Abstract

Zooplankton can consume toxic Alexandrium spp. dinoflagellates in the Gulf of Maine and retain paralytic shellfish poisoning (PSP) toxins, potentially acting as toxin vectors. We performed experiments to determine toxin budgets for common species of copepods (Acartia hudsonica, Eurytemora herdmani, Centropages hamatus) feeding on toxic Alexandrium fundyense, offered as monocultures or in mixtures of algal prey, by comparing calculated toxin ingestion rates and toxin content of copepod body tissue and fecal pellets. When fed monocultures, both copepod tissue and fecal pellet fractions accounted for ≤ 5% each of the calculated ingested toxin, and thus by difference ≥ 90% was lost as a dissolved fraction into the seawater medium. The presence of alternative food did not significantly alter the efficiency of toxin retention. Sloppy feeding or regurgitation are probable mechanisms for release of toxin to sea water. Experiments using varying concentrations of A. fundyense and alternativenon-toxic species did not show significant effects of cell concentration on toxin retention efficiency. Total toxin retained and efficiency of retention varied among copepod species. Toxin profiles (% molar composition) of dinoflagellates, copepod tissues and fecal pellets differed slightly, suggesting some metabolic transformation. Because of their low retention efficiency, copepod grazers can effectively disperse PSP toxins produced by Alexandrium spp. into the environment, where they are much less likely to be harmful—zooplankton act as a sink for PSP toxins. Nevertheless, sufficient toxin body burdens are attained to contribute to propagation of PSP toxins to other trophic levels.

INTRODUCTION

Paralytic shellfish poisoning (PSP) toxins are produced by dinoflagellates of the genus Alexandrium, which form blooms in the Gulf of Maine (Anderson et al., 1994). It is well known that suspension-feeding shellfish can consume toxic cells of Alexandrium spp. and accumulate PSP toxins, and that such contaminated shellfish are a threat to public health and result in economic loss to the fishing and aquaculture industries (Shumway et al., 1988). Zooplankton are also major consumers of phytoplankton, including Alexandrium spp.; the presence of PSP toxins in wild zooplankton populations of the Gulf of Maine has been documented in several studies (White, 1979, 1980, 1984). These studies have shown that mesozooplankton can attain body burdens of these toxins that are inimical or even fatal to vertebrate zooplanktivores such as clupeid fishes and whales. Larval fish are even more susceptible to adverse effects from consumption of zooplankton contaminated by PSP toxins (White et al., 1989; Robineau et al., 1991a,b).

Nevertheless, research suggests that only a fraction of the total toxin from Alexandrium spp. cells acquired during feeding activity is retained in the tissues ofzooplankton grazers. White conducted toxin accumulation studies with adult copepods, Acartia clausi (= A. hudsonica), and barnacle nauplii, Balanus sp. (= Semibalanus sp.) at high densities (10–13 individuals ml−1), fed Alexandrium fundyense (= Gonyaulax excavata) at ∼ 3 × 103 cells ml−1 (White, 1981). At such high densities of zooplankton and toxic cells, both grazer species accumulated high toxin levels [19–54 μg saxitoxin equivalents (μgSTXeq) g−1 wet weight] within 6 h. Toxin retained in copepod tissues, expressed as a percentage of total toxin ingested, ranged from ∼10% for A. clausi to ∼32% for Balanus sp. White used a modification of the Association of Official Analytical Chemists (AOAC) mouse bioassay (a relatively crude and low-sensitivity technique) for PSP toxin determinations in these early experiments (AOAC, 1980; White, 1981). Teegarden and Cembella measured consumption of two clonal cultures of Alexandrium spp. by the copepod species Acartia tonsa and Eurytemora herdmani, and toxin content of dinoflagellates and copepod tissues (Teegarden and Cembella, 1996). In these experiments, toxin content was analyzed by high-performance liquid chromatography with fluorescence detection (HPLC-FD) (Oshima, 1995a), a method that yields both qualitative and quantitative data on toxin composition. Retention efficiencies of the two copepod species were typically <5% of ingested toxins. Recently Guisande et al. found similar low retention of ingested toxins in the copepod Acartia clausi feeding on low toxicity Alexandrium minutum (Guisande et al., 2002). Low retention of toxins suggests that the bulk of ingested toxin was excreted, perhaps in fecal pellets. This hypothesis was addressed in the present study.

Poor retention of PSP toxins in copepods could also be an experimental artefact of physiological effects on copepod digestion when feeding on unnaturally high concentrations of Alexandrium spp. For example, toxic cells at high concentrations could trigger regurgitation [as documented by (Sykes and Huntley, 1987)], reduced absorption of liberated toxin in the gut, and/or an increased rate of excretion. If, however, cells of toxic Alexandrium spp. constitute only a minor proportion of available food, retention of toxic cells in the gut and absorption of toxins into tissues might be enhanced via dilution with non-toxic food. Analytical constraints have precluded the measurement of PSP toxins in the tissues of copepods feeding on low concentrations of Alexandrium spp. and in fecal pellets, but newer methods of micro-extraction and analysis have made it feasible to detect subnanomolar concentrations. We therefore tested copepods for differences in retention efficiency and total toxin body burden, when fed varying concentrations of toxic Alexandrium fundyense and other species of non-toxic phytoplankton.

The relative molar composition of retained toxins is also of interest because changes in toxin composition affect the potency of retained toxins, and thus the threat posed to zooplanktivores feeding on contaminated zooplankton. Alexandrium spp. typically contain three major classes of PSP toxins: non-sulfated carbamate toxins, 11-hydroxysulfated carbamates, and 21-N-sulfocarbamoyl derivatives (Cembella et al., 1987). The carbamate toxins, saxitoxin (STX) and its N-1-hydroxy derivative neosaxitoxin (NEO) are the most potent PSP toxins. On a molar basis, the 11-hydroxysulfated carbamate toxins, known as gonyautoxins (GTX), are about half as potent as STX. The 21-N-sulfocarbamoyl toxins (B- and C-toxins) are more than an order of magnitude less toxic on a molar basis than carbamate analogues. Biotransformation and chemical conversion of PSP toxins to related derivatives can occur via a number of enzymatic and non-enzymatic reactions, including oxidation, reduction and desulfation, as well as simple epimerization.

In early experiments that used HPLC analysis to track vectoral transfer of PSP toxins from dinoflagellates (Alexandrium tamarense = A. excavatum) through zooplankton to post-larval fish, such as Atlantic cod and winter flounder (Robineau et al., 1991a,b), changes in toxin profile were attributed to zooplankton metabolism or selective elimination of particular toxin analogues (though there is no apparent mechanism to account for the latter). Metabolic transformation of toxins has also been shown to occur in several species of bivalve molluscs (Bricelj et al., 1991; Cembella et al., 1993, 1994; Oshima, 1995b). Virtually all invertebrate species analyzed, including bivalve molluscs, crustaceans, copepods and tintinnids, show some time-dependent modification of toxin profile following ingestion of PSP toxin-producing dinoflagellates (A.D. Cembella, unpublished data).

Using an early HPLC technique, Boyer et al. specifically looked for post-ingestive PSP toxin transformations in copepods, Tigropus californicus, and found elevated levels of the N-1-hydroxy-toxins (GTX2 and GTX3) and the absence of NEO relative to the toxin profile of the ingested dinoflagellates Alexandrium catenella (= Protogonyaulax catenella) within several hours of the initiation of feeding (Boyer et al., 1985). Turriff et al. also found some evidence for transformation in the copepod Calanus finmarchicus, and suggested that C-toxins may be preferentially retained, because NEO and STX were sometimes lacking in copepod samples though present in the dinoflagellate food source (Turriff et al., 1995). Using an improved analytical method, Teegarden and Cembella analyzed toxin transformation in Acartia tonsa and Eurytemora herdmani and detected a more consistent pattern of toxin transformation, characterized by epimerization of N-sulfocarbamoyl toxin C2 to C1, and subsequent desulfation of C1 to the more potent carbamate toxin GTX2 (Teegarden and Cembella, 1996). They suggested that those toxins that do remain in zooplankton tissues could undergo facile reactions that serve to increase their molar toxicity.

In the present study, copepods were exposed to toxic cells of Alexandrium fundyense (Balech), either alone or in combination with non-toxic alternative phytoplankton in a series of experiments to address the following questions. What is the retention efficiency of ingested toxins, and does efficiency vary among grazer species? What is the fate of unassimilated toxin, and what mechanisms might explain observed retention efficiency? Does the presence of alternative food affect toxin retention efficiency, and toxin body burden? Is toxin retention affected by varying cell concentrations of Alexandrium spp.? Is there a general pattern among copepod species of metabolic transformation of ingested toxins?

METHOD

Zooplankton and phytoplankton collection and culture

During blooms of Alexandrium spp. in coastal Gulf of Maine waters (Casco Bay) in the spring of 1998 and 2000, Acartia hudsonica Pinhey was the dominant zooplankter; Eurytemora herdmani Thompson & Scott and Centropages hamatus Krøyer were also commonly found (Teegarden et al., 2001). Therefore these three species were selected for experiments. Zooplankton samples were collected with a 150 μm mesh net from the Damariscotta River estuary, Walpole, Maine, or from Harpswell Sound, Harpswell, Maine. The net was either suspended in the tidal flow, or towed from a small boat, to gently collect live zooplankton. Adult females of the copepod species Acartia hudsonica, Eurytemora herdmani and Centropages hamatus were sorted from the plankton and placed into 22 l containers of 5 μm-filtered sea water, at approximately 17°C and 31 p.s.u., with gentle aeration. Animals were acclimated at 17°C for at least 24 h before experiments, and fed a mixture of non-toxic dinoflagellates [Heterocapsa triquetra (Ehrenberg) Stein clone HT984, Gymnodinium sanguineum Hirasaka clone B4], diatoms [Thalassiosira weissflogii (Grunow) Fryxell and Hasle clone B9TW], and small flagellates [Rhodomonas salina (Wislouch) Hill and Wetherbee clone CCMP 1319]. Alexandrium fundyense Balech clones GTCA 28 (moderate toxicity) and BC1 (high toxicity) were also cultured for use in experiments. All phytoplankton clones were cultured in sea water (∼31 p.s.u.) enriched with f/2–Si (+Si for diatoms) (Guillard, 1975), in an incubator at 17°C with a14 h:10 h light:dark cycle, using cool-white 100 W fluorescent lights. Cultures were transferred every 2 weeks to keep the phytoplankton in exponential growth. Carbon content of the phytoplankton clones was determined by filtering samples onto pre-combusted glass fiber filters and combustion in a Carlo Erba CHN analyzer.

Experimental design

Alexandrium fundyense monoculture experiments

Feeding experiments were performed with adult females of the copepod species Acartia hudsonica, Centropages hamatus and Eurytemora herdmani to measure copepod ingestion of toxic Alexandrium fundyense, to determine the toxin content of copepod and fecal pellet fractions, and to construct toxin budgets. The experimental containers were 1 liter polycarbonate beakers, fitted with removable polycarbonate sleeves with 200 μm mesh bottoms, which were suspended ∼2 cm above the bottom of the container. This arrangement allowed fecal pellets and eggs to sink away from feeding zooplankton.

Experimental cell suspensions were prepared from 0.45 μm-filtered sea water, to which A. fundyense GTCA 28 was added to yield a final concentration of approximately 250 cells ml−1. For each copepod species tested, triplicate experimental and control containers with mesh-bottom sleeves were set up, and 1 liter of experimental cell suspension was placed into each container. Initial samples of 50 copepods of each species (two samples per species) were isolated by micropipette and pooled into vials, then frozen (–80°C) for later extraction and toxin analysis. Copepods (n = 50) were then quickly and gently added to each experimental container, which was then gently mixed to distribute copepods and A. fundyense cells. Subsamples of the initial cell suspension were preserved with Lugol’s iodine solution. Experimental and control containers were placed in a controlled environmental chamber at 17°C, 14 h:10 h light:dark cycle, at 19:00 h, and allowed to feed overnight for 12 h (until 07:00 h). During this time, containers were occasionally mixed gently with a small paddle to redistribute A. fundyense cells and copepods. Visual examination confirmed that A. fundyense cells remained in suspension.

At the end of each experiment, the polycarbonate sleeves were slowly removed, and the copepods were removed to a dish and examined for condition. Copepods were then isolated and pooled into vials, and frozen (–80°C) for subsequent extraction. Experimental cell suspensions were gently passed through a bottom-supported 40 μm mesh to remove fecal pellets, which were rinsed into 10 ml centrifuge tubes. Meshes were checked under a dissecting scope to ensure recovery of fecal material. A 50 ml subsample was taken from the experimental suspension and preserved with Lugol’s solution, and the remainder was concentrated on a 10 μm mesh and rinsed into a dish, and examined for remaining fecal pellets. Any pellets present (very few were found) were isolated with a pipette and added to the 10 ml centrifuge tube. The tubes were then centrifuged at 5000 g for 10 min, the supernatant was removed by aspiration, and the pellet was frozen (–80°C) for later extraction. Control cell suspensions were subjected to the same handling procedures and subsamples were preserved with Lugol’s solution. Cell counts of all replicates were performed with a Coulter® Multisizer particle counter. Ingestion rates were calculated with the Frost equations (Frost, 1972).

Mixed food experiments

Experiments were conducted with mixtures of Alexandrium fundyense clone GTCA 28 and Heterocapsa triquetra clone HT 984 to examine the effects of non-toxic alternative food sources on toxin assimilation efficiency. Experimental, control and initial containers were 500 ml polycarbonate wide-mouth bottles. Initial copepod samples were prepared as above, and cell suspensions were prepared with A. fundyense at ∼250 cells ml−1 andH. triquetra at ∼500 cells ml−1. Subsamples of initial suspensions were preserved with Lugol’s solution. Containers were incubated on a grazing wheel at 1 r.p.m. and 17°C for 12 h, to ensure even distribution of the two dinoflagellate species and grazers (i.e. to prevent possible separation of food types).

Feeding rates were measured in additional experiments with mixtures of four phytoplankton species—Alexandrium fundyense clone BC1, Heterocapsa triquetra HT984, Thalassiosira weissflogii B9TW, and Rhodomonas salina CCMP 1312— over widely varying concentrations of A. fundyense. For each experiment, in four treatments, copepods were offered the Alexandrium clone at approximate concentrations of 5 × 103, 104, 5 × 104, and 105 cells l−1, respectively. Within each treatment, the three alternative food species were offered at roughly equal carbon concentrations, and when combined with the A. fundyense clone, total food concentration was ∼300 μg C l−1. As A. fundyense concentrations increased from one treatment to the next, the contribution of the alternative phytoplankton was decreased to maintain ∼300 μg C l−1 for each treatment. Triplicate initial and control containers were prepared. For each treatment, 15–20 adult females were added to each of the 500 ml experimental containers. Containers were placed on a grazing wheel as before, and incubated for 18–24 h. Copepods from the experimental containers were later recovered for toxin analysis (triplicate samples).

At the end of all mixed food experiments, copepods were removed with a 200 μm mesh, and isolated and pooled into vials and frozen (–80°C) for later extraction. For solutions with <100 cells ml−1 of Alexandrium fundyense, 250 ml were concentrated five-fold by reverse filtration with a 10 μm mesh. Fifty millilitre subsamples of neat and concentrated solutions from experimental and control containers were preserved with Lugol’s solution. Fecal pellets could not be reliably isolated in these experiments, therefore fecal pellet toxin content was not determined. Preserved solutions were counted using Sedgwick–Rafter chambers or settling chambers at 100× with a phase-contrast microscope, and cell counts were converted to clearance and ingestion rates using the equations of Frost (Frost, 1972). Feeding selectivity in mixed food experiments was quantified using the selectivity coefficient and electivity index of Vanderploeg and Scavia (Vanderploeg and Scavia, 1979). The selection coefficients Wi for each phytoplankton species i were calculated from clearance (filtration) rates by 

\[\mathit{W}_{\mathit{i}}\ \mathit{=\ F}_{\mathit{i}}/S\mathit{F}_{\mathit{i}}\]
where Fi is the clearance rate of species i, and ΣFi is the sum of clearance rates on all species. The electivity index Ei* of grazers for each phytoplankton species was then calculated by 
\[\mathit{E}_{\mathit{i}}*\ =\ [\mathit{W}_{\mathit{i}}\ {\mbox{--}}\ (1/\mathit{n})]/[\mathit{W}_{\mathit{i}}\ +\ (1/\mathit{n})]\]
where n is the total number of phytoplankton species in the food complex. This value can theoretically vary between –1 and 1, where 0 signifies no electivity (no selective grazing), negative numbers correspond to negative selection (avoidance), and positive numbers correspond to selection for species in the food complex.

For all toxin budget experiments, toxin ingested over time was calculated from ingestion rate (cells copepod−1 h−1) × cell toxicity (pgSTXeq cell−1) × time (h). Ingested toxin was compared with measured toxin in copepod tissues (pg or ng STXeq copepod−1) and fecal pellet fraction (reported as total ngSTXeq, for all material collected from a container), to construct gross 12–24 h toxin budgets and retention efficiency.

Sample preparation and toxin analysis

Cell counts of Alexandrium fundyense clones from each experiment were performed with a Coulter® Multisizer, and triplicate samples of 1 × 105 cells were concentrated with a 10 μm mesh. Cells were rinsed into 15 ml centrifuge tubes, and the mesh was examined at 50× magnification to ensure complete cell recovery. Cells were then centrifuged at 5000 g for 15 min, and the supernatant was aspirated. The pellet was then frozen (–80°C) forlater extraction.

Alexandrium fundyense samples were extracted by adding 0.5 ml of 0.1 M acetic acid, and sonicating with a Fisher® model 60 sonic dismembrator equipped with a microtip until cells were completely disrupted (as confirmed by microscopic examination). Samples were kept cool in an ice bath or cold block during processing. Fecal pellet samples were processed by the same method.

Copepod samples (isolation described above) were lyophilized and stored frozen (–80°C) until extraction. Lyophilized samples were extracted in 300 μl of 0.1 M acetic acid by a freeze–thaw cycle followed by 30 s sonication(25 W) at 50% pulse–duty cycle with a stepped microtip.

Sonicated samples were centrifuged (at 4°C) in microcentrifuge vials at 10 000 g for 10 min. The supernatant of each vial was filtered through a 0.45 μm membrane in a 0.5 ml spin-cartridge (Millipore® Ultra-free MC) at 2000 g until the filter was dry.

The filtrate was analyzed for PSP toxins by HPLC-FD according to minor modifications of the post-column derivatization method described in detail for copepods (Teegarden and Cembella, 1996) and Alexandrium cells (Parkhill and Cembella, 1999). Three separate isocratic elutions were employed to separate the complete spectrum of PSP toxins at a flow rate of 0.8 ml min−1. Toxins were resolved by reverse-phase ion-pair chromatography using silica-base columns: Betabasic C-8 (4.6 × 250 mm, Keystone Scientific, Bellefonte, PA) for theSTX, decarbamoyl STX group (STX, NEO, dcSTX, dcNEO) and N-sulfocarbamoyl C-toxins (C1–C4), and Zorbax SB-C8 (4.6 × 250 mm, Hewlett-Packard, Palo Alto, CA) for the gonyautoxins (GTX1–GTX6). Toxins were oxidized to fluorescent derivatives using 0.05 M periodic acid (pH 7.8) and neutralized with 0.75 M nitric acid. For quantification, duplicate injections of 20 μl of extract were compared with external toxin standards (PSP-1C) provided by the Certified Reference Material Program (CRMP) of the Institute for Marine Biosciences, NRC, Halifax, Canada. Toxin concentrations (μmol l−1) were converted by the formula given in Parkhill and Cembella (Parkhill and Cembella, 1999) to toxicity units (STX equivalents cell−1 for Alexandrium and STX equivalents individual for copepods−1) using specific toxicity conversion factors (in mouse units μmol−1) provided in Oshima (Oshima, 1995a).

RESULTS

Alexandrium fundyense clones GTCA 28 and BC 1 remained moderately and highly toxic, respectively, throughout the experiments, with little variation in toxicity or molar composition within clones. Toxicity of GTCA 28 clones was 25.5 ± 4.3 pgSTXeq cell−1 in monoculture experiments, and 26.9 ± 3.5 pgSTXeq cell−1 in mixed food experiments; the BC 1 clone used in multi-species, varying concentration experiments measured 84.2 ±9.5 pgSTXeq cell−1 (mean ± SD, triplicate samples). In all experiments, copepods accumulated measurable toxin body burdens; a typical HPLC-FD chromatogram of an extract of Centropages hamatus copepods is depicted in Figure 1.

Toxin budget experiments

Toxin ingestion (calculated from Alexandrium fundyense cell ingestion), body burden, and gross retention efficiency varied among copepod species (Table I). Acartia hudsonica ingested low amounts of toxic A. fundyense (Table II), and retained very little of the ‘ingested’ toxin (<1%, Table I). A larger proportion was released in fecal pellets (3–6%), but most of the calculated ingested toxin was not recovered (> 90%), and was probably released as dissolved toxin after Alexandrium cell breakage (Table I). Centropages hamatus also ingested low amounts of A. fundyense (Table II), but retained a higher percentage of calculated ingested toxin in its tissues. Unrecovered (presumably dissolved) toxin was still approximately 85% (Table I). Eurytemora herdmani ingested more A. fundyense cells thanthe other two species [Table II, analysis of variance (ANOVA) P < 0.001], and accumulated a moderate body burden of toxin (Table I). The fecal pellet fraction from E. herdmani contained a relatively high amount of total toxin, but was still only a minor fraction of the total calculated for ingested toxin (3.3%). The relative proportions of toxin in E. herdmani copepod tissues, in fecal pellets, and in the dissolved fraction were not significantly different (ANOVA) from those of C. hamatus (Table I).

There were slight differences between toxin profiles of copepod tissue, fecal material, and Alexandrium fundyense fractions (Figure 2). Copepod tissues typically had a greater contribution of C1 toxin than dinoflagellates, probably owing to facile epimerization of C2 to C1 toxin. Copepods retained GTX3 and GTX4 in roughly similar proportions to that in the dinoflagellate. Tissues of C. hamatus and E. herdmani (but not A. hudsonica) contained a higher percentage of GTX1 than A. fundyense cells, probably because of non-enzymatic epimerization of GTX4 to GTX1. Since these epimers have similar molar toxicities, such reactions do not result in substantial changes in total toxic potential of copepod guts or other tissues. Traces of dcGTX3, not usually present in the dinoflagellate strain, were sometimes detected in copepods, but except for E. herdmani, both NEO and STX found in the dinoflagellate were below the detection limit in copepods.

The toxin profiles (in % molar composition) of fecal pellet fractions of all three copepod species were slightly different than both the A. fundyense food source and the copepod tissues. Fecal pellet C1:C2 ratios were similar to A. fundyense, suggesting probable excretion of intact cells and/or partially digested dinoflagellate material, but changes in other toxins (GTX1:GTX4 ratio, absence of dcGTX3, NEO) suggest partial post-ingestive modification that was somewhat less than in copepod tissues. It should be noted that the latter two toxins have a relatively weak molar fluorescence yield in the fluorescence HPLC method. The molar fluorescence yield of NEO when subjected to post-column oxidation in the fluorescence HPLC method is among the lowest of the PSP toxins, and it is therefore difficult to detect when present in low amounts.

Mixed food experiments

The presence of an abundant alternative food source did not significantly alter the efficiency of toxin retention (Table III). Despite roughly equal concentrations (μg Cl−1) of Heterocapsa triquetra offered with toxic A. fundyense, toxin retention over 12 h was <4% for all copepod species tested. Since toxin content of fecal pellets was not measured in this set of experiments, the remaining toxins were placed in the unaccounted category. Both total ingestion (ng C copepod−1 h−1, Table II) and toxin ingestion (ng STXeq copepod−1 h−1, Table III), inferred from disappearance of cells, was greater in mixed food assemblages than in monocultures. This may have been aided by the use of a grazing wheel (not used in monoculture budget experiments), which may have more evenly distributed the food, thus preventing sinking or patch formation. Despite greater calculated ingestion rates and higher total body burdens of toxin, retention of toxin was not dramatically higher (Table III), differing among species by a factor of 2.5, with low efficiency. Feeding selectivity was not consistent (Table II); electivity index (Ei) determinations indicate slight relative avoidance of toxic A. fundyense by Acartia hudsonica and Eurytemora herdmani, but not by Centropages hamatus.

Experiments with varying concentrations of toxicA. fundyense (from 5 × 103 to 100 × 103 cells l−1), along with other non-toxic phytoplankton, showed no dramatic effect of concentration on efficiency of toxin retention (Table IV), with the possible exception of Centropages hamatus. As in other experiments, Acartia hudsonica retained ≤1% of calculated ingested toxin, at all cell concentrations. Eurytemora herdmani retained a greater percentage of toxins, but still only ∼0.5–2% of calculated ingested toxin. While in some cases results suggested that retention efficiency rose with increasing A. fundyense cell concentration, differences were not significant (e.g. Eurytemora herdmani, ANOVA P = 0.056; n = 3). Centropages hamatus did display relatively high average retention efficiency (mean 24.9%; n = 3) at the lowest concentration of A. fundyense cells and mean retention efficiency values decreased as A. fundyense concentrations increased, but retention was highly variable, and differences between treatments were not significant (ANOVA P = 0.16; n = 3). Nevertheless, C. hamatus had the highest retention efficiencies of all copepod species in each treatment, consistent with the other series of experiments.

Copepods changed their feeding in response to varying concentrations of A. fundyense cells. Clearance rates of copepods on A. fundyense across treatments are depicted in Figure 3, and ingestion rates and electivity indices (Ei) for all food items in each treatment are given in Table V. As the contribution of A. fundyense increased from ∼2% (Treatment 1) to ∼65% (Treatment 4) of the total available food (as carbon), copepods generally avoided A. fundyense more strongly (Ei values in Table V). Avoidance was not universal; for each copepod species, actual positive selection of A. fundyense was apparent in the first two treatments, with concentrations of ≤10 cells ml−1. Centropages hamatus appeared to ingest A. fundyense selectively at all concentrations (Ei > 0), however, the strength of that selection decreased with increasing A. fundyense concentration, and the total carbon consumption of all prey declined substantially as A. fundyense became more dominant in the prey field (Table V). Acartia hudsonica and Eurytemora herdmani displayed avoidance (Ei < 0) when A. fundyense concentration reached 50 cells ml−1, and E. herdmani avoided A. fundyense more strongly at the highest concentration, whereas A. hudsonica remained roughly neutral at higher concentrations (Table V).

Clearance rates of copepods feeding on A. fundyense decreased with increasing cell concentration (Figure 3). While decreases are expected when feeding rates become saturated (Frost, 1972), the decreases were not a result of saturation or maximal ingestion. Electivity index (Ei) determinations indicate that at high concentrations A. fundyense was not generally consumed in proportion to its abundance (except for Centropages hamatus, Figure 3, Table V), and total carbon consumed in treatments with A. fundyense ≥50 cells ml−1 was always lower than carbon consumed in treatments with A. fundyense ≤10 cells ml−1 (Table V). Electivity index values (Table V) indicate increased reliance on non-dinoflagellate food as A. fundyense concentrations increased (e.g. Eurytemora herdmani and Centropages hamatus feeding on Thalassiosira weissflogii, Acartia hudsonica and C. hamatus feeding on Rhodomonas salina).

DISCUSSION

Toxin retention

Experimental results indicated that toxins were not efficiently assimilated into the tissues of copepods that had fed on Alexandrium fundyense, whether or not alternative food was present. The highest mean gross retention efficiency (% of ingested toxin retained) was 24.9%(± 19.7%) for Centropages hamatus (Table IV), but this value was highly variable, and virtually all other values of retention efficiency in every other experiment were between 0.1% and 8%. Values of total toxin retained (ngSTXeq copepod−1) and retention efficiency (%) both corresponded well with PSP toxin body burden and retention efficiency of Eurytmora herdmani and Acartia tonsa measured in a previous study (Teegarden and Cembella, 1996) that used Alexandrium spp. clones of similar toxicity. As in that previous study, even with low toxin retention efficiencies, the total toxin body burden in copepods was sufficient to pose a substantial risk of morbidity or mortality to other components of the marine food web through vectoral transfer. Using an average body burden of 0.5 ngSTXeq copepod−1 and an average wet weight of 100 μg copepod−1, 1 kg of zooplankton tissue from the Gulf of Mainewould contain 5 × 103 μgSTXeq [for comparison, the accepted safe limit for human consumption of shellfish is ≤800 μgSTXeq kg−1 total soft tissue (Shumway et al., 1988)]. The body toxin burdens measured in this study also correspond well with those measured in field-collected samples of zooplankton from the Gulf of Maine [(Campbell et al., 2000); R. Campbell et al., unpublished results], suggesting that they may represent an equilibrium body burden for copepods chronically exposed to cells of Alexandrium spp. in the Gulf of Maine.

Nevertheless, very low retention efficiencies argue that the magnitude of this toxin threat is not as great as one might expect. In general, there was a trend in toxin retention efficiency among species, such that Centropages hamatus > Eurytemora herdmani > Acartia hudsonica (Tables I, III and IV). The values were generally low enough, however, that any differences among species of copepod were insignificant compared with the calculated amount of toxin ingested, but not retained, in copepod tissues. The unassimilated toxin was not released primarily in the form of feces. Although only the first set of experiments, in which fecal pellet fractions were isolated, yielded reliable results, those results indicate that only 3.5–7.5% of the ingested toxin could be accounted for in the feces. Thus 70–90% of ingested toxin remains unaccounted for.

Toxin retention efficiency values of copepods are quite different from those published for bivalve shellfish. For example, Bricelj et al. reported a PSP toxin uptake (retention) efficiency of 78% for the blue mussel Mytilus edulis, which was higher than the concurrently measured assimilation of organic matter (estimated 60–64%) (Bricelj et al., 1990). In another study, the hard clam Mercenaria mercenaria exhibited a toxin retention efficiency of 35–40%, and organic matter assimilation of ∼60% (Bricelj et al., 1991). Bricelj et al. also reported moderate to high values of toxin ingestion and retention in the clams Mya arenaria and Spisula solidissima, and suggested that retention efficiency in bivalves may depend on the specific toxicity of Alexandrium cells ingested, and whether or not alternative food is concurrently available (Bricelj et al., 1996).

Clearly, species of bivalve molluscs have much higher toxin retention and assimilation efficiencies than do copepods. What is happening to the ‘ingested’ toxin that does not end up in copepod tissues or fecal pellets? Possible explanations are release as liquid excreta, metabolic degradation of toxins, or release via mechanisms other than excretion, such as sloppy feeding or regurgitation.

Recently Guisande et al. also found low assimilation efficiency of PSP toxins in tissues and fecal pellets by the copepod Acartia clausi feeding on Alexandrium minitum, and suggested detoxification and excretion of dissolved toxins as the mechanisms of toxin loss (Guisande et al., 2002). Excretion of nitrogenous soluble waste can take place via maxillary glands and cephalic nephrocytes (Le Borgne, 1986), but such excretion is principally or almost exclusively in the form of NH4+, and such a mechanism is not likely to be specially adapted to the removal of PSP toxins, nor is it likely to account for a 90% loss of an ingested substance. Metabolic detoxification is also an unlikely alternative. First, one would expect to see time-dependent increases in toxin degradation products in copepodtissues (e.g. metabolism of C-toxins to gonyautoxins and ultimately to STX), and generally higher levels of toxins in tissues as they were accumulated and subsequently detoxified. In this study, a previous study measuring time-dependent toxin accumulation (Teegarden and Cembella, 1996), field-collected samples (Campbell et al., 2000), and the study of Guisande et al. (Guisande et al., 2002), neither of these results has been observed. Second, such a mechanism would require specialized physiological adaptations in animals that have simple digestive systems, and that are exposed only intermittently to toxic food, usually mixed with non-toxic alternative food—probably not a sufficiently strong selective force to favor adaptive development of sophisticated detoxification enzyme systems. Furthermore, the toxin profile transformations that are observed are mostly facile epimerizations, not indicative of metabolic degradation.

There is evidence from this study and the literature to support the hypothesis that large quantities of an ‘ingested’ substance may be released by pathways other than excretion or detoxification. A re-examination of carbon assimilation data from previous work (Teegarden, 1999) suggests that not only toxin assimilation, but also carbon assimilation, are probably low for copepods feeding on toxic Alexandrium cells. In experiments where ingestion rates and changes in body carbon content of copepods feeding on monoculture diets of toxic and non-toxic Alexandrium spp. were measured (Teegarden, 1999), Acartia tonsa, Centropages hamatus and Eurytmora herdmani all fed well on non-toxic cells of Alexandrium tamarense isolate CCMP 115. These copepod species all exhibited an increase of total body carbon over 24 h (28%, 24% and 14%, respectively). When these copepods were fed on toxic A. fundyense GTCA 28, they either had no significant gains in body carbon over 24 h or lost significant body weight, even though the total carbon ‘ingested’ (32–63% body weight day−1) was not much lower than the values for ingestion of A. tamarense CCMP 115. This indicates a failure to assimilate ‘ingested’ material from the toxic dinoflagellate strain.

Dutz studied egg production and growth efficiency of the copepod Acartia clausi fed moderate to high concentrations of toxigenic Alexandrium lusitanicum (Dutz, 1998). Although ingestion was high, and increased with increasing food concentrations, growth and egg production were inefficient, and remained constant despite increasing calculated ingestion rates (inferred from disappearance of cells). These results, in combination with the very low values for toxin assimilation measured in the present study, suggest that much of the total organic material of prey cells is not being assimilated when copepods feed on toxic Alexandrium cells. This in turn suggests two alternatives that could account for these results: either toxic Alexandrium cells impair the digestive systems of copepods to the point where efficient assimilation of ingested material is not possible (perhaps resulting in rapid excretion of fecal material), or copepods are practising extremely ‘sloppy feeding’.

Dutz suggested that toxins might interfere with the digestive process, or that the metabolic costs of detoxification are so high that they interfere with assimilation of organic material and growth (Dutz, 1998). High rates of metabolic detoxification and associated costs are not a probable explanation for the reasons listed above. Inhibited digestion is also an unlikely alternative explanation for poor carbon assimilation efficiency in the presence of PSP toxins. Hassett reported that digestive enzyme function and gut epithelial condition of copepods was not in any way impaired by consumption of toxic cells of Alexandrium spp., and that copepods actually had higher levels of enzyme activity when fed toxic cells than when fed non-toxic A. tamarense (Hassett, 1996). If poor digestion resulted in cellular material (including toxin) being ingested and passed rapidly through the system, fecal pellet fractions from experiments in the present study should have contained much higher levels of toxin. Although it is not certain that fecal material recovery was 100% effective, low pellet production, and low toxin levels measured in pellets, argue against production of a relatively large amount of undigested fecal material containing toxic remnants of Alexandrium cells. Turner et al. noticed that fecal pellet production was greatlyreduced among Centropages hamatus feeding on high concentrations of toxic Alexandrium tamarense compared with C. hamatus feeding on non-toxic Scrippsiella trochoidea, consistent with the hypothesis that ‘ingested’ food is not being rapidly passed unassimilated through the gut (Turner et al., 1998). It is more probable that copepods are unable to assimilate organic matter (carbon or toxin) efficiently, because of physiological or behavioral rejection of ‘ingested’ material.

The second alternative, regurgitation or ‘sloppy feeding’, is thus the best explanation for the observed results. Sykes and Huntley reported frequent regurgitation of gut contents by Calanus pacificus feeding on Protoceratium reticulatum (= Gonyaulax grindleyi) (Sykes and Huntley, 1987). They hypothesized that toxic dinoflagellates may cause acute physiological reactions when ingested, inducing regurgitation of inimical material in the gut. At the time these grazing experiments were conducted, toxicity was not associated with P. reticulatum, but this dinoflagellate species is now known to be a primarysource of the lipophilic phycotoxins known as yessotoxins (Satake et al., 1997).

In any case, the presence of toxin may not be necessary to induce such a regurgitation reaction. Powell and Berry demonstrated that inert plastic beads were routinely ingested and regurgitated by copepods, and that little or no fecal material was produced in the process (Powell and Berry, 1990). Regurgitation by copepods may be relatively commonplace after ingestion of food that is inimical, indigestible, or simply unpalatable. Beyond such direct observations of gross expulsion of ingested material, ‘sloppy feeding’ of palatable phytoplankton by microcrustacea has been reported to result in losses to the environment of 20–70% of calculated ingested material (Dagg, 1974; Lampert, 1978; Roy et al., 1989).

Toxin retention efficiency of copepods does not improve when the diet contains alternative food, even when the alternative food dominates the total available food (possibly excepting Centropages hamatus, see below). This result would seem to favor sloppy feeding as a hypothesis to explain poor toxin retention. Low amounts of toxic A. fundyense in the diet probably would not trigger physiological rejection (i.e. regurgitation) of gut contents composed primarily of non-toxic food, in which case one might expect higher retention efficiency at low A. fundyense concentrations. If, however, toxic cells were crushed and rejected at the time of feeding, i.e. behavioral rejection or sloppy feeding, toxin retention would be low and inefficient at all A. fundyense concentrations. Regurgitation may be more important at higher cell concentrations and in monocultures. Copepods that normally reject toxic cells are known to feed on cells of Alexandrium spp. when no (or little) other food is available, and physiological effects are more often observed when no other food source is available to grazers (Teegarden, 1999).

A possible exception to this conclusion is apparent in results from Centropages hamatus feeding in mixed food with low concentrations of A. fundyense. Owing to high variability, differences in toxin retention between C. hamatus treatments were not significant, but C. hamatus clearly had higher retention of toxin than other copepod species when A. fundyense was present at concentrations <100cells ml−1. In addition, electivity index values suggest that A. fundyense was never actively avoided by C. hamatus in the presence of mixed food. Perhaps C. hamatus does not practise selective rejection of A. fundyense at lower concentrations (Table IV), while still displaying low retention efficiency at higher concentrations (Tables I and III), from rejection or regurgitation. In any case, one must conclude that response to toxic prey varies among grazer species, as has been demonstrated numerous times [(Teegarden and Cembella, 1996; Teegarden, 1999), review by (Turner and Tester, 1997)].

Feeding behavior

Feeding behavior of copepods in varying concentrations of A. fundyense suggests some concentration-dependent avoidance of toxic prey. Deterrence of grazers by high concentrations of toxic cells, but not low concentrations, has been hypothesized or tested numerous times (Hansen, 1989; Nielsen et al., 1990; Sykes, 1991; Smayda, 1992; Donaghay and Osborn, 1997). Strong avoidance of cells of toxic Alexandrium spp. at high concentrations has been demonstrated in the laboratory (Turriff et al., 1995; Teegarden, 1999), but field studies at low concentrations of toxic Alexandrium cells indicate that avoidance is not universal (Teegarden et al., 2001).

All of the copepod species tested demonstrated moderate positive selection for A. fundyense at low concentrations (≤10 cells ml−1). Since A. fundyense was the largest motile prey item offered, cells may have been more conspicuous and thus positively selected, at least for capture. Given the low encounter rates, capture and crushing of A. fundyense would not have represented a significant problem of either handling cost or toxin exposure—particularly if toxic cells were rejected after being crushed, in hypothesized sloppy feeding. As A. fundyense concentrations increased and alternative food decreased, copepods displayed avoidance or decreased selection of A. fundyense, though the response of each copepod species varied. Results with Acartia hudsonica and Eurytemora herdmani suggest that at 50 cells ml−1 of A. fundyense, the concentration is sufficient to induce some avoidance (Ei < 0). At higher concentrations, E. herdmani displayed stronger avoidance, and much reduced total food consumption, while A. hudsonica did not practise strong selective feeding, but did substantially reduce total food consumption. Centropages hamatus never displayed strong avoidance of A. fundyense, but did reduce selective feeding on A. fundyense and total food consumption as A. fundyense concentrations increased.

One may generalize that as Alexandrium spp. concentrations increase, zooplankton are likely to avoid toxic prey cells increasingly, or reduce total ingestion rates, or both, though no one model can be made to fit each species response. From this study one could suggest that Alexandrium spp. concentrations of ≥ 5 × 104 cells l−1 represent a threshold concentration that triggers selective feeding. However, such a figure is not reliable for all species, and zooplankton feeding behavior probably responds to a continuum of conditions (toxicity cell−1, concentrations, alternative prey, etc.) rather than a threshold. If substantial rejection of toxic cell material is taking place (as suggested by this study), the concept of selective feeding may need clarification to include actual selective ingestion, rather than selective ingestion inferred or calculated from cell disappearance in grazing experiments.

Conclusions

Copepods feeding on toxic Alexandrium spp. regularly displayed very low retention of toxin, despite calculated high ingestion rates, across a continuum of conditions. Toxin analyses of field-collected zooplankton during blooms of Alexandrium spp. support the hypothesis that low levels of toxin are accumulated in zooplankton (Campbell et al., 2000; Turner et al., 2000; Campbell et al., unpublished). Body burdens of PSP toxins in these experiments and the field studies mentioned indicate that there is a threat of transfer of toxins in marine food webs via zooplankton. Should zooplankton be considered a significant vector of toxins? The relatively large amounts of toxic Alexandrium spp. consumed and the very small fraction of toxin retained in the consumers, suggest that zooplankton effectively disperse PSP toxins into the surrounding environment. Even after fecal material is accounted for,∼90% of toxin from cells removed by grazing activity is not made available to higher consumers. The zooplankton component of the food web is a sink into which PSP toxins pass, but that does not pass PSP toxins efficiently to other trophic components; in the Gulf of Maine, zooplankton appear to be inefficient, but perhaps not ineffective, vectors of PSP toxins.

Table I:

Toxin budgets of the three copepod species feeding on monocultures of Alexandrium fundyense

 Toxin ingested Toxin retained  Total toxin in fecal pellets Unaccounted toxin 
 ng STXeq copped ng STXeq copped % of toxin ingested ng STXeq % of toxin ingested ng STXeq % of toxin ingested 
For each copepod species category, the first column indicates calculated toxin ingested (total cells removed copepod−1 × toxicity cell−1). Following columns indicate measured values for toxin content copepod−1 of copepod samples, total toxin in fecal pellet samples, and calculated unaccounted toxin, expressed as ng saxitoxin equivalents (ngSTXeq), and the per cent of the total calculated ingested toxin for each category. 
Acartia 5.4 0.01 0.2 16.4 5.8 263.5 94.0 
hudsonica 4.3 0.01 0.2 6.6 3.1 208.2 96.7 
 5.8 0.05 0.9 12.4 4.1 289.7 95.1 
 5.2 ± 0.8 0.02 ± 0.02 0.4 ± 0.4 11.8 ± 4.9 4.3 ± 1.4 254 ± 41.6 95.3 ± 1.4 
Centropages 8.1 0.4 4.4 8.9 2.2 380.4 93.4 
hamatus 9.3 0.4 4.3 18.0 3.9 427.4 91.8 
 2.5 0.4 16.3 20.7 16.5 84.3 67.2 
 6.6 ± 3.7 0.4 ± 0.02 8.3 ± 6.9 15.9 ± 6.2 7.5 ± 7.8 297 ± 186 84.2 ± 14.7 
Eurytemora 28.2 0.3 1.2 0.01 0.0 1435.0 98.90 
herdmani 32.0 0.8 2.5 56.6 3.3 1598.2 94.2 
 26.7 0.8 3.2 41.8 3.3 1197.4 93.6 
 29.0 ± 2.8 0.7 ± 0.3 2.3 ± 0.5 32.8 ± 29.3 2.2 ± 1.9 1398 ± 283 94 ± 0.4 
 Toxin ingested Toxin retained  Total toxin in fecal pellets Unaccounted toxin 
 ng STXeq copped ng STXeq copped % of toxin ingested ng STXeq % of toxin ingested ng STXeq % of toxin ingested 
For each copepod species category, the first column indicates calculated toxin ingested (total cells removed copepod−1 × toxicity cell−1). Following columns indicate measured values for toxin content copepod−1 of copepod samples, total toxin in fecal pellet samples, and calculated unaccounted toxin, expressed as ng saxitoxin equivalents (ngSTXeq), and the per cent of the total calculated ingested toxin for each category. 
Acartia 5.4 0.01 0.2 16.4 5.8 263.5 94.0 
hudsonica 4.3 0.01 0.2 6.6 3.1 208.2 96.7 
 5.8 0.05 0.9 12.4 4.1 289.7 95.1 
 5.2 ± 0.8 0.02 ± 0.02 0.4 ± 0.4 11.8 ± 4.9 4.3 ± 1.4 254 ± 41.6 95.3 ± 1.4 
Centropages 8.1 0.4 4.4 8.9 2.2 380.4 93.4 
hamatus 9.3 0.4 4.3 18.0 3.9 427.4 91.8 
 2.5 0.4 16.3 20.7 16.5 84.3 67.2 
 6.6 ± 3.7 0.4 ± 0.02 8.3 ± 6.9 15.9 ± 6.2 7.5 ± 7.8 297 ± 186 84.2 ± 14.7 
Eurytemora 28.2 0.3 1.2 0.01 0.0 1435.0 98.90 
herdmani 32.0 0.8 2.5 56.6 3.3 1598.2 94.2 
 26.7 0.8 3.2 41.8 3.3 1197.4 93.6 
 29.0 ± 2.8 0.7 ± 0.3 2.3 ± 0.5 32.8 ± 29.3 2.2 ± 1.9 1398 ± 283 94 ± 0.4 
Table II:

Calculated ingestion rates of the three copepod species for monoculture feeding experiments and two species mixed algae feeding experiments

 Monoculture Mixed algae 
 Ingestion rate (ng C cop.−1 h−1% of body C day−1 Ingestion rate (ng C cop.−1 h−1)(A. fundyense/H. triquetra% of body C day−1 (A. fundyense/H. triquetra
For each copepod species and experimental treatment category, left columns indicate ingestion rate in ng carbon copepod−1 h−1, and right columns indicate daily ingestion as a percentage of estimated body carbon individual−1. Mean ± SD is indicated under each column, and for mixed food experiments, electivity index Ei is given for both food items at the bottom. 
Acartia 44.7 23.6 161.6 / 54.7 85.3 / 28.8 
hudsonica 32.9 17.4 76.9 / 88.8 40.6 / 46.8 
 40.8 21.5 205.9 / 92.5 108.6 / 48.8 
 39.5 ± 6.0 20.8 ± 3.2 148.2 ± 65.6 / 78.6 ± 20.8 78.2 ± 34.6 / 41.5 ±11.0 
   Ei = –0.124 / 0.099  
Centropages 62.27 14.9 348.0 / 49.8 83.5 / 11.9 
hamatus 65.67 15.8 225.3 / 47.5 54.1 / 11.4 
 16.71 4.0 243.7 / 45.4 58.5 / 10.9 
 48.2 ± 27.3 11.6 ± 6.6 272.3 ± 66.2 / 47.6 ± 2.2 65.4 ± 15.9 / 11.4 ± 0.5 
   Ei = 0.168 / –0.252  
Eurytemora 135.5 54.2 152.5 / 67.6 61.0 / 27.1 
herdmani 137.8 55.1 106.4 / 70.4 42.6 / 28.1 
 104.0 41.6 152.2 / 65.9 60.9 / 26.4 
 125.8 ± 18.9 50.3 ± 7.6 137.0 ± 26.5 / 68.0 ± 2.2 54.8 ± 10.6 / 27.2 ± 0.9 
   Ei = –0.101 / 0.084  
 Monoculture Mixed algae 
 Ingestion rate (ng C cop.−1 h−1% of body C day−1 Ingestion rate (ng C cop.−1 h−1)(A. fundyense/H. triquetra% of body C day−1 (A. fundyense/H. triquetra
For each copepod species and experimental treatment category, left columns indicate ingestion rate in ng carbon copepod−1 h−1, and right columns indicate daily ingestion as a percentage of estimated body carbon individual−1. Mean ± SD is indicated under each column, and for mixed food experiments, electivity index Ei is given for both food items at the bottom. 
Acartia 44.7 23.6 161.6 / 54.7 85.3 / 28.8 
hudsonica 32.9 17.4 76.9 / 88.8 40.6 / 46.8 
 40.8 21.5 205.9 / 92.5 108.6 / 48.8 
 39.5 ± 6.0 20.8 ± 3.2 148.2 ± 65.6 / 78.6 ± 20.8 78.2 ± 34.6 / 41.5 ±11.0 
   Ei = –0.124 / 0.099  
Centropages 62.27 14.9 348.0 / 49.8 83.5 / 11.9 
hamatus 65.67 15.8 225.3 / 47.5 54.1 / 11.4 
 16.71 4.0 243.7 / 45.4 58.5 / 10.9 
 48.2 ± 27.3 11.6 ± 6.6 272.3 ± 66.2 / 47.6 ± 2.2 65.4 ± 15.9 / 11.4 ± 0.5 
   Ei = 0.168 / –0.252  
Eurytemora 135.5 54.2 152.5 / 67.6 61.0 / 27.1 
herdmani 137.8 55.1 106.4 / 70.4 42.6 / 28.1 
 104.0 41.6 152.2 / 65.9 60.9 / 26.4 
 125.8 ± 18.9 50.3 ± 7.6 137.0 ± 26.5 / 68.0 ± 2.2 54.8 ± 10.6 / 27.2 ± 0.9 
   Ei = –0.101 / 0.084  
Table III:

Toxin budgets of the three copepod species feeding on mixtures of Alexandrium fundyense and Heterocapsa triquetra

 Toxin ingested Toxin retained Unaccounted toxin 
 ng STXeq copepod−1 ng STXeq copepod−1 % of toxin ingested ng STXeq % of toxin ingested 
Ingestion of toxin based on ingestion of A. fundyense only, because H. triquetra is non-toxic. Legend as in Table I. Fecal toxin content was not measured in this experiment, so toxin not retained in body tissues is included in unaccounted toxin. 
Acartia 33.5 0.5 1.4 1617.5 98.6 
hudsonica 14.5 0.3 2.4 692.1 97.6 
 11.7 0.3 2.9 559.1 97.1 
 19.9 ± 11.9 0.4 ± 0.1 2.2 ± 0.8 956.2 ± 576.5 97.3 ± 1.4 
Centropages  0.6 – – – 
hamatus 19.9 0.8 3.9 773.6 96.1 
 24.2 0.9 3.8 795.8 96.2 
 22.1 ± 3.1 0.8 ± 0.2 3.8 ± 0.1 784.7 ± 15.6 96.2 ± 0.1 
Eurytemora 4.6 – – – – 
herdmani 33.3 0.7 2.2 1528.2 97.8 
 37.5 1.4 3.7 1841.6 96.3 
 25.1 ± 17.9 1.1 ± 0.5 3.0 ± 1.1 1685.0 ± 221.6 96.8 ± 0.7 
 Toxin ingested Toxin retained Unaccounted toxin 
 ng STXeq copepod−1 ng STXeq copepod−1 % of toxin ingested ng STXeq % of toxin ingested 
Ingestion of toxin based on ingestion of A. fundyense only, because H. triquetra is non-toxic. Legend as in Table I. Fecal toxin content was not measured in this experiment, so toxin not retained in body tissues is included in unaccounted toxin. 
Acartia 33.5 0.5 1.4 1617.5 98.6 
hudsonica 14.5 0.3 2.4 692.1 97.6 
 11.7 0.3 2.9 559.1 97.1 
 19.9 ± 11.9 0.4 ± 0.1 2.2 ± 0.8 956.2 ± 576.5 97.3 ± 1.4 
Centropages  0.6 – – – 
hamatus 19.9 0.8 3.9 773.6 96.1 
 24.2 0.9 3.8 795.8 96.2 
 22.1 ± 3.1 0.8 ± 0.2 3.8 ± 0.1 784.7 ± 15.6 96.2 ± 0.1 
Eurytemora 4.6 – – – – 
herdmani 33.3 0.7 2.2 1528.2 97.8 
 37.5 1.4 3.7 1841.6 96.3 
 25.1 ± 17.9 1.1 ± 0.5 3.0 ± 1.1 1685.0 ± 221.6 96.8 ± 0.7 
Table IV:

Toxin budgets of the three copepod species feeding on mixtures of Alexandrium fundyense and three alternative species of phytoplankton (see methods text)

 Treatment Toxin ingested Toxin retained Unaccounted toxin 
 (Alexandrium concentrationng STXeq copepod ng STXeq copepod % of toxin ingested ng STXeq % of toxin ingested 
Columns indicate treatment indicating A. fundyense concentration, toxin ingested, retained, or unaccounted, expressed as ngSTXeq (left column under each category heading, mean ± SD for each treatment), and percentage of ingested toxin (right column under category, mean ± SD). 
Acartia 5 ×103 l−1 6.4 ± 1.0 0.01 ± 0.01 0.1 ± 0.04 127.8 99.9 
hudsonica 104 l−1 22.1 ± 2.8 0.01 ± 0.01 0.1 ± 0.01 441.7 99.9 
 5 ×104 l−1 21.7 ± 19.0 0.1 ± 0.04 2.0 ± 2.8 432.9 98.0 
 105 l−1 133.9 ± 9.7 0.16 ± 0.03 0.1 ± 0.02 2674.5 99.9 
Centropages 5 ×103 l−1 6.2 ± 1.6 1.4 ± 0.8 24.9 ± 19.7 71.7 75.1 
hamatus 104 l−1 17.2 ± 1.1 1.3 ± 0.3 7.7 ± 1.5 238.4 92.3 
 5 ×104 l−1 76.3 ± 16.3 5.8 ± 0.2 7.8 ± 1.8 1058.1 92.2 
 105 l−1 95.1 ± 18.3 5.6 ± 0.2 6.0 ± 1.5 1342.5 96.0 
Eurytemora 5 ×103 l−1 12.9 ± 1.2 0.1 ± 0.1 0.6 ± 0.8 256.4 98.8 
herdmani 104 l−1 31.6 ± 0.7 0.4 ± 0.1 1.1 ± 0.2 624.8 97.7 
 5 ×104 l−1 60.7 ± 3.2 0.6 ± 0.2 1.0 ± 0.3 1203.0 98.0 
 105 l−1 46.7 ± 3.6 0.9 ± 0.1 1.8 ± 0.4 916.0 96.3 
 Treatment Toxin ingested Toxin retained Unaccounted toxin 
 (Alexandrium concentrationng STXeq copepod ng STXeq copepod % of toxin ingested ng STXeq % of toxin ingested 
Columns indicate treatment indicating A. fundyense concentration, toxin ingested, retained, or unaccounted, expressed as ngSTXeq (left column under each category heading, mean ± SD for each treatment), and percentage of ingested toxin (right column under category, mean ± SD). 
Acartia 5 ×103 l−1 6.4 ± 1.0 0.01 ± 0.01 0.1 ± 0.04 127.8 99.9 
hudsonica 104 l−1 22.1 ± 2.8 0.01 ± 0.01 0.1 ± 0.01 441.7 99.9 
 5 ×104 l−1 21.7 ± 19.0 0.1 ± 0.04 2.0 ± 2.8 432.9 98.0 
 105 l−1 133.9 ± 9.7 0.16 ± 0.03 0.1 ± 0.02 2674.5 99.9 
Centropages 5 ×103 l−1 6.2 ± 1.6 1.4 ± 0.8 24.9 ± 19.7 71.7 75.1 
hamatus 104 l−1 17.2 ± 1.1 1.3 ± 0.3 7.7 ± 1.5 238.4 92.3 
 5 ×104 l−1 76.3 ± 16.3 5.8 ± 0.2 7.8 ± 1.8 1058.1 92.2 
 105 l−1 95.1 ± 18.3 5.6 ± 0.2 6.0 ± 1.5 1342.5 96.0 
Eurytemora 5 ×103 l−1 12.9 ± 1.2 0.1 ± 0.1 0.6 ± 0.8 256.4 98.8 
herdmani 104 l−1 31.6 ± 0.7 0.4 ± 0.1 1.1 ± 0.2 624.8 97.7 
 5 ×104 l−1 60.7 ± 3.2 0.6 ± 0.2 1.0 ± 0.3 1203.0 98.0 
 105 l−1 46.7 ± 3.6 0.9 ± 0.1 1.8 ± 0.4 916.0 96.3 
Table V:

Ingestion rates of the three copepod species in mixed food/varying Alexandrium fundyense concentration experiments

Treatment Alexandriumcells l−1 Species ingested Acartia hudsonica Centropages hamatus Eurytemora herdmani 
  Ingestion Ei Ingestion Ei Ingestion Ei 
Left column, A. fundyense concentration treatment, followed by phytoplankton species ingested, and subsequently ingestion rate in ng C copepod−1 h−1 (mean ± SD, left column) and electivity index Ei (right column) for each copepod species. 
5 × 103 A. fundyense 7.0 ± 1.0 0.3 7.0 ± 1.8 0.3 20.0 ± 1.8 0.2 
 H. triquetra 51.8 ± 11.1 –0.1 84.8 ± 14.2 –0.1 78.4 ± 3.4 –0.3 
 T. weissflogii 139.0 ± 31.7 0.2 97.8 ± 51.7 –0.2 135.8 ± 3.2 0.3 
 R. salina 8.0 ± 4.0 –1.0 33.2 ± 13.6 –0.4 11.2 ± 4.9 –0.9 
 Total 205.9  222.8  245.4  
1 × 104 A. fundyense 22.6 ± 2.9 0.1 18.2 ± 1.2 0.4 47.0 ± 1.0 0.2 
 H. triquetra 77.5 ± 12.9 –0.2 78.5 ± 21.7 –0.2 61.1 ± 4.6 –0.3 
 T. weissflogii 198.2 ± 2.5 0.2 104.1 ± 19.3 –0.2 137.4 ± 0.2 0.3 
 R. salina 31.2 ± 12.6 –0.2 33.5 ± 21.0 –0.4 18.4 ± 0.8 –0.8 
 Total 329.5  234.3  263.9  
5 × 104 A. fundyense 22.3 ± 19.4 –0.5 83.3 ± 17.6 0.3 89.5 ± 4.9 –0.1 
 H. triquetra 73.2 ± 7.2 0.2 42.8 ± 26.4 –0.1 43.9 ± 2.2 0.0 
 T. weissflogii 104.4 ± 23.7 0.2 57.6 ± 14.8 –0.1 70.9 ± 1.6 0.4 
 R. salina 4.2 ± 3.7 –0.2 16.0 ± 11.6 –0.3 4.2 ± 4.1 –0.8 
 Total 203.9  199.8  208.4  
1 × 105 A. fundyense 131.9 ± 10.1 0.0 100.2 ± 19.2 0.1 67.2 ± 5.1 –0.5 
 H. triquetra 4.2 ± 2.9 –0.5 5.5 ± 1.0 –0.5 21.7 ± 5.0 0.0 
 T. weissflogii 27.3 ± 9.9 0.0 35.5 ± 1.5 0.3 54.3 ± 3.1 0.4 
 R. salina 0.8 ± 1.3 0.3 1.4 ± 2.4 –0.2 0.3 ± 0.4 –1.0 
 Total 164.2  142.5  143.4  
Treatment Alexandriumcells l−1 Species ingested Acartia hudsonica Centropages hamatus Eurytemora herdmani 
  Ingestion Ei Ingestion Ei Ingestion Ei 
Left column, A. fundyense concentration treatment, followed by phytoplankton species ingested, and subsequently ingestion rate in ng C copepod−1 h−1 (mean ± SD, left column) and electivity index Ei (right column) for each copepod species. 
5 × 103 A. fundyense 7.0 ± 1.0 0.3 7.0 ± 1.8 0.3 20.0 ± 1.8 0.2 
 H. triquetra 51.8 ± 11.1 –0.1 84.8 ± 14.2 –0.1 78.4 ± 3.4 –0.3 
 T. weissflogii 139.0 ± 31.7 0.2 97.8 ± 51.7 –0.2 135.8 ± 3.2 0.3 
 R. salina 8.0 ± 4.0 –1.0 33.2 ± 13.6 –0.4 11.2 ± 4.9 –0.9 
 Total 205.9  222.8  245.4  
1 × 104 A. fundyense 22.6 ± 2.9 0.1 18.2 ± 1.2 0.4 47.0 ± 1.0 0.2 
 H. triquetra 77.5 ± 12.9 –0.2 78.5 ± 21.7 –0.2 61.1 ± 4.6 –0.3 
 T. weissflogii 198.2 ± 2.5 0.2 104.1 ± 19.3 –0.2 137.4 ± 0.2 0.3 
 R. salina 31.2 ± 12.6 –0.2 33.5 ± 21.0 –0.4 18.4 ± 0.8 –0.8 
 Total 329.5  234.3  263.9  
5 × 104 A. fundyense 22.3 ± 19.4 –0.5 83.3 ± 17.6 0.3 89.5 ± 4.9 –0.1 
 H. triquetra 73.2 ± 7.2 0.2 42.8 ± 26.4 –0.1 43.9 ± 2.2 0.0 
 T. weissflogii 104.4 ± 23.7 0.2 57.6 ± 14.8 –0.1 70.9 ± 1.6 0.4 
 R. salina 4.2 ± 3.7 –0.2 16.0 ± 11.6 –0.3 4.2 ± 4.1 –0.8 
 Total 203.9  199.8  208.4  
1 × 105 A. fundyense 131.9 ± 10.1 0.0 100.2 ± 19.2 0.1 67.2 ± 5.1 –0.5 
 H. triquetra 4.2 ± 2.9 –0.5 5.5 ± 1.0 –0.5 21.7 ± 5.0 0.0 
 T. weissflogii 27.3 ± 9.9 0.0 35.5 ± 1.5 0.3 54.3 ± 3.1 0.4 
 R. salina 0.8 ± 1.3 0.3 1.4 ± 2.4 –0.2 0.3 ± 0.4 –1.0 
 Total 164.2  142.5  143.4  
Fig. 1.

Sample chromatogram from HPLC-FD analysis of PSP toxins retained in Centropages hamatus fed Alexandrium fundyense. C1/C2 =N-sulfocarbamoyl ‘C-toxins’; STX = saxitoxin, NEO = neosaxitoxin, GTX1–GTX4 = gonyautoxins 1,2,3 and 4.

Fig. 1.

Sample chromatogram from HPLC-FD analysis of PSP toxins retained in Centropages hamatus fed Alexandrium fundyense. C1/C2 =N-sulfocarbamoyl ‘C-toxins’; STX = saxitoxin, NEO = neosaxitoxin, GTX1–GTX4 = gonyautoxins 1,2,3 and 4.

Fig. 2.

Toxin composition profiles of (a) Alexandrium fundyense clone GTCA 28, (b) tissue samples of whole Eurytemora herdmani copepods after feeding on GTCA 28, and (c) copepod fecal pellet samples from experiments in which copepods fed on GTCA 28 (mean values of triplicate samples from monoculture toxin budget experiments).

Fig. 2.

Toxin composition profiles of (a) Alexandrium fundyense clone GTCA 28, (b) tissue samples of whole Eurytemora herdmani copepods after feeding on GTCA 28, and (c) copepod fecal pellet samples from experiments in which copepods fed on GTCA 28 (mean values of triplicate samples from monoculture toxin budget experiments).

Fig. 3.

Clearance rates of (a) Acartia hudsonica, (b) Centropages hamatus, and (c) Eurytemora herdmani, in four treatments of varying concentrations of Alexandrium fundyense (clearance of A. fundyense only shown, for other phytoplankton refer to Table V). Electivity index Ei is given next to each clearance rate.

Clearance rates of (a) Acartia hudsonica, (b) Centropages hamatus, and (c) Eurytemora herdmani, in four treatments of varying concentrations of Alexandrium fundyense (clearance of A. fundyense only shown, for other phytoplankton refer to Table V). Electivity index Ei is given next to each clearance rate.

We thank N. Lewis, K. Thomas and K. Reeves, IMB-NRC, for HPLC-FD analyses and chromatogram processing. T. Miller of the Darling Marine Center, University of Maine, assisted with laboratory set-up. Alexandrium fundyense clone BC1 was contributed by B. Thompson of Bigelow Laboratory for Ocean Science, and GTCA 28 was provided by the laboratory of D. M. Anderson, Woods Hole Oceanographic Institution. This study was funded by NSF grant OCE-9726261. The publication is NRCC # 42360.

REFERENCES

Anderson, D. M., Kulis, D. M., Doucette, G. J., Gallagher, J. C. and Balech, E. (
1994
) Biogeography of toxic dinoflagellates in the genus Alexandrium from the northeastern United States and Canada.
Mar. Biol.
 ,
120
,
467
–468.
AOAC (1980) Paralytic shellfish poisons. In Horwitz, W. (ed.), Official Methods of the Association of Official Analytical Chemists, 13th edition. AOAC, Washington, DC, pp. 298–299.
Boyer, G. L., Sullivan, J. J., LeBlanc, M. and Andersen, R. J. (1985) Assimilation of PSP toxins by the copepod Tigriopus californicus from dietary Protogonyaulax catanella. In Anderson, D. M., White, A. W.and Baden, D. G. (eds), Toxic Dinoflagellates. Elsevier, New York, pp. 407–412.
Bricelj, V. M., Lee, J. H., Cembella, A. D. and Anderson, D. M. (
1990
) Uptake kinetics of paralytic shellfish toxins from the dinoflagellate Alexandrium fundyense in the mussel Mytilus edulis.
Mar. Ecol. Prog. Ser.
 ,
74
,
33
–46.
Bricelj, V. M., Lee, J. H. and Cembella, A. D. (
1991
) Influence of dinoflagellate cell toxicity on uptake and loss of paralytic shellfish toxins in the northern quahog Mercenaria mercenaria.
Mar. Ecol. Prog. Ser.
 ,
74
,
33
–46.
Bricelj, V. M., Cembella, A. D., Laby, D., Shumway, S. E. and Cucci, T. L. (1996) Comparative physiological and behavioral responses to PSP toxins in two bivalve molluscs, the softshell clam, Mya arenaria, and surfclam, Spisula solidissima. In Yasumoto, T., Oshima, Y. and Fukuyo, Y. (eds), Harmful and Toxic Algal Blooms. Intergovernmental Oceanographic Commission of UNESCO, Paris, pp. 405–408.
Campbell, R. G., Teegarden, G. J., Cembella, A. D. and Durbin, E. G. (2000) The role of zooplankton grazers in Alexandrium sp. bloom dynamics in the near-shore environment of the Gulf of Maine. In Anderson, D. M. (ed.), Symposium on Harmful Marine Algae in the US. Marine Biological Laboratory, Woods Hole MA, Abstract p. 100.
Cembella, A. D., Sullivan, J. J., Boyer, G. L., Taylor, F. J. R. and Andersen, R. J. (
1987
) Variation in paralytic shellfish toxin composition within the Protogonyaulax tamarensis/catenella species complex.
Biochem. System. Ecol.
 ,
15
,
171
–186.
Cembella, A. D., Shumway, S. E. and Lewis, N. (
1993
) A comparison of anatomical distribution and spatio-temporal variation of paralytic shellfish toxin composition in two bivalve species from the Gulf of Maine.
J. Shellfish Res.
 ,
12
,
389
–403.
Cembella, A. D., Shumway, S. E. and Larocque, R. (
1994
) Sequestering and putative biotransformation of paralytic shellfish toxins by the giant sea scallop Placopecten magellanicus: seasonal and spatial scales in natural populations.
J. Exp. Mar. Biol. Ecol.
 ,
180
,
1
–22.
Dagg, M. J. (
1974
) Loss of prey body contents during feeding by an aquatic predator.
Ecology
 ,
55
,
903
–906.
Donaghay, P. L. and Osborn, T. R. (
1997
) Toward a theory of biological-physical control of harmful algal bloom dynamics and impacts.
Limnol. Oceanogr.
 ,
42
,
1283
–1296.
Dutz, J. (
1998
) Repression of fecundity in the neritic copepod Acartia clausi exposed to the toxic dinoflagellate Alexandrium lusitanicum: relationship between feeding and egg production.
Mar. Ecol. Prog. Ser.
 ,
175
,
97
–107.
Frost, B. W. (
1972
) Effects of size and concentration of food particles on the feeding behavior of the marine planktonic copepod Calanus pacificus.
Limnol. Oceanogr.
 ,
17
,
805
–815.
Guillard, R. R. L. (1975) Culture of marine phytoplankton for feeding marine invertebrates. In Smith, W. L. and Chaney, M. H. (eds), Culture of Marine Invertebrate Animals. Plenum, New York, pp. 29–60.
Guisande, C., Frangopulos, M., Carotenuto, Y., Maneiro, I., Riveiro, I. and Vergara, A. R. (
2002
) Fate of paralytic shellfish poisoning toxins ingested by the copepod
Acartia clausi. Mar. Ecol. Prog. Ser.
 ,
240
,
105
–115.
Hansen, P. J. (
1989
) The red tide dinoflagellate Alexandrium tamarense: effects on behavior and growth of a tintinnid ciliate.
Mar. Ecol. Prog. Ser.
 ,
53
,
105
–116.
Hassett, R. P. (
1996
) Effects of the red-tide dinoflagellate Alexandrium tamarense on copepod digestion: toxin level does not predict physiological impact.
EOS Transactions
 ,
76
,
OS12G
–3.
Lampert, W. (
1978
) Release of dissolved organic carbon by grazing zooplankton.
Limnol. Oceanogr.
 ,
23
,
831
–834.
Le Borgne, R. (1986) The release of soluble end products of metabolism. In Corner, E. D. S. and O’Hara, S. C. M. (eds), The Biological Chemistry of Marine Copepods. Clarendon Press, Oxford, Chapter 3.
Nielsen, T. G., Kiørboe, T. and Bjørnsen, P. K. (
1990
) Effect of a Chrysochromulina polylepis subsurface bloom on the planktonic community.
Mar. Ecol. Prog. Ser.
 ,
62
,
21
–35.
Oshima, Y. (1995a) Post-column derivatization HPLC methods for paralytic shellfish poisons. In Hallegraeff, G. M., Anderson, D. M., Cembella, A. D. and Enevoldsen, H. O. (eds), Manual On Harmful Marine Microalgae. International Oceanographic Commission of UNESCO, Paris, pp. 81–94.
Oshima, Y. (1995b) Chemical and enzymatic transformations of paralytic shellfish toxins in marine organisms. In Lassus, P., Arzul, G., Erard–Le Denn, E., Gentien, P. and Marcaillou-Le Baut, C. (eds), Harmful Marine Algal Blooms. Lavoisier Science Publishers, Paris, pp. 475–480.
Parkhill, J. P. and Cembella, A. D. (
1999
) Effects of salinity, light, and inorganic nitrogen on growth and toxigenicity of the marine dinoflagellate Alexandrium tamarense from northeastern Canada.
J. Plankton Res.
 ,
21
,
939
–955.
Powell, M. D. and Berry, A. J. (
1990
) Ingestion and regurgitation of living and inert materials by the estuarine copepod Eurytemora affinis (Poppe) and the influence of salinity.
Estuarine Coastal Shelf Sci.
 ,
31
,
763
–773.
Robineau, B., Gagné, J. A., Fortier, L. and Cembella, A. D. (
1991
) Potential impact of a toxic dinoflagellate (Alexandrium excavatum) bloom on survival of fish and crustacean larvae.
Mar. Biol.
 ,
108
,
293
–301.
Robineau, B., Gagné, J.A., Fortier, L. and Cembella, A. D. (
1991
) Comparison of the response of five larval fish species to the toxic dinoflagellate Alexandrium excavatum (Braarud) Balech.
J. Exp. Mar. Biol. Ecol.
 ,
152
,
225
–242.
Roy, S., Harris, R. P. and Poulet, S. A. (
1989
) Inefficient feeding by Calanus helgolandicus and Temora longicornis on Coscinodiscus wailesii: quantitative estimation using chlorophyll-type pigments and effects on dissolved free amino acids.
Mar. Ecol. Prog. Ser.
 ,
52
,
145
–153.
Satake, M., McKenzie, L. and Yusumoto, T. (
1997
) Identification of Protoceratium reticulation as the biogenetic origin of yessotoxin.
Nat. Toxins
 ,
5
,
164
–167.
Shumway, S. E., Sherman-Caswell, S. and Hurst, J. W. (
1988
) Paralytic shellfish poisoning in Maine: monitoring a monster.
J. Shellfish Res.
 ,
7
,
643
–652.
Smayda, T. (1992) Global epidemic of noxious phytoplankton blooms and food chain consequences in large ecosystems. In Sherman, K., Alexander, L. M. and Gold, B. D. (eds), Food Chains, Yields, Models and Management of Large Marine Ecosystems. Westview Press, Boulder, pp. 275–307.
Sykes, P. F. (1991) Physiological-ecology and chemical-ecology of copepod-dinoflagellate interactions. PhD dissertation, University of California, San Diego.
Sykes, P. F. and Huntley, M. E. (
1987
) Acute physiological reactions of Calanus pacificus to selected dinoflagellates: direct observations.
Mar. Biol.
 ,
94
,
19
–24.
Teegarden, G. J. (
1999
) Copepod grazing selection and particle discrimination on the basis of PSP toxin content.
Mar. Ecol. Prog. Ser.
 ,
181
,
163
–176.
Teegarden, G. J. and Cembella, A. D. (
1996
) Grazing of toxic dinoflagellates, Alexandrium spp., by adult copepods of coastal Maine: implications for the fate of paralytic shellfish toxins in marine food webs.
J. Exp. Mar. Biol. Ecol.
 ,
196
,
145
–176.
Teegarden, G. J., Campbell, R.G. and Durbin, E. G. (
2001
) Zooplankton feeding behavior and particle selection in natural assemblages containing toxic Alexandrium spp.
Mar. Ecol. Prog. Ser.
 ,
218
,
213
–226.
Turner, J. T. and Tester, P. A. (
1997
) Toxic marine phytoplankton, zooplankton grazers, and pelagic food webs.
Limnol. Oceanogr.
 ,
42
,
1203
–1214.
Turner, J. T., Lincoln, J. A. and Cembella, A. D. (1998) Effects of toxic and non–toxic dinoflagellates on copepod grazing, egg production and egg hatching success. In Reguera, B., Blanco, J., Fernández, M. L. and Wyatt, T. (eds), Harmful Algae. Xunta de Galicia and Intergovernmental Oceanographic Commission of UNESCO, Paris, pp. 379– 381.
Turner, J. T., Powell, C. L., Kulis, D. M., Keafer, B. A., Anderson, D. M. and Doucette, G. J. (2000) Accumulation of PSP toxins in zooplankton assemblages in the Gulf of Maine. In Anderson, D. M. (ed.), Symposium on Harmful Marine Algae in the US. Marine Biological Laboratory, Woods Hole MA, Abstract p. 74.
Turriff, N., Runge, J. A. and Cembella, A. D. (
1995
) Toxin accumulation and feeding behaviour of the planktonic copepod Calanus finmarchicus exposed to the red-tide dinoflagellate Alexandrium excavatum.
Mar. Biol.
 ,
123
,
55
–64.
Vanderploeg, H. A. and Scavia, D. (
1979
) Two electivity indices for feeding with special reference to zooplankton grazing.
J. Fish. Res. Bd. Can.
 ,
36
,
362
–365.
White, A. W. (1979) Dinoflagellate toxins in phytoplankton and zooplankton fractions during a bloom of Gonyaulax excavata. In Taylor, F. J. R. and Seliger, H. H. (eds), Toxic Dinoflagellate Blooms. Elsevier, New York, pp. 381–384.
White, A. W. (
1980
) Recurrence of kills of Atlantic herring (Clupea harengus harengus) caused by dinoflagellate toxins transferred through herbivorous zooplankton.
Can. J. Fish. Aquat. Sci.
 ,
37
,
2262
–2265.
White, A. W. (
1981
) Marine zooplankton can accumulate and retain dinoflagellate toxins and cause fish kills.
Limnol. Oceanogr.
 ,
26
,
103
–109.
White, A. W. (1984) Paralytic shellfish toxins and finfish. In Ragelis, E. P. (ed.), Seafood Toxins. American Chemistry Society, Washington DC, pp. 171–180.
White, A. W., Fukuhara, O. and Anraku, M. (1989) Mortality of fish larvae from eating toxic dinoflagellates or zooplankton containing dinoflagellate toxins. In Okaichi, T., Anderson, D. M. and Nemoto, T. (eds), Red Tides: Biology, Environmental Science, and Toxicology. Elsevier, New York, pp. 395–398.