Abstract

Three strains of the marine dinoflagellate Alexandrium ostenfeldii of different geographic origin were tested for their short-term deleterious effects on a diversity of marine protists. All A. ostenfeldii strains were capable of eliciting an apparent allelochemical response, but the various protistan target species were differentially affected. Protists that were negatively affected by exposure to cells of A. ostenfeldii and associated extracellular metabolites comprised both autotrophs (Rhodomonas sp., Dunaliella salina, Thalassiosira weissflogii) and heterotrophs (Oxyrrhis marina, Amphidinium crassum, Rimostrombidium caudatum). Observed effects included immobilisation (e.g. of O. marina), morphological changes (e.g. in D. salina) and/or aberrant behaviour (e.g. of R. caudatum), mainly as preliminary stages of cell lysis. Immobilization and lytic effects against O. marina were strongly dependent on A. ostenfeldii cell concentrations. Effects also differed substantially among strains and different batch cultures of the same strain. Values of EC50, defined as the A. ostenfeldii cell concentration causing lysis of 50% of O. marina cells, ranged from 0.3 to 1.9 × 103 mL−1, depending on the A. ostenfeldii strain. The autotrophic dinoflagellate Scrippsiella trochoidea reacted to exposure to A. ostenfeldii cells by formation of temporary (ecdysal) cysts, whereas, in contrast, the flagellates Emiliania huxleyi and Prymnesium parvum and the ciliate Strombidium sp. were relatively refractory or even unaffected. As long as cells did not lyse, the fluorescence yield of target autotrophs, estimated by pulse-amplitude modulation fluorometry, did not significantly change during the first 3 h of incubation, suggesting that allelochemicals produced by A. ostenfeldii caused no short-term negative effects on the photosynthetic apparatus. Overall, the allelochemical responses of target species showed no obvious relationship to cell quota or extracellular concentrations of either toxic macrocyclic imines (spirolides) or tetrahydropurine neurotoxins (saxitoxin and analogues) produced by various strains of A. ostenfeldii. Instead, the potency of A. ostenfeldii, eliciting immobilization and lytic species-specific responses in potential predators and competitors, is consistent with the existence of an allelochemical mechanism unrelated to the bioactivity of known phycotoxins of the genus Alexandrium.

INTRODUCTION

The marine dinoflagellate Alexandrium (Halim) Balech is perhaps the most intensively studied genus of free-living planktonic dinoflagellates because of the production of potent neurotoxins associated with paralytic shellfish poisoning (PSP) (Anderson, 1998; Cembella, 1998; Taylor and Fukuyo, 1998). Among the species of this genus, A. ostenfeldii (Paulsen) Balech & Tangen, originally described from northern Iceland as Goniodoma ostenfeldii (Paulsen, 1904), has been reported from many coastal temperate waters from both hemispheres, but particularly from the north Atlantic (Hansen et al., 1992; Balech, 1995). The species is also prominent in New Zealand waters (Mackenzie et al., 1996). Alexandrium ostenfeldii warrants special attention within the genus because of numerous peculiarities. For example, A. ostenfeldii is the only species of Alexandrium within which food vacuoles containing ciliates or phytoplankton cells have been observed (Jacobson and Anderson, 1996), occasionally in quite high abundance (up to 12% of cells). Production of PSP toxins was demonstrated for certain A. ostenfeldii strains (Hansen et al., 1992; Mackenzie et al., 1996), although wide divergence among populations from different geographical regions indicates that toxigenic capacity is unusual rather than universal (Cembella, 1998). Alexandrium ostenfeldii was also identified as the primary, if not unique, source of spirolides, a novel group of macrocyclic imines characterized as marine fast-acting toxins (Cembella et al., 2000). Finally, A. ostenfeldii exudates have been shown to be toxic to the tintinnid Favella ehrenbergii (Hansen et al., 1992) and certain strains can cause short-term toxic effects towards heterotrophic dinoflagellates (Tillmann and John, 2002).

Allelochemicals are generally characterized as specific metabolites that stimulate or suppress growth of other organisms, or elicit other physiological responses in target cells. In marine environments, allelochemicals may function in chemical defence as agents capable of incapacitating or even killing competitors and/or deterring grazers (McClintock and Baker, 2001; Cembella, 2003; Legrand et al., 2003; Granéli and Hansen, 2006). Such allelochemicals act as mediators of species-specific biological interactions, including resource competition and predator–prey interactions, and are thus a potentially important adaptive factor in many biotic associations. Most investigations of chemically mediated interactions in the plankton have focused either on algal-algal interactions (traditionally defined as allelopathy, see Legrand et al., 2003) or on grazer inhibition-deterrence properties of algal–borne secondary metabolites (Wolfe, 2000). Negative effects on both autotrophic and heterotrophic planktonic protists have rarely been simultaneously analysed.

Lytic activity of extracellular metabolites and other negative effects upon other microalgae (Blanco and Campos, 1988; Arzul et al., 1999; Fistarol et al., 2004b) and towards heterotrophic protists (Hansen, 1989; Hansen et al., 1992; Matsuoka et al., 2000; Tillmann and John, 2002; Fistarol et al., 2004b) is rather widely expressed within the genus Alexandrium. On the basis of early experiments with Alexandrium tamarense (Hansen, 1989), deleterious effects on the tintinnid F. ehrenbergii were provisionally attributed to the presence of extracellular PSP toxins. Yet these compounds were not specifically measured in the medium and the magnitude of the response was poorly correlated with cellular PSP toxin content. Further experiments with A. ostenfeldii from Denmark with the same tintinnid species (Hansen et al., 1992) showed a striking avoidance behaviour (retrograde swimming) of the tintinnid against dinoflagellate cells that contained only very low cell concentrations of PSP toxins. In general, allelochemical interactions of Alexandrium cells against other protists have been attributable to extracellular unknown substances but that were clearly independent of PSP toxin content (Tillmann and John, 2002; Fistarol et al., 2004b).

Chemical analysis of A. ostenfeldii strains and field populations from Nova Scotia (Cembella et al., 2000) and clonal cultures from Limfjord, Denmark (MacKinnon et al., 2006), has now confirmed unique spirolide toxin profiles among these strains. This provokes the question of whether or not the observed allelochemical response is related to the presence of spirolides in the cells or in the culture medium.

In the present study, our objective is to investigate the allelochemical potency of A. ostenfeldii by analysing and quantifying the effects of exposure to cells and culture medium of this dinoflagellate towards a variety of protistan species. To evaluate the potential allelochemical role of spirolides in interactions with these protists, we selected A. ostenfeldii strains that produced a range of spirolide concentrations and originated from different geographical regions.

METHOD

Culture of Alexandrium strains and target species

A stock culture of A. ostenfeldii strain BAHME136, originally isolated by N. Berkett in Timaru (New Zealand), was obtained from the BAH Culture Collection on Sylt, Germany (via Malte Elbrächter). Strain K-0287, isolated by Per Juel Hansen from Limfjord (Denmark), was provided by the Copenhagen University Culture Collection (via Øjvind Moestrup). The AOSH1 strain of A. ostenfeldii, isolated from Ship Harbour, Nova Scotia, Canada, originated from the IMB Culture Collection, NRC, Halifax, Canada (via Nancy Lewis). All three strains were grown as unialgal cultures with sterile technique on K growth medium (Keller et al., 1987) as a nutrient enrichment to North Sea seawater from Helgoland, Germany, supplemented with selenite (Dahl et al., 1989). Experimental cultures were maintained in 500 mL Erlenmeyer flasks under controlled conditions at 15°C with artificial Cool-White fluorescent light at a photon flux density of 100 µmol m−2 s−1 on a 16 : 8 h light–dark photocycle. Culture sub-samples for experiments were obtained from exponential growth phase. Prior to sub-sampling by pipette, culture flasks were shaken gently by hand to achieve a homogenous cell distribution.

The heterotrophic dinoflagellate Oxyrrhis marina strain B21.89, isolated from Helgoland and obtained from the Göttingen Culture Collection, was grown with the chlorophyte Dunaliella salina as food. Stock cultures were maintained at 20°C on ambient room light. Cultures of O. marina used in the experiments were grown to high cell concentrations (>3 × 103 cells mL−1) until they became almost deprived of food. The heterotrophic dinoflagellate Amphidinium crassum was isolated from the Mediterranean Sea at Barcelona Harbour, Spain. Stock cultures were grown at 15°C at a low photon flux density of 20 µmol m−2 s−1 with the cryptophyte Rhodomonas sp. as food. The oligotrich ciliate Strombidium sp. was isolated from seawater originating from Helgoland. The ciliate Rimostrombidium caudatum originated from a seawater sample from the Wadden Sea at Sylt, Germany. Stock cultures of both ciliates were grown at 15°C at a photon flux density of 20 µmol m−2 s−1 with the chrysophyte Isochrysis galbana as food. All heterotrophic cultures were transferred once or twice per week to fresh food medium.

Autotrophic test algae used in the experiments (Rhodomonas sp., D. salina, Thalassiosira weissflogii, Emiliania huxleyi, Prymnesium parvum, Scrippsiella trochoidea) were taken from the AWI Culture Collection, Bremerhaven, Germany. Cultures were grown in f/2 medium at 15°C at a photon flux density of 100 µmol m−2 s−1 and transferred weekly to fresh medium.

Interactions between A. ostenfeldii and O. marina

The effects of A. ostenfeldii on the protistan grazer O. marina were quantified by previously described methods (Tillmann and John, 2002). Briefly, 0.5 mL aliquots of a dense O. marina culture (3–6 × 103 cells mL−1) were mixed with 1 mL of A. ostenfeldii culture in wells of a 24-well plate (Corning Inc., Corning, USA). After 1 h exposure at 15°C and under at photon flux density of 100 µmol m−2 s−1, the number of motile O. marina cells was estimated by a droplet-counting procedure: 100 µL of cell suspension was separated into 25–30 small droplets in a Petri dish and the number of swimming O. marina cells was counted under a stereo-microscope. Triplicate preparations of K-medium, instead of the A. ostenfeldii culture, served as a control. After 3 h exposure, 0.5 mL of culture were transferred into small settling chambers (diameter 10 mm), fixed with Lugoĺs iodine solution and counted on an inverted microscope at 200 × magnification. In the fixed samples, O. marina was only scored if the normal cell shape was still visible.

In a preliminary experiment, a single batch culture of A. ostenfeldii AOSH1 was set up and tested in triplicate with sub-samples as described at several different points of the growth curve representing different cell concentrations (Experiment 1a). Subsequently, the entire experiment was repeated using another batch culture (Experiment 1b). In another experiment, three different A. ostenfeldii strains (AOSH1, BAHME136, K-0287) were tested for the algal cell concentration-dependence of the immobilization and lytic effect on O. marina (Experiment 2). Triplicate batch cultures were set up for each A. ostenfeldii strain and were tested as described at several different points of the growth curve representing different cell concentrations (NB: cells taken at different stages of the growth curve may differ in activity). For each of the triplicate batch cultures, immobilization and lysis were calculated as the mean of duplicate sub-samples. Results were always expressed as percentage of motile cells or intact cells compared with the control. Concurrent with samples taken for the Oxyrrhis test, sub-samples were collected for cell counts and for spirolide analysis. For all experiments, percentages of immobilization and lysis were transformed to probits (Hewlett and Placklet, 1979). Values of EC50, defined as the Alexandrium cell concentration inducing 50% immobilization after 1 h of incubation or 50% cell lysis after 3 h incubation with the target species, were calculated using linear regression analysis of probits against log-transformed A. ostenfeldii cell concentrations.

Effects of A. ostenfeldii on other protists

All subsequently described experiments were carried out on individual batch cultures of each A. ostenfeldii strain, which were arbitrarily diluted about every 2 weeks. Samples used for the experiments were always taken at mid- to late-exponential growth phase.

The autotrophic dinoflagellate S. trochoidea was chosen to study possible immobilization effects. Since clear differentiation between A. ostenfeldii and S. trochoidea cells is rather difficult at low magnification with a stereo-microscope, A. ostenfeldii culture filtrate, instead of whole cell cultures, was tested. Filtrate from A. ostenfeldii was obtained by gently pouring culture through a 10 µm Nitex mesh. All A. ostenfeldii cells were removed, as confirmed by microscopic examination. Two millilitres of a mixture of A. ostenfeldii filtrate (volume corresponding to a final “cell” concentration of 1.5 × 103 mL−1) and S. trochoidea (final concentration: 1.0 × 103 cells mL−1) were prepared in 20 mL glass vials. K-medium served as control. After 1, 3 and 24 h incubation at 15°C and under a photon flux density of 100 µmol m−2 s−1, the number of motile S. trochoidea cells in 200 µL was counted as outlined above. Results were calculated as percentage of control. In addition, 0.5 mL was fixed after 3 h incubation and counted on an inverted microscope.

Possible lytic effects against a range of different protist species were investigated with triplicate 3 mL mixtures of A. ostenfeldii culture (final concentration ca. 1.3–1.5 × 103 cells mL−1) and different target species in 20 mL glass-vials, with K-medium as control. Table I gives an overview of the target species tested, the final cell concentration of A. ostenfeldii and the respective target species. After 3 and 24 h incubation at 15°C and under at photon flux density of 100 µmol m−2 s−1, 0.5 mL (for Rhodomonas sp., D. salina, P. parvum and E. huxleyi), 1 mL (for A. crassum, Strombidium sp. and T. weissflogii) or 2 mL (for R. caudatum) of culture were fixed and inspected under an inverted microscope. Cell counts comprised either the whole sub-sample volume or a volume corresponding to at least 800 cells in the control. In addition to counting the A. ostenfeldii cells, the number of intact cells of the target species (see Fig. 1 and description in the text) was scored in order to quantify lytic effects.

Fig. 1.

Microscopic observations of lytic effects of A. ostenfeldii on different target species. Photos in the left row (A, D, G, J, L, O and R) show unaffected cells from control treatment. (AC) Rhodomonas sp.; (B) beginning of cell lysis, note the formation of blisters; (C) completely lysed cell. (DF) Dunaliella salina; (E) elongated cell frequently found after short-term incubation (hours); (F) completely lysed cell. (GI) Thalassiosira weissflogii; (H) cell with partly granulated cytoplasm; (I) completely lysed cell. (J and K) Scrippsiella trochoidea; (K) temporary cysts with empty theca. (LN) Oxyrrhis marina (round dark particles are remains of Dunaliella salina used as food in O. marina cultures; (M) round cell; (N) lysed cell. (OQ) Amphidinium crassum; (P) round cell; (Q) lysed cell. (RT) Rimostrombidium caudatum; (S) round cell; (T) lysed cell. Scale bar represents 10 µm. Note that the sequence of prints is not a true time course.

Fig. 1.

Microscopic observations of lytic effects of A. ostenfeldii on different target species. Photos in the left row (A, D, G, J, L, O and R) show unaffected cells from control treatment. (AC) Rhodomonas sp.; (B) beginning of cell lysis, note the formation of blisters; (C) completely lysed cell. (DF) Dunaliella salina; (E) elongated cell frequently found after short-term incubation (hours); (F) completely lysed cell. (GI) Thalassiosira weissflogii; (H) cell with partly granulated cytoplasm; (I) completely lysed cell. (J and K) Scrippsiella trochoidea; (K) temporary cysts with empty theca. (LN) Oxyrrhis marina (round dark particles are remains of Dunaliella salina used as food in O. marina cultures; (M) round cell; (N) lysed cell. (OQ) Amphidinium crassum; (P) round cell; (Q) lysed cell. (RT) Rimostrombidium caudatum; (S) round cell; (T) lysed cell. Scale bar represents 10 µm. Note that the sequence of prints is not a true time course.

Table I:

Cell concentrations of A. ostenfeldii strains and of target species as used in the short-term exposure experiments

A. ostenfeldii strain Cell concentration (mL−1Target species Target concentration (mL−1
K-0287 1.26 × 103 Rhodomonas sp. 2.01 × 104 
BAHME136 1.35 × 103 Rhodomonas sp. 2.01 × 104 
AOSH1 1.25 × 103 Rhodomonas sp. 2.01 × 104 
K-0287 1.32 × 103 D. salina 1.63 × 104 
BAHME136 1.31 × 103 D. salina 1.63 × 104 
AOSH1 1.30 × 103 D. salina 1.63 × 104 
K-0287 1.39 × 103 P. parvum 1.84 × 104 
BAHME136 1.43 × 103 P. parvum 1.84 × 104 
AOSH1 1.55 × 103 P. parvum 1.84 × 104 
K-0287 1.60 × 103 A. crassum 381 
BAHME136 1.50 × 103 A. crassum 381 
AOSH1 1.39 × 103 A. crassum 381 
K-0287 1.38 × 103 Strombidium sp. 144 
BAHME136 1.42 × 103 Strombidium sp. 144 
AOSH1 1.37 × 103 Strombidium sp. 144 
K-0278 1.43 × 103 R. caudatum 92 
BAHME136 1.43 × 103 R. caudatum 92 
AOSH1 1.51 × 103 R. caudatum 92 
K-0287 1.46 × 103 E. huxleyi 1.09 × 105 
BAHME136 1.36 × 103 E. huxleyi 1.09 × 105 
AOSH1 — — — 
K-0287 1.43 × 103 T. weissflogii 1.18 × 104 
BAHME136 1.43 × 103 T. weissflogii 1.18 × 104 
AOSH1 — — — 
A. ostenfeldii strain Cell concentration (mL−1Target species Target concentration (mL−1
K-0287 1.26 × 103 Rhodomonas sp. 2.01 × 104 
BAHME136 1.35 × 103 Rhodomonas sp. 2.01 × 104 
AOSH1 1.25 × 103 Rhodomonas sp. 2.01 × 104 
K-0287 1.32 × 103 D. salina 1.63 × 104 
BAHME136 1.31 × 103 D. salina 1.63 × 104 
AOSH1 1.30 × 103 D. salina 1.63 × 104 
K-0287 1.39 × 103 P. parvum 1.84 × 104 
BAHME136 1.43 × 103 P. parvum 1.84 × 104 
AOSH1 1.55 × 103 P. parvum 1.84 × 104 
K-0287 1.60 × 103 A. crassum 381 
BAHME136 1.50 × 103 A. crassum 381 
AOSH1 1.39 × 103 A. crassum 381 
K-0287 1.38 × 103 Strombidium sp. 144 
BAHME136 1.42 × 103 Strombidium sp. 144 
AOSH1 1.37 × 103 Strombidium sp. 144 
K-0278 1.43 × 103 R. caudatum 92 
BAHME136 1.43 × 103 R. caudatum 92 
AOSH1 1.51 × 103 R. caudatum 92 
K-0287 1.46 × 103 E. huxleyi 1.09 × 105 
BAHME136 1.36 × 103 E. huxleyi 1.09 × 105 
AOSH1 — — — 
K-0287 1.43 × 103 T. weissflogii 1.18 × 104 
BAHME136 1.43 × 103 T. weissflogii 1.18 × 104 
AOSH1 — — — 

Effect of bacteria and ageing of filtrate

To examine the potential role of bacteria in “ageing” of filtrate and allelochemical effects, culture filtrate was produced either by sieving through 10 µm mesh (removing A. ostenfeldii cells but with bacteria included) or by filtration through a 0.2 µm membrane (both A. ostenfeldii and bacterial cells excluded). Accordingly, 20 mL of exponentially growing A. ostenfeldii strain BAHME136 (at 4.2 × 103 cells mL−1) or 20 mL K-medium as control were stored in 20 mL glass vials. In parallel, 20 mL each of algal culture were gently filtered through either 10 µm Nitex mesh or 0.2 µm membrane filters (Sartorius Minisart) and stored in 20 mL glass vials. All vials were kept at 15°C at a photon flux density of 100 µmol m−2 s−1 on light–dark cycle of 16 : 8 h. Immediately after filtration (t = 0), as well as after 1, 2 and 3 days of ageing, triplicates of 1 mL of protist culture, K-medium (control) or the two different types of filtrate were added to new glass vials, mixed with 2 mL fresh K-medium and spiked with 0.3 mL of Rhodomonas sp., resulting in final cell concentrations of 1.4 × 103 and 2.0 × 104 mL−1 for A. ostenfeldii and Rhodomonas sp., respectively.

After 3 h incubation under standard conditions as above, 0.5 mL sub-samples were transferred into small settling chambers (diameter 10 mm), fixed with Lugoĺs solution and counted on an inverted microscope.

Allelochemical effect of different target cell concentrations

A series of equivalent inocula of A. ostenfeldii BAHME136 cells (final concentration: 500 mL−1) or seawater as control were exposed to five different concentrations of Rhodomonas sp. (0.7, 3.3, 27, 97 and 178 × 103 cells mL−1) prepared by dilution. To rule out any effect of adding different amount of dissolved organic carbon or bacteria when adding different volumes of Rhodomonas sp., dilutions were made from filtrate of the same culture of Rhodomonas sp. obtained by gentle gravity filtration of 100 mL of culture through a 3 µm pore size polycarbonate filter. Three replicates and three controls were set up for each algal concentration. After 3 h incubation under standard conditions, varying amounts were fixed with Lugoĺs iodine solution to obtain > 700 Rhodomonas cells to count in the entire sample volume of the control sample. Only Rhodomonas cells with a normal cell shape were counted. Percent mortality was calculated by comparing cell numbers of samples and seawater control.

Pulse-amplitude-modulation fluorometry measurements

Potential sub-lethal effects of A. ostenfeldii allelochemicals on the photosynthetic physiology of autotrophic target species were evaluated by pulse-amplitude-modulation (PAM) fluorometry. Aliquots of batch cultures of all three A. ostenfeldii strains (concentration ca. 5 × 103 cells mL−1) were sub-sampled for cell enumeration and then gently filtered through 10 µm Nitex mesh. Twenty millilitre mixtures of filtrate (corresponding to a final “cell” concentration of 2 × 103 mL−1), or K-medium as control, and different target species (final concentrations: Rhodomonas sp.: 2.32 × 104 mL−1; D. salina: 2.96 × 104 mL−1; T. weissflogii: 1.04 × 104 mL−1; E. huxleyi: 1.07 × 105 mL−1; S. trochoidea: 2.36 × 103 mL−1) were prepared in 20 mL glass vials and kept in the dark at 15°C. From these vials, sub-samples were removed after 3 h incubation and fixed in Lugoĺs solution for cell enumeration. In addition, immediately after mixing (t = 0) and at 15, 30, 60 and 180 min, 2 mL of culture (in triplicate) were pipetted into a quartz cuvette and the photosynthetic response was analysed by PAM fluorometry (Schreiber et al., 1986; Juneau and Harrison, 2005). The quantum yield was determined in triplicate for each treatment with the Xe-PAM (Heinz Walz GmbH, Germany) equipped with a temperature control unit and a magnetic stirrer. The quantum yield was calculated from fluorescence readings of illuminated samples as the ratio of variable fluorescence (Fv) to the maximum fluorescence (Fm), which can be quantitatively related to photochemical efficiency. In the experiment with T. weissflogii and E. huxleyi, additional samples were taken after 24 h incubation in the dark and fixed for cell counts.

Spirolide analysis by liquid chromatography-mass spectrometry

The micro-extraction procedure for spirolides from A. ostenfeldii cells has been thoroughly described and validated in a previous study (Cembella et al., 1999). For each stage of the growth cycle, 2 mL samples (n = 3) of cell suspension were harvested from each of the triplicate batch cultures and filtered by centrifugation (ca. 1 min at 2000 g) through a 0.45 µm spin filter (Ultrafree-MC, Durapore Membrane, Millipore, Bedford, MA, USA). One millilitre of filtrate was retained for direct analysis after addition of 1.0 mL methanol. Spirolides were extracted from the filtered cells by addition of 1.0 mL methanol to the cartridge filter and then by centrifugation after incubation on the filter for 10 min (Maclean et al., 2003).

Spirolides were analysed by liquid chromatography-mass spectrometry, according to the method of Cembella et al. (Cembella et al., 2000), on an Agilent (Palo Alto, CA, USA) HP1100 LC system with detection by a Perkin-Elmer SCIEX (Concord, Ontario, Canada) API-165 atmospheric pressure ionization mass spectrometer. Chromatographic separation was achieved on a 50 × 2.1 mm i.d. analytical column packed with 3 µm Hypersil-BDS octylsilica (C8) packing and 0.2 mL min−1 of mobile phase composed of acetonitrile/water (35 : 65, v/v) with 2 mM ammonium formate and 50 mM formic acid. Analyses were conducted in positive ion electrospray mode using a Turbo IonSpray interface, with drying nitrogen maintained at 350°C and selected ion monitoring of the [M + H]+ ions. Calibration was performed against standards of pure spirolides B and D prepared from scallop viscera (Hu et al., 1995). Relative molar response factors for other spirolides were assumed to be the same.

RESULTS

Effects on O. marina

Exposure to A. ostenfeldii caused cells of the heterotrophic dinoflagellate O. marina to lose motility and then to become rounded before finally lysing (Fig. 1L–N). As quantified along the growth curve in batch culture, immobilization and lytic effects were strongly dependent on the A. ostenfeldii cell concentrations. When first tested on two different batch cultures of A. ostenfeldii AOSH1 (Experiment 1), negative short-term effects were observed at cell concentrations of > 100 cells mL−1 (Fig. 2). Growth rates of the two batch cultures were statistically equal (Culture 1 : µ = 0.106 day−1; Culture 2: µ = 0.102 day−1; ANOVA: P > 0.5); however, on a per-cell basis, the potency of Culture 2 was slightly higher than that of Culture 1. The corresponding EC50 concentrations are summarized in Table II. In Experiment 2, immobilization and lytic effects of three different strains of A. ostenfeldii were compared and the growth rate (µ) was calculated from mean cell numbers. Growth rates of the three replicate batch cultures for each strain were not significantly different (ANOVA: P > 0.1 for all species) but strain K-0287 (µ = 0.18 day−1) grew significantly faster (pair-wise ANOVA: P < 0.01) than strains BAHME136 (µ = 0.14 day−1) and AOSH1 (µ = 0.13 day−1). Again, the observed effects showed a strong dependence on the A. ostenfeldii cell concentration (Fig. 3). The corresponding EC50 values (Table II) ranged from 0.6 to 2 × 103 mL−1. Covariance analysis of the probit regression lines indicated that probit regression lines for cell immobilization and cell lysis did not differ significantly (P > 0.05) except for strain BAHME 136 (P = 0.025). Probit regression lines for immobilization for both AOSH1 cultures in Experiment 1a and b did differ significantly (P = 0.012), whereas cell lysis values were not statistically different (P > 0.05). For both immobilization and cell lysis, probit regression lines of both AOSH1 cultures from Experiment 1 were significantly different (P < 0.05) from that of the AOSH1 culture from Experiment 2. Finally, probit regression lines for both immobilization and cell lysis differed significantly (P < 0.05) between all strains in Experiment 2.

Fig. 2.

Two different batch cultures of A. ostenfeldii (strain AOSH1) grown at 15°C and photon flux density of 100 µmol m−2 s−1. (A) Growth curves. (B) Percentage of motile O. marina cells after 1 h exposure (live counts) as a function of algal cell concentration when measured along the growth curve. (C) Percentage of intact O. marina cells after 3 h incubation (fixed cell counts) as a function of different cell concentrations when measured along the growth curve. Results in (B) and (C) are expressed as triplicate mean ± 1 SD.

Fig. 2.

Two different batch cultures of A. ostenfeldii (strain AOSH1) grown at 15°C and photon flux density of 100 µmol m−2 s−1. (A) Growth curves. (B) Percentage of motile O. marina cells after 1 h exposure (live counts) as a function of algal cell concentration when measured along the growth curve. (C) Percentage of intact O. marina cells after 3 h incubation (fixed cell counts) as a function of different cell concentrations when measured along the growth curve. Results in (B) and (C) are expressed as triplicate mean ± 1 SD.

Fig. 3.

Percentage of motile O. marina cells after 1 h incubation (open circles) and intact O. marina after 3 h incubation (filled circles) as a function of algal cell concentration for (A) A. ostenfeldii strain K-0287, (B) strain BAHME 136 and (C) strain AOSH1 when measured along the growth curve. Results are expressed as triplicate mean ± 1 SD.

Fig. 3.

Percentage of motile O. marina cells after 1 h incubation (open circles) and intact O. marina after 3 h incubation (filled circles) as a function of algal cell concentration for (A) A. ostenfeldii strain K-0287, (B) strain BAHME 136 and (C) strain AOSH1 when measured along the growth curve. Results are expressed as triplicate mean ± 1 SD.

Table II:

Cell concentrations of three different A. ostenfeldii strains evoking effect on 50% of target cells of O. marina (EC50) for immobilisation (1 h incubation) or lysis (3 h incubation)

Experiment number A. ostenfeldii strain EC50 (immobilization) (cells mL−1EC50 (lysis) (cells mL−1
1a AOSH1 0.4 × 103 0.3 × 103 
1b AOSH1 0.2 × 103 0.3 × 103 
K-0287 2.0 × 103 1.9 × 103 
BAHME136 0.9 × 103 0.6 × 103 
AOSH1 1.2 × 103 0.9 × 103 
Experiment number A. ostenfeldii strain EC50 (immobilization) (cells mL−1EC50 (lysis) (cells mL−1
1a AOSH1 0.4 × 103 0.3 × 103 
1b AOSH1 0.2 × 103 0.3 × 103 
K-0287 2.0 × 103 1.9 × 103 
BAHME136 0.9 × 103 0.6 × 103 
AOSH1 1.2 × 103 0.9 × 103 

Effects on S. trochoidea

Extracellular bioactive compounds from A. ostenfeldii rapidly immobilized cells of the autotrophic dinoflagellate S. trochoidea (Fig. 4). Within 1 h after mixing, motility of S. trochoidea cells was drastically decreased for all three A. ostenfeldii strains. However, in contrast to the findings with O. marina, cell counts of fixed samples after 3 h of incubation showed no signs of cell lysis. Motility of S. trochoidea recorded after 3 and 24 h incubation clearly show a “recovery” of the autotrophic dinoflagellate. After 24 h, motility again reached high values (70–90%) for all three A. ostenfeldii strains.

Fig. 4.

Effects of A. ostenfeldii filtrate on S. trochoidea. (A): Percent of motile S. trochoidea cells (compared with seawater control) when incubated for 1, 3 or 24 h with culture filtrate of A. ostenfeldii strain K-0287, BAHME136 or AOSH1. (B): Number of intact S. trochoidea cells after 3 h incubation (fixed samples). Results expressed as triplicate mean ± 1 SD. Asterisks above bars indicate values significantly (P < 0.01) different from control.

Fig. 4.

Effects of A. ostenfeldii filtrate on S. trochoidea. (A): Percent of motile S. trochoidea cells (compared with seawater control) when incubated for 1, 3 or 24 h with culture filtrate of A. ostenfeldii strain K-0287, BAHME136 or AOSH1. (B): Number of intact S. trochoidea cells after 3 h incubation (fixed samples). Results expressed as triplicate mean ± 1 SD. Asterisks above bars indicate values significantly (P < 0.01) different from control.

Effects on other target species

The toxic effects of A. ostenfeldii whole-cell cultures against a range of different target species are shown in Fig. 5. Microscopic observations revealed that the cryptophyte Rhodomonas sp. was particularly heavily affected when mixed with A. ostenfeldii cells. Cells of Rhodomonas sp. formed blisters and subsequently lysed (Fig. 1A–C). After 3 h incubation, almost no intact Rhodomonas cells were found after exposure to A. ostenfeldii strains BAHME136 and AOSH1, whereas for strain K-0287 the number of intact Rhodomonas cells was reduced to 4.7 × 103 cells mL−1, compared with 22 × 103 cells mL−1 in the control sample (Fig. 5A). After 24 h incubation with strain K-0287, the number of intact Rhodomonas cells further declined to 0.6 × 103 mL−1.

Fig. 5.

Numbers of intact cells of different target organisms exposed to fresh culture medium (control) or to whole-cell culture of three A. ostenfeldii strains (K-0287, BAHME136, AOSH1). Samples were taken after 3 h (black bars) and in some cases after 24 h (white bars) after incubation. (A) Rhodomonas sp.; (B) D. salina; (C) T. weissflogii; (D) A. crassum; (E) R. caudatum; (F) Strombidium sp.; (G) P. parvum; (H) E. huxleyi. Results expressed as triplicate mean ± 1 SD. Asterisks above bars indicate values significantly (P < 0.01) different from control.

Fig. 5.

Numbers of intact cells of different target organisms exposed to fresh culture medium (control) or to whole-cell culture of three A. ostenfeldii strains (K-0287, BAHME136, AOSH1). Samples were taken after 3 h (black bars) and in some cases after 24 h (white bars) after incubation. (A) Rhodomonas sp.; (B) D. salina; (C) T. weissflogii; (D) A. crassum; (E) R. caudatum; (F) Strombidium sp.; (G) P. parvum; (H) E. huxleyi. Results expressed as triplicate mean ± 1 SD. Asterisks above bars indicate values significantly (P < 0.01) different from control.

The chlorophyte D. salina was less affected than Rhodomonas (Fig. 5B). No apparent cell lysis occurred after 3 h incubation, but a large proportion of D. salina cells expressed a conspicuously elongated cell form in mixtures with A. ostenfeldii BAHME136 and AOSH1 (Fig. 1E). Nevertheless, because of many transitional forms between normal ellipsoid and elongated cells, all these cells were scored as “normal” for cell counts. After 24 h incubation, numbers of D. salina cells had increased in the control samples due to cell division. Growth of D. salina was also obvious in the K-0287 treatment, although final cell numbers were slightly, but significantly (t-test; P < 0.01), lower than in the control. In contrast, the decline of cell numbers in the BAHME136 and AOSH1 treatments clearly indicated cell lysis (Fig. 5B).

With the diatom T. weissflogii as target, lytic effects of strain BAHME136 were obvious after 24 h incubation (note that no data are available for strain AOSH1), whereas cell numbers of T. weissflogii increased in both control and K-0287 treatments (Fig. 5C). After 3 h incubation, cells of T. weissflogii incubated with A ostenfeldii BAHME136 appeared to be in poor condition, as indicated by morphological changes. In a number of cells (not quantified), parts of the cell content was conspicuously granular and/or somewhat concentrated at both valvar sides, leaving a hyaline band in the centre (Fig. 1H). All these transitional cells were scored as “normal”; only cells with completely granular cytoplasm, as well as broken or empty frustules, were scored as abnormal.

For the heterotrophic dinoflagellate A. crassum, mortality was almost 100% after only 3 h incubation with all three A. ostenfeldii strains (Fig. 5D). The same was true for the ciliate species R. caudatum (Fig. 5E). To rule out potential negative effects of elevated pH, R. caudatum was simultaneously incubated with a dense culture of the non-toxic dinoflagellate S. trochoidea as control. Under these conditions, cell numbers after 3 h incubation were the same as in the control (results not shown). Upon mixing with A. ostenfeldii, R. caudatum cells immediately started to swim in a conspicuously irregular manner; helical swimming with an unusual high speed was interrupted by shorter periods of backwards swimming. In contrast, no negative short-term effect towards the oligotrich ciliate Strombidium sp. could be detected. Microscopic examinations carried out at irregular intervals during the incubation period showed no signs of altered swimming behaviour of the ciliate in the presence of any of the three A. ostenfeldii strains. In addition, cell numbers were unchanged relative to the control (Fig. 5F). The same holds true for the prymnesiophyte P. parvum. Here again, there was no detectible cell lysis, and cell numbers in both control and A. ostenfeldii treatments remained constant (Fig. 5G). In this experiment, the prymnesiophyte E. huxleyi was also not negatively affected. Cell numbers for both strains tested (note that no data are available for strain AOSH1) increased after 24 h incubation. Final cell number for strain BAHME136, although slightly lower than the control (Fig. 5H) was not significantly different from the control (t-test, P > 0.01).

Ageing culture filtrate

When measured immediately after filtration, mortality of Rhodomonas sp. was 100% for exposure to both 10 µm Nitex mesh and 0.2 µm membrane filtrates, as well as for the whole algal culture preparation (Fig. 6). Mortality of Rhodomonas sp. stayed high (100%) throughout the whole dinoflagellate culture treatment, i.e. over the entire time course of ageing (3 days). In contrast, mortality of Rhodomonas cells, when incubated with culture filtrate, declined with time of ageing. After 3 days of ageing, no mortality of Rhodomonas cells was detected. Reduction in mortality with ageing time was identical for both <10 and <0.2 µm filtrates.

Fig. 6.

Effect of ageing culture filtrate of A. ostenfeldii strain BAHME136. Rhodomonas sp. mortality (% of control) was recorded after 0, 1, 2 and 3 days of ageing of whole algal culture, of filtrate from 10 µm gauze filtration or of filtrate from by 0.2 µm pore-size filtration. Results expressed as triplicate mean ± 1 SD.

Fig. 6.

Effect of ageing culture filtrate of A. ostenfeldii strain BAHME136. Rhodomonas sp. mortality (% of control) was recorded after 0, 1, 2 and 3 days of ageing of whole algal culture, of filtrate from 10 µm gauze filtration or of filtrate from by 0.2 µm pore-size filtration. Results expressed as triplicate mean ± 1 SD.

Effect of different target concentration

The percentage cell mortality of Rhodomonas sp. was clearly decreased by increasing the target cell concentration (Fig. 7). When exposed to filtrate from A. ostenfeldii BAHME136, corresponding to a concentration of 500 cells mL−1, mortality was around 80% for the two lowest Rhodomonas cell concentrations (≤3 × 103 mL−1). With increasing Rhodomonas cell concentration, mortality declined almost linearly to about 16% at a concentration of 1.8 × 105Rhodomonas cells mL−1.

Fig. 7.

Mortality of Rhodomonas sp. (% of control) as a function of initial cell concentration. Different cell concentrations of Rhodomonas sp. were exposed to cells of A. ostenfeldii BAHME136 (final concentration: 500 mL−1) for 3 h. Data points represent mean values (n = 3) with error bars (±1SD).

Fig. 7.

Mortality of Rhodomonas sp. (% of control) as a function of initial cell concentration. Different cell concentrations of Rhodomonas sp. were exposed to cells of A. ostenfeldii BAHME136 (final concentration: 500 mL−1) for 3 h. Data points represent mean values (n = 3) with error bars (±1SD).

PAM measurements

The time course of fluorescence yield as an indicator of the photosynthetic performance is shown in Fig. 8. For Rhodomonas sp. spiked with A. ostenfeldii filtrate, the fluorescence yield of Rhodomonas cells in the mixtures with BAHME136 and AOSH1 filtrate was drastically decreased after 15 min incubation, but for K-0287 filtrate, a drop in fluorescence yield was apparent only after 3 h. The pattern of cell lysis corresponded closely to the respective drop in fluorescence yield among the filtrate exposure treatments for the three A. ostenfeldii strains (Fig. 9A). In comparison, the fluorescence yield of D. salina was quite refractory to filtrate effects over the time course of the experiment (Fig. 8B).

Fig. 8.

Fluorescence yield (Fv/Fm) of different target algae as a function of incubation time with fresh culture medium (control, filled circles) or 10 µm gauze filtrate of A. ostenfeldii K-0287 (open circles), strain BAHME136 (filled squares) or strain AOSH1 (open squares). (A) Rhodomonas sp.; (B) D. salina; (C) T. weissflogii; (D) E. huxleyi; (E) S. trochoidea. Data points represent mean values (n = 3) with error bars (±1SD).

Fig. 8.

Fluorescence yield (Fv/Fm) of different target algae as a function of incubation time with fresh culture medium (control, filled circles) or 10 µm gauze filtrate of A. ostenfeldii K-0287 (open circles), strain BAHME136 (filled squares) or strain AOSH1 (open squares). (A) Rhodomonas sp.; (B) D. salina; (C) T. weissflogii; (D) E. huxleyi; (E) S. trochoidea. Data points represent mean values (n = 3) with error bars (±1SD).

Fig. 9.

Number of intact cells from the PAM fluorometry experiment for (A) Rhodomonas sp., (B) D. salina, (C) T. weissflogii, (D) E. huxleyi and (E) S. trochoidea. Black bars: cell counts after 3 h incubation. Open bars (for T. weissflogii and E. huxleyi): cell counts after 24 h incubation. Dotted area in (E) represent fraction of temporary cysts. Data represent mean value (n = 3) with error bars (±1 SD). Asterisks above bars indicate values significantly (P < 0.01) different from control.

Fig. 9.

Number of intact cells from the PAM fluorometry experiment for (A) Rhodomonas sp., (B) D. salina, (C) T. weissflogii, (D) E. huxleyi and (E) S. trochoidea. Black bars: cell counts after 3 h incubation. Open bars (for T. weissflogii and E. huxleyi): cell counts after 24 h incubation. Dotted area in (E) represent fraction of temporary cysts. Data represent mean value (n = 3) with error bars (±1 SD). Asterisks above bars indicate values significantly (P < 0.01) different from control.

For the diatom T. weissflogii, the slight initial increase in fluorescence yield was followed a slight depression of the yield after 3 h of incubation for all A. ostenfeldii treatments (Fig. 8C). A greater decrease was observed when target cells exposed to filtrate of BAHME136 and AOSH1 than for strain K-0287.

For the prymnesiophyte E. huxleyi, fluorescence yield remained more or less constant over the incubation period (Fig. 8D). Corresponding cell numbers after 3 h incubation were also essentially unaltered (Fig. 9D), but after 24 h, cell numbers of E. huxleyi were reduced for all A. ostenfeldii strains.

The same trend of an unaltered fluorescence yield was observed when S. trochoidea was incubated with A. ostenfeldii filtrates (Fig. 8E). The total number of cells in samples after 3 h incubation was not affected (Fig. 9E), but the majority of cells in all filtrate treatments were round and without thecae, representing temporary cysts (Fig. 1K).

Spirolide composition

Alexandrium ostenfeldii BAHME136 from Timaru, New Zealand, contained no measurable spirolides in the cell fraction (detection limit: <0.01 pg cell−1) nor in the culture medium. The strain K-0287 from Limfjord, Denmark, also produced no detectable spirolides in either the cell fraction or the medium. The AOSH1 strain from Ship Harbour, Nova Scotia, Canada, consistently yielded spirolides from the cells, but total spirolide concentrations varied widely from mean values (n = 9; three cultures × three replicates) of 3.4 ± 3.0 to 63.3 ± 26.2 pg cell−1 over the culture cycle (seven time points) and among batch cultures. In general, the cell spirolide quota tended to be highest towards the end of the 3-week batch culture cycle. Low levels of spirolides (<0.35 ng mL−1), particularly of desmethyl spirolide C, were occasionally detected in the medium but only towards the end of the experiment and at the highest cell concentrations. The toxin profile (% total; mean values: n = 18) was rather stable among replicates through the culture cycle and comprised the following spirolide derivatives (in descending order of relative concentration): desmethyl C (86.0%), desmethyl D (8.0%), C3 (4.7%), C (2.2%) and D3 (0.1%).

DISCUSSION

Our experiments clearly showed that certain strains of A. ostenfeldii produce extracellular compounds that have deleterious effects on various protistan species. Observed negative effects of allelochemicals from A. ostenfeldii include immobilization (e.g. O. marina) and morphological (e.g. D. salina) and/or behavioural (R. caudatum) changes as preliminary stages of cell lysis. These allelochemical effects are important to distinguish from those of physico-chemical factors because a number of autotrophic (Schmidt and Hansen, 2001) and heterotrophic (Pedersen and Hansen, 2003) eukaryotes are negatively affected simply by the high pH often generated in dense cultures. Although we did not continuously control the pH in our study, it is unlikely that elevated pH in Alexandrium cultures contributed to the observed short-term negative effects for the following reasons: (i) some target species (e.g. O. marina) are known to be extremely tolerant to high pH (Pedersen and Hansen, 2003); (ii) sporadic pH measurements in cell suspensions with 1–3 × 103Alexandrium cells mL−1 revealed values of 8.2–8.6, about the same range found in stock cultures of the target species; (iii) some target species were negatively affected by only certain A. ostenfeldii strains, e.g. cultures of strain K-0278 (assumed to have a similar pH) elicited no negative effects and (iv) the sensitive ciliate species R. caudatum was not affected by a dense culture of the non-toxic dinoflagellate S. trochoidea, which again is assumed to have a high pH comparable to that of A. ostenfeldii cultures.

Bacteria may be either directly or indirectly associated with phycotoxin production (Doucette et al., 1998) and their role in allelochemical interactions has often been described. The potential of extracellular bacteria in causing the effects observed for the non-axenic Alexandrium cultures must therefore be considered. As reported earlier (Tillmann and John, 2002), immobilization evoked by different whole-cell Alexandrium cultures (including bacteria) was higher than that produced by the culture filtrate, indicating a sequential release of toxic substances. Yet there was no difference between the effects produced by xenic (<10 µm) versus axenic (<0.2 µm) filtrates. Furthermore, the decrease in toxicity with ageing time was identical for both types of filtrate; it is therefore unlikely that significant extracellular toxins were produced by free bacteria.

Observed negative effects of allelochemicals from A. ostenfeldii include immobilization (e.g. O. marina) and changes in morphology (e.g. D. salina) and/or behaviour (e.g. R. caudatum) as preliminary stages of cell lysis. The autotrophic dinoflagellate S. trochoidea reacts to A. ostenfeldii by formation of ecdysal (temporary) cysts, a typical stress response for many thecate dinoflagellates. Non-motile ecdysal cells are formed by shedding the outer cell layers and jettisoning the flagellae in response to adverse conditions such as ageing of cultures (Jensen and Moestrup, 1997), deficiencies in specific nutrients (Doucette et al., 1989) or changes in temperature (Schmitter, 1979) and salinity. Ecdysal cyst formation may occur as a result of gradual deleterious changes in growth conditions but can also happen suddenly (within minutes) in response to sudden physical–chemical shock or stress (Taylor, 1987). Our observations that S. trochoidea reacts to A. ostenfeldii by formation of ecdysal (temporary) cysts is consistent with recent findings of Fistarol et al. (Fistarol et al., 2004a), who showed that three toxic microalgal species (A. tamarense, Karenia mikimotoi and Chrysochromulina polylepis) induced temporary cyst formation in natural populations of S. trochoidea. The clear increase of motile S. trochoidea vegetative cells after 24 h incubation with A. ostenfeldii filtrate (Fig. 4) indicates that these cysts can readily reconvert to the motile stage and resume mitosis. Since other growth factors remained constant, the most likely explanation for this reversion is time-dependent degradation or inactivation of the allelochemicals that provoked encystment.

The function of ecdysal cysts is unknown but there may be metabolic advantages to sacrificing motility and cell division (but not photosynthetic activity) in the short term. The temporary cysts of S. trochoidea completely retain their photosynthetic capacity, as PAM measurements showed no changes in photochemical yield (Fig. 8E). The allelochemical response of S. trochoidea may represent an inducible defence mechanism in response to deleterious chemical cues, especially those affecting cell membranes. This is in line with some recent findings that indicate that ecdysis significantly increased upon exposure to chemical cues from parasite attack (Toth et al., 2004) or to exudates of various toxic microalgae (Legrand et al., 2003; Fistarol et al., 2004a).

The capacity to express extracellular lytic activity is rather widespread among members of the genus Alexandrium. Negative effects on other algae (Blanco and Campos, 1988; Arzul et al., 1999; Fistarol et al., 2004b) and on heterotrophic protists (Hansen, 1989; Hansen et al., 1992; Matsuoka et al., 2000; Tillmann and John, 2002) have been described for a number of species. In addition, exudates of A. tamarense were shown to affect whole plankton communities by decreasing the growth rate of both autotrophic and heterotrophic protists (ciliates) and thus changes the whole community structure (Fistarol et al., 2004b). The structural and functional elucidation of the allelochemicals produced by Alexandrium species remains elusive, but culture experiments with A. tamarense have shown that the lytic potential was unrelated to the presence or cell quota of PSP toxins (Tillmann and John, 2002).

Extracellular bioactive compounds of A. ostenfeldii affect a wide range of protists, including both autotrophic and heterotrophic species, but the mode of action is still undefined. Microscopic observations of cell blistering, deformation and subsequent lysis (Fig. 1) imply that these toxins act primarily upon the outer cell membrane of the target organisms. Here we have demonstrated that these lytic effects are unrelated to the presence or cell concentration of spirolides. Two A. ostenfeldii strains produced no detectable spirolides in either the cell or the culture medium but, nevertheless, exhibited allelochemical activity. We were somewhat surprised by the lack of spirolides in the Danish strain K-0287 because recent chemical investigations have demonstrated low but consistent spirolide production by certain strains from Limfjord, Denmark (MacKinnon et al., 2006). In any case, the AOSH1 strain from Nova Scotia has consistently yielded high spirolide levels but only in the cell fraction, dominated by desmethyl spirolide C as found previously (Cembella et al., 2000). The low amount of spirolide detected in the culture medium, consisting only of the desmethyl C derivative, and found only in mature cultures at high cell concentrations, provides evidence that spirolides are unlikely to act as exotoxins against predators or competitors. This interpretation is consistent with toxicological observations on spirolides, indicating that they may behave as neurotoxins in mammalian systems (Richard et al., 2001), but they are not necessarily highly cytolytic.

The allelochemicals produced by Alexandrium spp. are thus distinguishable from the known phycotoxins, such as those associated with PSP or spirolide toxicity, which can be vectored through the food web and are often accompanied by broad-based trophodynamic effects. The allelochemical activity of Alexandrium spp. appears to be “non-specific”, but this may only be a reflection of the lack of knowledge regarding the mode of action and cellular targets. Intriguingly, the effects of allelochemicals from Alexandrium spp. resemble the toxic effects of haptophytes, many of which are ichthyotoxic and also produce poorly described cytolytic compounds. Exotoxins of the haptophyte P. parvum provoke cell lysis of autotrophic protists (Granéli and Johansson, 2003b; Skovgaard et al., 2003), as well as heterotrophs (Granéli and Johansson, 2003a; Tillmann, 2003). In this respect, it is interesting that P. parvum seems to be insensitive to Alexandrium toxins at the tested nominal levels, although we cannot rule out that Prymnesium might be affected at higher toxin concentrations. The distribution of Prymnesium is mainly restricted to brackish waters; thus, A. ostenfeldii and Prymnesium blooms rarely if ever co-occur. Consequently, it is unlikely that the insensitivity (or at least low sensitivity) of Prymnesium to Alexandrium allelochemicals is the result of a selective adaptation. We speculate that this insensitivity could result from auto-immunity of Prymnesium against its lytic compounds and thus may reflect either structural or functional similarities of Prymnesium and Alexandrium exotoxins.

Some allelochemicals effectively inhibit photosynthetic electron flow and, consequently, the growth of other photosynthetic organisms (von Ehlert and Jüttner, 1997), whereas growth of non-photosynthetic microorganisms may not be inhibited (Gleason and Baxa, 1986). Our PAM measurements showed no evidence that the allelochemicals produced by A. ostenfeldii caused specific short-term negative effects on the photosynthetic apparatus of other algae. Fluorescence yield of target algae was unaffected during the first 3 h of incubation with A. ostenfeldii filtrates (Figs 8 and 9), whereas the reaction time in response to selective photochemical inhibitors is normally in the range of a few minutes (Sukenik et al., 2002).

Interestingly, the percentage mortality of Rhodomonas sp. was significantly decreased by increasing the target cell concentration (Fig. 7). The inverse relationship between target cell concentration and the magnitude of the response is indicative of a saturation effect, perhaps by removal of allelochemical compounds from the system via binding to the target cells, as has been recently shown for Prymnesium exotoxins (Tillmann, 2003). Culture filtrate was used to dilute Rhodomonas; therefore, target cell concentration, and not differences in dissolved organic carbon or bacterial cell concentration, was responsible for the observed effect. This result underlines the need to standardize toxicity tests applied to A. ostenfeldii, with respect to target cell concentration, to obtain reproducible and comparable results. In addition, it clearly shows the difficulties in comparing the sensitivity of different target organisms tested against widely differing parameters, including cell concentrations, cell size and surface area to volume ratios. From an ecological point of view, this result shows that the actual density of accompanying protist (or probably density of all absorbing particles) may largely influence the effectiveness of extracellular allelochemicals.

There is no apparent simple and universal relationship between extracellular concentration of lytic compounds and A. ostenfeldii cell number. Although all three A. ostenfeldii strains used in these experiments produced lytic allelochemicals, there were clear differences in toxicity among them that were independent of cell concentration and culture age. Furthermore, for strain AOSH1, the EC50 values for negative effects on O. marina varied even between different batch cultures grown separately at different times. Finally, strain BAHME136 was originally reported (Tillmann and John, 2002) to be only weakly toxic towards O. marina at a cell concentration of 3000 mL−1, using the same methods as presented here. In contrast, in the present investigation, cultures of this strain were found to be highly toxic to O. marina. The BAHME136 culture in the previous experiments grew at a slightly lower rate (µ = 0.11 day−1) than in the present study (µ = 0.14 day−1), but this slight discrepancy cannot explain the difference in potency.

The reasons for this variability of toxicity are unknown but probably reflect a subtle interplay between both genetic and environmental factors which might directly affect the rate of production and/or excretion of allelochemicals or indirectly affect toxicity by alterations in the physical and chemical properties of the lytic compounds. Such complex interactions make it difficult to interpret the observed variability in extracellular lytic activity. For example, according to the first experiment, E. huxleyi seems to be insensitive (Fig. 5H), whereas in the PAM experiment, lytic effects upon exposure to two A. ostenfeldii strains were obvious after 24 h incubation (Fig. 9D). These differences may be partly due to slightly higher cell concentrations in this second experiment (2000 versus 1500 mL−1), but are more likely caused by large differences in toxicity per cell of the different A. ostenfeldii subcultures in these experiments. Where different target species were subjected simultaneously to the same batch culture of A. ostenfeldii, as in the PAM experiments, some target species seem to differ substantially in their sensitivity towards A. ostenfeldii allelochemicals, but direct comparison is difficult because the target species differed in cell concentration and size and thus in surface area/volume.

Among the heterotrophic targets, athecate dinoflagellates (O. marina, A. crassum) were highly susceptible. Although no thecate heterotrophic dinoflagellate species were tested in these experiments, quantitative comparison experiments have shown that both thecate and athecate heterotrophic dinoflagellate species are immobilized by Alexandrium exotoxins, with the thecate species (Oblea rotunda) even more affected than the athecate O. marina (Tillmann and John, 2002).

Many dinoflagellates have considerable facultative heterotrophic abilities, both for prey ingestion (phagotrophy) and uptake of soluble organic nutrients. It is therefore plausible that immobilizing substances and lysins play a role in food acquisition and competition. For mixotrophic algae, lytic toxins may serve to immobilize and kill potential prey organisms (Tillmann, 1998; Skovgaard and Hansen, 2003; Tillmann, 2003). Although Alexandrium spp. are mainly autotrophic, food vacuoles containing ciliates or phytoplankton cells have been seen in A. ostenfeldii (Jacobson and Anderson, 1996). No food vacuoles were observed in A. ostenfeldii in the present study, but this species probably benefits from enhanced concentrations of dissolved organic matter released as a consequence of its lytic activity. Alexandrium species are known to have the capacity to take up high molecular weight organic molecules (Carlsson et al., 1998; Legrand and Carlsson, 1998), to utilize organic N-substances for growth and toxin production (Ogata et al., 1996) and to remove dissolved free amino acids from the surrounding medium, down to concentrations similar to those found in natural waters (John and Flynn, 1999).

The observed potent short-term effects, such as immobilization and/or lysis of potential predators, also provides evidence of a plausible defence mechanism at the individual cell level. Extracellular release of these lytic allelochemicals may create “toxic” or “repellent” microzones around each cell, although the existence and stability of such microzone gradients is controversial (Mitchell et al., 1985; Alldredge and Cohen, 1987). In any case, experimental evidence convincingly showed that the tintinnid F. ehrenbergii, a ciliate predator upon Alexandrium spp., is killed by exposure to lytic exotoxins of A. tamarense and A. ostenfeldii at high prey cell concentrations (Hansen, 1989; Hansen et al., 1992) . When fed at low Alexandrium cell concentrations (<1 × 103 mL−1) of the same strain, however, Favella readily ingested the prey cells and exhibited rapid growth. This implies that these lytic allelochemicals may only function in chemical defence at a concentration in the aqueous medium above a certain threshold and furthermore that this critical value is at least partially a function of the concentration of the producing species. With respect to predator–prey and microalgal competition interactions, there is a clear need to perform mixed growth experiments starting at low cell concentration to reveal if lytic allelochemicals do indeed act as allelopathic agents at early stages of bloom development.

The important question remains as to whether or not lytic activity of extracellular allelochemicals of Alexandrium is of ecological significance in situ. Ideally, dose–response curves from laboratory studies should be compared with in situ measurements of compound concentrations. However, as long as a chemical determination and quantification of these compounds is not possible, only cell numbers of Alexandrium spp. remain as a basis of comparison between laboratory and field populations. Species of Alexandrium are often considered to be “background” bloom species, in that they are often outnumbered by co-occurring phytoplankton, constituting on a small fraction of total biomass, and rarely form near-monospecific blooms (Anderson, 1998). Nevertheless, members of the A. tamarense/fundyense species complex do occasionally form dense cell aggregations (Cembella et al., 2002), at concentrations exceeding 0.5 × 106 cells L−1. In contrast, populations of A. ostenfeldii have never been recorded at high cell concentration (typical maxima are <5 × 106 cells L−1) and this species tends not to dominate even the thecate dinoflagellate community. Alexandrium species are difficult to discriminate by conventional light microscopy and A. ostenfeldii often co-occurs with other Alexandrium species, in the North Sea (John et al., 2003), Gulf of Maine (Gribble et al., 2005) and on the Nova Scotian coast (Cembella et al., 1999). The fact that A. ostenfeldii is often overlooked in mixed Alexandrium populations from the North Sea (John et al., 2003), Gulf of Maine (Gribble et al., 2005) and on the Nova Scotian coast (Cembella et al., 1999) further underscores the tendency of this species to remain cryptic and at low cell concentrations. With respect to the allelochemical potential of A. ostenfeldii in natural populations, it is important to note that typical field concentrations of 0.1–31 cells mL−1 (Konovalowa, 1993; John et al., 2003) are low relative to the EC50 concentrations estimated for this species in this study (Table II).

Estimates of plankton cell concentrations are typically depth-averaged over a fraction of the water column or are generated from discrete depth sampling. Yet the importance of patchy microzones, cell aggregation and the formation of thin layers of plankton at pycnoclines or nutriclines is increasingly recognized (Rines et al., 2002). Formation of dense patches or thin layers can provide the opportunity for enhanced allelochemical interactions at these steep gradients. However, growing in dense patches or thin layers may also impose locally unfavourable conditions, such as macronutrient limitation or high pH. Patch formation can also enhance the risk of being grazed, as micrograzers are known to actively exploit and remain within patches of food (Buskey and Stoecker, 1988; Fenchel and Blackburn, 1999). Consequently, exudation of noxious chemicals could be a strategy to prevent micrograzers from invading dense algal layers. Indeed, a number of marine algal species [Alexandrium spp., K. mikimotoi (= Gyrodinium aureolum), C. polylepis], which often appear in dense layers associated with the pycnocline (Richardson, 1997), have also been shown to produce extracellular toxins (Gentien, 1998; Schmidt and Hansen, 2001; John et al., 2002; Tillmann and John, 2002). Hence, growth in patches may allow extracellular allelochemicals to reach concentration levels high enough to act efficiently against micrograzers and thus may confer an enhanced defence capability (Vardi et al., 2002).

In conclusion, we have demonstrated the allelochemical potential of uncharacterized exogenous compounds produced by A. ostenfeldii against a variety of putative target predators or competitors. Nevertheless, it does not necessarily follow that these lytic compounds evolved in direct response to predation or competition. It is possible to envisage a scenario in which Alexandrium cell concentrations are high enough in local microzones or thin layers to form an “allelochemical cloud” of high bioactivity. Nevertheless, if evolved simply as a chemical defence, it is difficult to imagine why these compounds are so highly labile and work effectively only at relatively high producing cell concentrations. There is little knowledge on potential long-term effects of sublethal concentrations of Alexandrium allelochemicals on other protists in natural bloom populations. Future work must concentrate not only on the structural elucidation and toxic potency determination of these unknown allelochemicals, but also on unravelling the functional significance and effectiveness of these compounds in natural bloom populations.

ACKNOWLEDGEMENTS

We gratefully note the important contributions provided by researchers at the Institute for Marine Biosciences, National Research Council, Halifax, Canada. In particular, Michael Quilliam and William Hardstaff conducted the analysis of spirolides, and Nancy Lewis assisted with processing of analytical data.

REFERENCES

Alldredge
A. L.
Cohen
J.
Can microscale chemical patches persist in the sea? Microelectrode study of marine snow, fecal pellets
Science
 , 
1987
, vol. 
235
 (pg. 
689
-
691
)
Anderson
D. M.
Anderson
D. M.
Cembella
A. D.
Hallegraeff
G. M.
Physiology and bloom dynamics of toxic Alexandrium species, with emphasis on life cycle transitions
Physiological Ecology of Harmful Algae Blooms
 , 
1998
Berlin/Heidelberg/New York
Springer
(pg. 
29
-
48
)
Arzul
G.
Seguel
M.
Guzman
L.
, et al.  . 
Comparison of allelopathic properties in three toxic Alexandrium species
J. Exp. Mar. Biol. Ecol.
 , 
1999
, vol. 
232
 (pg. 
285
-
295
)
Balech
E.
The Genus Alexandrium Halim (Dinoflagellata)
 , 
1995
Ireland
Sherkin Island Marine Station Publication, Sherkin Island, Co, Cork
Blanco
J.
Campos
M. J.
The effect of water conditioned by a PSP-producing dinoflagellate on the growth of four algal species used as food for invertebrates
Aquaculture
 , 
1988
, vol. 
68
 (pg. 
289
-
298
)
Buskey
E. J.
Stoecker
D. K.
Locomotory patterns of the planktonic ciliate Favella sp.: adaptations for remaining within food patches
Bull. Mar. Sci.
 , 
1988
, vol. 
43
 (pg. 
783
-
796
)
Carlsson
P.
Edling
H.
Béchemin
C.
Interactions between a marine dinoflagellate (Alexandrium catenella) and a bacterial community utilizing riverine humic substances
Aquat. Microb. Ecol.
 , 
1998
, vol. 
16
 (pg. 
65
-
80
)
Cembella
A. D.
Anderson
D. M.
Cembella
A. D.
Hallegraeff
G. M.
Ecophysiology and metabolism of paralytic shellfish toxins in marine microalgae
Physiological Ecology of Harmful Algal Blooms
 , 
1998
Berlin/Heidelberg/New York
Springer Verlag
(pg. 
381
-
403
)
Cembella
A. D.
Chemical ecology of eukaryotic microalgae in marine ecosystems
Phycologia
 , 
2003
, vol. 
42
 (pg. 
420
-
447
)
Cembella
A. D.
Lewis
N. I.
Quilliam
M. A.
Spirolide composition in micro-extracted pooled cells isolated from natural plankton assemblages and from cultures of the dinoflagellate Alexandrium ostenfeldii
Nat. Toxins
 , 
1999
, vol. 
7
 (pg. 
197
-
206
)
Cembella
A. D.
Lewis
N. I.
Quilliam
M. A.
The marine dinoflagellate Alexandrium ostenfeldii (Dinophyceae) as the causative organism of spirolide shellfish toxins
Phycologia
 , 
2000
, vol. 
39
 (pg. 
67
-
74
)
Cembella
A. D.
Quilliam
M. A.
Lewis
N. I.
, et al.  . 
The toxigenic marine dinoflagellate Alexandrium tamarense as the probable cause of mortality of caged salmon in Nova Scotia
Harmful Algae
 , 
2002
, vol. 
1
 (pg. 
313
-
325
)
Dahl
E.
Lindahl
O.
Paasche
E.
, et al.  . 
Cosper
E. M.
Bricelj
V. M.
Carpenter
E. J.
The Chrysochromulina polylepis bloom in Scandinavian waters during spring 1988
Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms
 , 
1989
Berlin
Springer Verlag
(pg. 
383
-
405
)
Doucette
G. J.
Cembella
A. D.
Boyer
G. L.
Cyst formation in the red tide dinoflagellate Alexandrium tamarense (Dinophyceae): effects of iron stress
J. Phycol.
 , 
1989
, vol. 
25
 (pg. 
721
-
731
)
Doucette
G. J.
Kodama
M.
Franca
S.
, et al.  . 
Anderson
D. M.
Cembella
A. D.
Hallegraeff
G. M.
Bacterial interaction with harmful algal bloom species: bloom ecology, toxigenesis, and cytology
Physiological Ecology of Harmful Algal Blooms
 , 
1998
Berlin/Heidelberg/New York
Springer Verlag
(pg. 
619
-
647
)
von Ehlert
E.
Jüttner
F.
Phosphorus limitation and not light controls the extracellular release of allelopathic compounds by Trichormus doliolum (Cyanobacteria)
Limnol. Oceanogr.
 , 
1997
, vol. 
42
 (pg. 
1796
-
1802
)
Fenchel
T.
Blackburn
N.
Motile chemosensory behaviour of phagotrophic protists: mechanisms for and efficiency in congregating at food patches
Protists
 , 
1999
, vol. 
150
 (pg. 
325
-
336
)
Fistarol
G. O.
Legrand
C.
Rengefors
K.
, et al.  . 
Temporary cysts formation in phytoplankton: a response to allelopathic competitors?
Environ. Microbiol.
 , 
2004
, vol. 
6
 (pg. 
791
-
798
)
Fistarol
G. O.
Legrand
C.
Selander
E.
, et al.  . 
Allelopathy in Alexandrium spp.: effect on natural plankton community and on algal monocultures
Aquat. Microb. Ecol.
 , 
2004
, vol. 
35
 (pg. 
45
-
56
)
Gentien
P.
Anderson
D. A.
Cembella
A. D.
Hallegraeff
G. M.
Bloom dynamics and ecophysiology of the Gymnodinium mikimotoi species complex
Physiological Ecology of Harmful Algae Blooms
 , 
1998
Berlin/Heidelberg/New York
Springer
(pg. 
155
-
173
)
Gleason
F. K.
Baxa
C. A.
Activity of the natural algicide, cyanobacterin, on eukaryotic microorganisms
FEMS Microbiol. Lett.
 , 
1986
, vol. 
33
 (pg. 
85
-
88
)
Granéli
E.
Hansen
P. J.
Granéli
E.
Turner
J. T.
Allelopathy in harmful algae: a mechanism to compete for resources?
Ecology of Harmful Algae
 , 
2006
Berlin
Springer
(pg. 
189
-
201
)
Granéli
E.
Johansson
N.
Effects of the toxic haptophyte Prymnesium parvum on the survival and feeding of a ciliate: the influence of different nutrient conditions
Mar. Ecol. Prog. Ser.
 , 
2003
, vol. 
254
 (pg. 
49
-
56
)
Granéli
E.
Johansson
N.
Increase in the production of allelopathic substances by Prymnesium parvum cells grown under N- or P-deficient conditions
Harmful Algae
 , 
2003
, vol. 
2
 (pg. 
135
-
145
)
Gribble
K. E.
Keafer
B. A.
Quilliam
M. A.
, et al.  . 
Distribution and toxicity of Alexandrium ostenfeldii (Dinophyceae) in the Gulf of Maine, USA
Deep-Sea Res. Pt. II
 , 
2005
, vol. 
52
 (pg. 
2745
-
2763
)
Hansen
P. J.
The red tide dinoflagellate Alexandrium tamarense: effects on behaviour and growth of a tintinnid ciliate
Mar. Ecol. Prog. Ser.
 , 
1989
, vol. 
53
 (pg. 
105
-
116
)
Hansen
P. J.
Cembella
A. D.
Moestrup
O.
The marine dinoflagellate Alexandrium ostenfeldii: paralytic shellfish toxin concentration, composition, and toxicity to a tintinnid ciliate
J. Phycol.
 , 
1992
, vol. 
28
 (pg. 
597
-
603
)
Hewlett
P. S.
Placklet
R. L.
Placklet
R. L.
Placklet
R. L.
The probit and similar transformation. In Hewlett
An Introduction to the Interpretation of Quantal Response in Biology
 , 
1979
London
Edward Arnold Ltd
(pg. 
12
-
17
)
Hu
T.
Curtis
J. M.
Oshima
Y.
, et al.  . 
Spirolides B and D, two novel macrocycles isolated from the digestive glands of shellfish
J. Chem. Soc. Chem. Commun., 2159–2161
 , 
1995
Jacobson
D. M.
Anderson
D. M.
Widespread phagocytosis of ciliates and other protists by mixotrophic and heterotrophic thecate dinoflagellates
J. Phycol.
 , 
1996
, vol. 
32
 (pg. 
279
-
285
)
Jensen
M. O.
Moestrup
O.
Autecology of the toxic dinoflagellate Alexandrium ostenfeldii: life history and growth at different temperatures and salinities
Eur. J. Phycol.
 , 
1997
, vol. 
32
 (pg. 
9
-
18
)
John
E. H.
Flynn
K. J.
Amino acid uptake by the toxic dinoflagellate Alexandrium fundyense
Mar. Biol.
 , 
1999
, vol. 
133
 (pg. 
11
-
19
)
John
U.
Tillmann
U.
Medlin
L.
A comparative approach to study inhibition of grazing and lipid composition of a toxic and non-toxic clone of Chrysochromulina polylepis (Prymnesiophyceae)
Harmful Algae
 , 
2002
, vol. 
1
 (pg. 
45
-
57
)
John
U.
Cembella
A. D.
Hummert
C.
, et al.  . 
Discrimination of the toxigenic dinoflagellates Alexandrium tamarense and A. ostenfeldii in co-occurring natural populations from Scottish coastal waters
Eur. J. Phycol.
 , 
2003
, vol. 
38
 (pg. 
25
-
40
)
Juneau
P.
Harrison
P. J.
Comparison by PAM fluorometry of photosynthetic activity of nine marine phytoplankton grown under identical conditions
Photochem. Photobiol.
 , 
2005
, vol. 
81
 (pg. 
649
-
653
)
Keller
M. D.
Selvin
R. C.
Claus
W.
, et al.  . 
Media for the culture of oceanic ultraphytoplankton
J. Phycol.
 , 
1987
, vol. 
23
 (pg. 
633
-
638
)
Konovalowa
G. V.
Smayda
T. J.
Shimizu
Y.
Toxic and potentially toxic dinoflagellates from the far east coastal waters of Russia
Toxic phytoplankton blooms in the sea
 , 
1993
Amsterdam
Elsevier Science Publishers B.V
(pg. 
275
-
279
)
Legrand
C.
Carlsson
P.
Uptake of high molecular weight dextran by the dinoflagellate Alexandrium catenella. Aquat
Microb. Ecol.
 , 
1998
, vol. 
16
 (pg. 
81
-
86
)
Legrand
C.
Rengefors
K.
Fistarol
G. O.
, et al.  . 
Allelopathy in phytoplankton—biochemical, ecological and evolutionary aspects
Phycologia
 , 
2003
, vol. 
42
 (pg. 
406
-
419
)
Mackenzie
L.
White
D.
Oshima
Y.
, et al.  . 
The resting cysts and toxicity of Alexandrium ostenfeldii (Dinophyceae) in New Zealand
Phycologia
 , 
1996
, vol. 
35
 (pg. 
148
-
155
)
MacKinnon
S. L.
Walter
J. A.
Quilliam
M. A.
, et al.  . 
Spirolides isolated from Danish strains of the toxigenic dinoflagellate Alexandrium ostenfeldii
J. Nat. Prod.
 , 
2006
, vol. 
69
 (pg. 
983
-
987
)
Maclean
C.
Cembella
A. D.
Quilliam
M. A.
Effects of light, salinity and inorganic nitrogen on cell growth and spirolide production in the marine dinoflagellate Alexandrium ostenfeldii (Paulsen) Balech et Tangen
Bot. Mar.
 , 
2003
, vol. 
46
 (pg. 
466
-
474
)
Matsuoka
K.
Cho
H. J.
Jacobson
D. M.
Observation of the feeding behaviour and growth rates of the heterotrophic dinoflagellate Polykrikos kofoidii (Polykrikaceae, Dinophyceae)
Phycologia
 , 
2000
, vol. 
39
 (pg. 
82
-
86
)
McClintock
J. B.
Baker
B. J.
Marine Chemical Ecology
 , 
2001
Boca Raton, Florida
CRC Press
Mitchell
J. G.
Okubo
A.
Fuhrmann
J. A.
Microzones surrounding phytoplankton form the basis for a stratified marine microbial ecosystem
Nature
 , 
1985
, vol. 
316
 (pg. 
58
-
59
)
Ogata
T.
Koike
K.
Nomura
S.
, et al.  . 
Yasumoto
T.
Oshima
Y.
Fukuyo
Y.
Utilization of organic substances for growth and toxin production by Alexandrium tamarense
Harmful and Toxic Algal Blooms
 , 
1996
Paris
Intergovernmental Oceanographic Commission of UNESCO
(pg. 
343
-
346
)
Paulsen
O.
Plankton-investigations in the waters round Iceland in 1903
Medd. Kommn. Havunders. Kobenh. Ser. Plankt.
 , 
1904
, vol. 
1
 (pg. 
1
-
40
)
Pedersen
M. F.
Hansen
P. J.
Effects of high pH on the growth and survival of six marine heterotrophic protists
Mar. Ecol. Prog. Ser.
 , 
2003
, vol. 
33
 (pg. 
33
-
41
)
Richard
D.
Arsenault
E.
Cembella
A. D.
, et al.  . 
Hallegraeff
G.M.
Blackburn
S.I.
Bolch
C.J.S.
Lewis
R. J.
Investigations into the toxicology and pharmacology of spirolides, a novel group of shellfish toxins
Harmful Algal Blooms 2000
 , 
2001
Paris
UNESCO
(pg. 
383
-
386
)
Richardson
K.
Harmful or exceptional phytoplankton blooms in the marine ecosystem
Adv. Mar. Biol.
 , 
1997
, vol. 
31
 (pg. 
301
-
385
)
Rines
J. E. B.
Donaghay
P. L.
Dekshenieks
M. M.
, et al.  . 
Thin layers and camouflage: hidden Pseudo-nitzschia spp. (Bacillariophyceae) populations in a fjord in the San Juan Islands, Washington, USA
Mar. Ecol. Prog. Ser.
 , 
2002
, vol. 
225
 (pg. 
123
-
137
)
Schmidt
L. E.
Hansen
P. J.
Allelopathy in the prymnesiophyte Chrysochromulina polylepis: effect of cell concentration, growth phase and pH
Mar. Ecol. Prog. Ser.
 , 
2001
, vol. 
216
 (pg. 
67
-
81
)
Schmitter
R. E.
Taylor
D. L.
Seliger
H. H.
Temporary cysts of Gonyaulax excavata: effects of temperature and light
Toxic Dinoflagellate Blooms
 , 
1979
New York
Elsevier North Holland
pg. 
123
 
Schreiber
U.
Schliewa
U.
Bilger
W.
Continuous recording of photochemical and non-photochemical chlorophyll fluorescence quenching with a new type of modulation fluorometer
Photosynthesis Res.
 , 
1986
, vol. 
10
 (pg. 
51
-
62
)
Skovgaard
A.
Hansen
P. J.
Food uptake in the harmful alga Prymnesium parvum mediated by excreted toxins
Limnol. Oceanogr.
 , 
2003
, vol. 
48
 (pg. 
1161
-
1166
)
Skovgaard
A.
Legrand
C.
Hansen
P. J.
, et al.  . 
Effects of nutrient limitation on food uptake in the toxic Haptophyte Prymnesium parvum
Aquat. Microb. Ecol.
 , 
2003
, vol. 
31
 (pg. 
259
-
265
)
Sukenik
A.
Ashkol
R.
Livne
A.
, et al.  . 
Inhibition of growth and photosynthesis of the dinoflagellate Peridinium gatunense by Microcystis sp. (cyanobacteria): a novel allelopathic mechanism
Limnol. Oceanogr.
 , 
2002
, vol. 
46
 (pg. 
1656
-
1663
)
Taylor
F. J. R.
Fukuyo
Y.
Anderson
D. M.
Cembella
A. D.
Hallegraeff
G. M.
The neurotoxigenic dinoflagellate genus Alexandrium halim: general introduction
Physiological Ecology of Harmful Algal Blooms
 , 
1998
, vol. 
41
 
Berlin
Springer
(pg. 
3
-
12
)
Taylor
J. F. R.
The Biology of Dinoflagellates
 , 
1987
Oxford
Blackwell Scientific Publications
Tillmann
U.
Phagotrophy by a plastidic haptophyte, Prymnesium patelliferum
Aquat. Microb. Ecol.
 , 
1998
, vol. 
14
 (pg. 
155
-
160
)
Tillmann
U.
Kill and eat your predator: a winning strategy of the planktonic flagellate Prymnesium parvum
Aquat. Microb. Ecol.
 , 
2003
, vol. 
32
 (pg. 
73
-
84
)
Tillmann
U.
John
U.
Toxic effects of Alexandrium spp. on heterotrophic dinoflagellates: an allelochemical defence mechanism independent of PSP toxins
Mar. Ecol. Prog. Ser.
 , 
2002
, vol. 
230
 (pg. 
47
-
58
)
Toth
G. B.
Norén
F.
Selander
E.
, et al.  . 
Marine dinoflagellates show induced life-history shifts to escape parasite infection in response to water-borne signals
P. Roy. Soc. Lond. B. Bio.
 , 
2004
, vol. 
271
 (pg. 
733
-
738
)
Vardi
A.
Schatz
D.
Beeri
K.
, et al.  . 
Dinoflagellate-cyanobacterium communication may determine the composition of phytoplankton assemblage in a mesotrophic lake
Curr. Biol.
 , 
2002
, vol. 
12
 (pg. 
1767
-
1772
)
Wolfe
G. V.
The chemical defense ecology of marine unicellular plankton: constrains, mechanisms, and impacts
Biol. Bull.
 , 
2000
, vol. 
198
 (pg. 
225
-
244
)

Author notes

Communicating editor: K.J. Flynn