″. The plate tabulation is unlike that of any other armored dinoflagellate. Nodules often demark the suture lines underneath the outer membrane, but fixation protocols can influence the detection of plates. Amoeboid benthic stages can be filose to lobose, are thecate, and have a reticulate or spiculate appearance. Amoeboid stages have a eukaryotic nuclear profile and are phagocytic. Cyst stages include a small spherical stage with a honeycomb, reticulate surface and possibly another stage that is elongate and oval to spherical with chrysophyte-like scales that can have long bracts. The species is placed in a new family, Pfiesteriaceae, and the order Dinamoebales is emended.     

KLEPTOPLASTIDY IN THE TOXIC DINOFLAGELLATE PFIESTERIA PISCICIDA (DINOPHYCEAE)

The ichthyotoxic dinoflagellate Pfiesteria piscicida Steidinger et Burkholder has a complex life cycle with several heterotrophic flagellated and amoeboid stages. A prevalent flagellated form, the nontoxic zoospore stage, has a proficient grazing ability, especially on cryptophyte prey. Although P. piscicida zoospores lack the genetic capability to synthesize chloroplasts, they can obtain functional chloroplasts from algal prey (i.e. kleptoplastidy), as demonstrated here with a cryptophyte prey. Zoospores grown with Rhodomonas sp. Karsten CCMP757 (Cryptophyceae) grazed the cryptophyte population to minimal densities. After placing the cultures in near darkness where cryptophyte recovery was restricted and further prey ingestion did not occur, the time-course patterns in growth, prey chloroplast content·zoospore⁻¹, and prey nucleus content·zoospore⁻¹ were followed. Ingested chloroplasts were selectively retained in the dinoflagellate, as indicated by the decline and, ultimately, near absence of cryptophyte nuclei in plastid-containing zoospores. Chloroplasts retained inside P. piscicida cells for at least a week were photosynthetically active, as indicated by starch accumulation and microscope-autoradiographic measurements of bicarbonate uptake. Recognition that P. piscicida can function as a phototroph broadens our perspective of the physiological ecology of the dinoflagellate because it suggests that, at least during part of its life cycle, P. piscicida 's growth and survival might be affected by photoregulation and nutritional control of photosynthesis.     

Grazing activity of Pfiesteria piscicida (Dinophyceae) and susceptibility to ciliate predation vary with toxicity status

Variability has been reported in the toxicity potential of Pfiesteria piscicida that is partly a function of the history of exposure to live fish. Grazing properties of P. piscicida and its susceptibility to ciliate predation were compared in three functional types or toxicity states of this species: actively toxic cultures, cultures with temporary loss of demonstrable toxicity, and cultures with no demonstrable toxicity. Pronounced differences in predator–prey interactions were found between actively toxic cultures and cultures with reduced toxicity. When grown with Rhodomonas sp. (Cryptophyceae) prey, specific growth rates were relatively low in actively toxic cultures under both relatively high and low irradiances. In the cultures with reduced toxicity, prey chloroplast material was apparent in nearly 100% of dinoflagellate cells 3 h after feeding, while chloroplast inclusions were found in <40% of actively toxic cells for ≤16 h (high light) and ≤23 h (low light). These results suggest a relatively high reliance on phagotrophic carbon assimilation and more rapid response to algal prey availability in Pfiesteria cells with lower toxicity. Grazing by two euplotid benthic ciliates (Euplotes vannus and E. woodruffi) on P. piscicida also varied among functional types. Grazing on actively toxic P. piscicida cells did not occur, whereas net positive ingestion rates were calculated for the other prey cultures. These results support concurrent experimental findings that a natural assemblage of microzooplankton displayed lower grazing potential on actively toxic P. piscicida than on cultures with reduced toxicity. In summary, pronounced differences in trophic interactions were found between actively toxic cultures and those with reduced or undetectable toxicity, providing additional evidence of the importance of cellular toxicity in the trophic ecology of Pfiesteria.     

Response of two zooptankton grazers to an ichthyotoxic estuarine dinoflagellate

The dinoflagellate Pfiesteria piscicida (gen. et sp. nov.).a toxic ‘ambush predator’, has been implicated asa causative agent of major fish kills in estuanne ecosystemsof the southeastern USA. Here we report the first experimentaltests of interactions between P.piscicida and estuarine zooplanktonpredators. specifically the rotifer Brachionus plicatilis andthe calanoid copepod Acartia tonsa. Short-term (10 day) exposureof adult B.plicatilis to P.piscicida as a food resource, aloneor in combination with the non-toxic green algae Nannochlorisand Tetraselmis. did not increase rotifer mortality relativeto animals that were given only non-toxic greens Similarly,short-term (3 day) feeding trials using adult A.tonsa indicatedthat the copepods survived equally well on either P.piscicidaor the non-toxic diatom Thalassiosira pseudonana. Copepods giventoxic dinoflagellates exhibited erratic behavior, however, relativeto animals given diatom prey. The fecundity of B.plicatiliswhen fed the toxic dinoflagellate was comparable to or higherthan that of rotifers fed only non-toxic greens We concludethat, on a short-term basis, toxic stages of P.piscicida canbe readily utilized as a nutritional resource by these commonestuarine zooplankters. More long-term effects of P.piscicidaon zooplankton, the potential for toxin bioaccumulation acrosstrophic levels, and the utility of zooplankton as biologicalcontrol agents for this toxic dinoflagellate. remain importantunanswered questions.     

Comparative culture and toxicity studies between the toxic dinoflagellate Pfiesteria piscicida and a morphologically similar cryptoperidiniopsoid dinoflagellate

A series of fish bioassays using cultures of the toxic dinoflagellate, Pfiesteria piscicida and a cryptoperidiniopsoid dinoflagellate indicated various degrees of toxicity for Pfiesteria piscicida and no toxicity by the cryptoperidiniopsoid. P. piscicida maintained toxicity in the presence of live fish, and this toxicity was perpetuated following a series of inoculations to other culture vessels. Differences in the onset and magnitude of the fish deaths occurred, requiring 16 days for the initial fish death when using P. piscicida from a culture that had previously been maintained on algal cells, to kills within hours when using a culture that had recently (previous day) killed fish. Autopsies of moribund fish from the test and control fish bioassays indicated a general lack of bacterial infection, which ensued following death of other autopsied fish. Moreover, bacterial comparisons of waters in the fish bioassay and control fish cultures indicated that similar bacterial concentrations were present. Neither oxygen or ammonia levels were determined to be factors in the fish death. Life stages of a cryptoperidiniopsoid dinoflagellate from Virginia estuaries were also identified, including motile zoospore, gametes, planozygote, amoebae, and cyst stages. The cryptoperidiniopsioid did not initiate fish deaths in bioassays conducted over a 14-week period at zoospore concentrations of ca. 700-800 cells ml(-1). Elemental X-ray analysis of the scales from cysts of this dinoflagellate and P. piscicida indicate that they both contain silicon. Overall, the data from this study demonstrate that the cryptoperidiniopsoid possesses several similar life stages and feeding patterns as P. piscicida, but was not toxic to fish.     

Predator/Prey Interaction between <Emphasis Type="Italic">Pfiesteria piscicida</Emphasis> and <Emphasis Type="Italic">Rhodomonas</Emphasis> Mediated by a Marine Alpha Proteobacterium

The dinoflagellate Pfiesteria piscicida coexists with bacteria in aquatic environments and as such, may interact with them at the physiological level. This study was designed to investigate the influence of bacteria, present in a clonal culture of Pfiesteria piscicida, on the predator/prey relationship of this dinoflagellate with the alga Rhodomonas. A series of replenishment experiments with bacteria isolated from P. piscicida clonal culture and the bacteria-free P. piscicida derived from the same culture were carried out. In the presence of bacteria, the number of P. piscicida increased significantly when incubated with alga Rhodomonas. This enhanced growth was almost entirely due to the increased consumption rate of Rhodomonas by P. piscicida since in bacteria-free (axenic) cultures Rhodomonas were consumed at significantly reduced rates relative to cultures with bacteria. Subsequent replenishment experiments with individual bacterial isolates showed that a single isolate was responsible for the increased predation rate of P. piscicida. The presence or absence of this specific bacterium determined the outcome of the interaction between P. piscicida and Rhodomonas. Partial sequence analysis of the 16S rDNA of this isolate indicated that it was a novel marine alpha proteobacterium with sequence similarities to a Roseobacter sp. and a bacterium recently isolated from a toxic dinoflagellate Alexandrium sp.     

Direct uptake of nitrogen by Pfiesteria piscicida and Pfiesteria shumwayae, and nitrogen nutritional preferences

The rates of uptake of a range of forms of nitrogenous nutrients were measured in cultures of Pfiesteria piscicida and Pfiesteria shumwayae maintained at varying physiological states. The measured rates of dissolved N uptake under some conditions approached the rates of N uptake that are achieved through phagotrophy. Rates of dissolved N uptake by P. piscicida contributed <10% of the cellular N of flagellated cells feeding on algae, but were equal to or greater than phagotrophic N acquisition in cells recently removed from fish cultures. Specific N uptake rates (V, h⁻¹) were higher for cells that were maintained on algal prey for long periods (months) than those that were grown with live fish. However, rates of N uptake on a cellular basis for cells grown on or recently removed from fish were comparable to those maintained on algal prey, likely reflecting differences in the sizes of cells of different physiological condition. Preferences for form of N generally followed a decreasing trend of amino acids > urea > NH₄⁺ > NO₃⁻. Nitrate consistently was not a preferred form of N. Although Pfiesteria spp. are often found in eutrophic environments, the relationship between Pfiesteria spp. and nutrient availability is likely to be primarily indirect, mediated through the production of various prey on which Pfiesteria spp. feed. These findings also confirm, however, that when dissolved N concentrations are elevated, they can contribute to the supplemental nutrition of these cells, and thus may provide a significant source of N to Pfiesteria spp. in nature.     

REPLY TO COMMENT ON THE LIFE CYCLE AND TOXICITY OF PFIESTERIA PISCICIDA REVISITED1

Free‐living, marine dinoflagellates are typified by a well‐defined, haplontic life cycle with relatively few stages. The most unusual departure from this life cycle is one reported for the heterotrophic dinoflagellate Pfiesteria piscicida Steidinger et Burkholder. This species is alleged to have at least 24 life cycle stages including amoebae and a chrysophyte‐like cyst form ( Burkholder et al. 1992 , Burkholder and Glasgow 1997a ) not previously known in free‐living marine dinoflagellates. Litaker et al. (2002) redescribed the life cycle of P. piscicida from single‐cell isolates and found only life cycle stages typical of free‐living marine dinoflagellates. The discrepancy between these observations and the life cycle reported in the literature prompted a rigorous study to resolve the life cycle of P. piscicida. Burkholder and Glasgow (2002) took exception to this study, arguing that Litaker et al. (2002) misunderstood the life cycle of P. piscicida and ignored recent publications. We present a rebuttal of their criticisms and suggest a simple way to resolve the discrepancies in the P. piscicida life cycle.     

A bacterium that inhibits the growth of Pfiesteria piscicida and other dinoflagellates

Toxic dinoflagellate blooms have increased in estuaries of the east coast of the United States in recent years, and the discovery of Pfiesteria piscicida has brought renewed attention to the problem of harmful algal blooms (HAB) in general. Many bacteria and viruses have been isolated that have algicidal or algistatic effects on phytoplankton, including HAB species. Twenty-two bacterial isolates from the Delaware Inland Bays were screened for algicidal activity. One isolate (Shewanella IRI-160) had a growth-inhibiting effect on all three dinoflagellate species tested, including P. piscicida (potentially toxic zoospores), Prorocentrum minimum, and Gyrodinium uncatenum. This bacterium did not have a negative effect on the growth of any of the other four common estuarine non-dinoflagellate species tested, and in fact had a slight stimulatory effect on a diatom, a prasinophyte, a cryptophyte, and a raphidophyte. Shewanella IRI-160 is the first non-microzooplankton example of a microbe with the ability to control and inhibit the growth of P. piscicida, suggesting that bacteria in the natural environment could play a role in controlling the growth and abundance of P. piscicida and other dinoflagellates. Such bacteria could also potentially be used as management tools to prevent the proliferation of potentially harmful dinoflagellates in estuaries and coastal waters.     

Effect of small-scale shear on grazing and growth of the dinoflagellate Pfiesteria piscicida

Low turbulence environments have been hypothesized to be necessary for toxic outbreaks of the heterotrophic/mixotrophic dinoflagellate Pfiesteria piscicida. A toxic Pfiesteria outbreak occurred in the shallow waters (the flats) of the lower Pocomoke River, MD, USA, in 1997. During August 1999 and May and August 2000, data were collected with a Shallow Water ADV Turbulence Tripod (SWATT) to estimate turbulent shear at a location monitored for Pfiesteria on the flats. Approximately 78% of the observed shears were ≤1 s⁻¹ and 98% were ≤2 s⁻¹. Densities of Pfiesteria-like dinoflagellates were low in the Pocomoke River during 1999 and 2000 and no toxic outbreaks occurred. In the laboratory, Couette cylinders were used to determine the effect of small-scale shear on feeding and growth of cultured P. piscicida growing on cryptophyte prey. Shear of 1 s⁻¹ had little or no effect on feeding or growth but 3 s⁻¹ reduced feeding and survival/growth of Pfiesteria zoospores. Small, rapidly dividing Pfiesteria cells were most susceptible to negative effects of shear. Pfiesteria appears to be more sensitive to small-scale shear than are most autotrophic dinoflagellates. However, in the lower Pocomoke River, turbulent shear is rarely high enough to inhibit the growth of non-toxic Pfiesteria zoospores. Toxic functional types (TOX-A and TOX-B) may be more sensitive to small-scale shear; it will be important to determine the responses of these types of Pfiesteria in order to predict the affects of wind and tidal mixing on toxic outbreaks.     

The amoebae of Idionectes vortex (Cutosea,Amoebozoa): Motility,cytoskeleton architecture and extracellular scales

The Cutosea represent a deep-branching lineage within the phylum Amoebozoa that is still relatively poorly explored. Currently, there are four cutosean representatives known – the monotypic genera Armaparvus, Idionectes, Sapocribrum, and Squamamoeba – with marked genetic distances. Idionectes vortex is the deepest-branching species and differs markedly from the other Cutosea in ecology, life history, and most importantly, in its ability to form a flagellated swarmer with an exceptional swimming mechanism. As far as we know, the other Cutosea lack flagella and rather represent small, marine amoebae with a characteristic cell coat. The present paper focuses on the amoeboid life history stage of the algivorous amoeboflagellate Idionectes vortex to provide data for a first in-depth comparison with other Cutosea and to document structural specialties. The amoeboid stage of Idionectes is mainly associated with the specific feeding process, that is, the interaction with algal prey cells and phagocytosis of protoplast material. Yet, the present data from time-lapse microscopy, cytochemical stainings, and electron microscopy demonstrate clear similarities with the other cutosean species concerning amoeboid locomotion and cell coat ultrastructure. Furthermore, Idionectes amoebae exhibit a well-developed microtubular cytoskeleton, and an unusual basal apparatus that seems to undergo marked changes during the life history of this exceptional amoebozoan.     

Feeding preferences and grazing rates of Pfiesteria piscicida and a cryptoperidiniopsoid preying on fish blood cells and algal prey

The grazing rates and feeding preferences of the dinoflagellates Pfiesteria piscicida and a cryptoperidiniopsoid on the alga Rhodomonas sp. and fish blood cells were calculated at different ratios of the two food types and at different total food densities. Data from 6 h grazing periods within microcosms were used to calculate grazing rates. Grazing rates of both dinoflagellates increased linearly with an increased ratio of blood cells to Rhodomonas, and P. piscicida had a higher maximum grazing rate than the cryptoperidiniopsoid. The grazing rate of P. piscicida on Rhodomonas also increased with increased Rhodomonas densities relative to the blood cells, but increased densities of Rhodomonas did not increase the grazing rate of the cryptoperidiniopsoid, suggesting a lower feeding threshold for this species. Both dinoflagellates demonstrated a preference for fish blood cells over Rhodomonas cells, with no significant difference in the index of preference between the two species. Total food abundance affected the degree of preference differently for each dinoflagellate species. A higher index of feeding preference was attained by P. piscicida when resource levels were high, while the cryptoperidiniopsoid did not show this response. A preference for fish blood cells occurred at all food ratios for both dinoflagellates, including when blood cells were scarce relative to the alternate food type (15% of total available food). These results suggest that these strains of P. piscicida and the cryptoperidiniopsoid share similar feeding preferences for the prey types tested, although cryptoperidiniopsoids have not been associated with fish kills.     

Comparative toxicity of Pfiesteria spp., prolonging toxicity of P. piscicida in culture and evaluation of toxin(s) stability

Toxicity of Pfiesteria piscicida (strain CAAE #2200) in the presence of fish (juvenile hybrid tilapia, Oreochromis sp., total length 3–6 cm) has been maintained in the laboratory for 19 months by serial transfer of toxic cells using a modified maintenance protocol. Toxicity was re-induced when toxin-producing P. piscicida cells were separated from fish and cultured on algal prey for 50 days and then re-introduced to new tanks containing fish. We confirmed toxicity in a strain of P. shumwayae (strain CAAE #101272). Toxicity to fish was demonstrated in culture filtrates (0.2 μm) derived from cultures of both Pfiesteria spp., however, it was markedly reduced in comparison to unfiltered water. Filtrates retained toxic activity when stored at −20 °C for up to 6 months. Toxicity to fish was retained when filtrates were held at room temperature for 48 h, at 70 °C for 30 min or at 88–92 °C for 2 h. P. piscicida killed all finfish species tested. Grass shrimp (Paleomonetes pugio; adult 2–3 cm), blue crab (Callinectes sapidus; juvenile 4–7 cm) and brine shrimp (Artemia sp.; 18–24 h post-hatch) were unaffected by concentrations of toxin(s) that killed juvenile tilapia in 4–24 h. Ichthyotoxic activity of filtrates from fish-killing cultures and stability of the toxic activity were similar among P. piscicida and P. shumwayae. These results confirm previously reported observations on toxicity of P. piscicidaand P. shumwayae to finfish. We have maintained toxicity in the laboratory for longer periods than have previously been routinely achieved, and we have demonstrated that the toxic activity is heat stable. In contrast to previous studies with other toxic P. piscicida strains, we did not observe toxic activity to blue crabs or other crustaceans.     

The Life Cycle of Cryptochlora perforans (Chlorarachniophyta)

Observations on the behaviour of different life cycle stages, gamete fusions, and measurements of nuclear DNA contents in Cryptochlora perforans resulted in a first concept concerning life histories in Chlorarachniophyta: the life cycle of Cr. perforans is diplohaplontic (gamete fusion with karyogamy - mitosis - meiosis - mitosis). In the haploid as well as in the diploid life cycle phases amoeboid and coccoid stages occur. The isomorphic gametes are modified amoebae frequently without filopodia. Only haploid flagellate stages are known representing mito- or meiozoospores. Diploid coccoid stages have a granular cytoplasmic structure and may be somewhat larger than haploid ones. Nevertheless, positive identification of haploid (gametophytic) and diploid (sporophytic) stages is only possible on the basis of nuclear DNA contents.     

Distribution of Pfiesteria piscicida cyst populations in sediments of the Delaware Inland Bays, USA

The toxic dinoflagellate, Pfiesteria piscicida, is a common constituent of the phytoplankton community in the Delaware Inland Bays, USA. In this study, molecular methods were used to investigate the distributions of benthic stages (cysts) of P. piscicida in sediment cores from the Delaware Inland Bays. Cores from 35 sites were partitioned into nephloid and anoxic layers and analyzed for P. piscicida by nested amplification of the 18S rDNA gene using P. piscicida-specific primers. The presence of inhibitory substances in the PCR reaction was evaluated by inclusion of an exogenous control DNA in the extraction buffer, thus eliminating samples that may yield false-negative results. Our results indicate a patchy distribution of P. piscicida in sediments of the Delaware Inland Bays, with distinct differences between each of the three bays. Overall, P. piscicida was found more frequently in sediments from Rehoboth Bay compared to Indian River and Little Assawoman Bays. These differences suggest (i) that populations of P. piscicida may be more widely distributed in Rehoboth Bay, (ii) that populations of P. piscicida may have been introduced to Rehoboth Bay at an earlier time, (iii) that past blooms of P. piscicida in Rehoboth Bay estuaries may have seeded the sediments with higher numbers of cysts, and/or (iv) that Rehoboth Bay sediments may be more resistant to clearing due to storm turbulence.     

Predator–prey interactions between a planktonic ciliate Strombidium sp. (Ciliophora, Oligotrichida) and the dinoflagellate Pfiesteria piscicida (Dinamoebiales, Pyrrophyta)

The functional response of a planktonic ciliate, Strombidium sp. feeding on the dinoflagellate Pfiesteria piscicida non-toxic zoospores (NTZ) was experimentally studied with four different prey concentrations (43–3153 cells ml⁻¹). Data from direct observations (NTZ inside individual Strombidium sp.) was used to calculate predator–prey specific ingestion and clearance rates. The ingestion rates varied between 0.68 and 14.26 NTZ ind⁻¹ h⁻¹, and with the predator–prey specific handling time of 2.83 min the U_max was 21.18 NTZ ind⁻¹ h⁻¹. The increase in the prey concentration between approximately 700 and 3000 NTZ ml⁻¹ did not increase the uptake of prey, and at the lowest Pfiesteria NTZ concentrations the feeding efficiency of Strombidium sp. was lowered, possibly indicating a situation of threshold feeding. When data from direct observations of ingested Pfiesteria NTZ were compared with values of total NTZ loss from the experimental water during the experiment, ingestion was found to represent only a fraction of the total NTZ loss in the presence of ciliates. This discrepancy was concluded to be due to other grazer related factors than actual ciliate grazing. The control of the initial growth of Pfiesteria community, in a pre-bloom situation, would require only a small ciliate abundance (less than 5 ml⁻¹), but when the Pfiesteria NTZ are scarce, relatively more ciliates are needed to limit the population growth of the dinoflagellate community because of the apparent feeding threshold. It is concluded that the formation of non-toxic P. piscicida blooms require periods of low grazing pressure or a means to escape grazing.