AQUATIC MICROBIAL ECOLOGYAquat Microb Ecol

Vol. 74: 29–41, 2015doi: 10.3354/ame01724

Published online January 12

INTRODUCTION

Species of Pseudo-nitzschia are common membersof phytoplankton communities throughout the world,and some of them can form persistent as well as near-monospecific blooms (Subba Rao et al. 1988, Bates etal. 1989, Martin et al. 1990, Stonik et al. 2001). Todate, 12 Pseudo-nitzschia species are confirmed pro-ducers of domoic acid (DA) (Trainer et al. 2012),

which can accumulate in the food webs duringblooms and cause amnesic shellfish poisoning (ASP)in an array of animals, including marine mammals,sea birds, and humans (Bates et al. 1989, Kotaki et al.2000, Scholin et al. 2000, Bargu et al. 2002, Trainer etal. 2012).

Blooms of Pseudo-nitzschia occur in both coastaland open ocean environments. In estuarine systems,many Pseudo-nitzschia blooms have been associated

© Inter-Research 2015 · www.int-res.com*These authors contributed equally to this work**Corresponding author: christopher.gobler@stonybrook.edu

Ability of the marine diatoms Pseudo-nitzschiamultiseries and P. pungens to inhibit the growth of co-occurring phytoplankton via allelopathy

Ning Xu1,*, Ying Zhong Tang2,3,*, Junlian Qin1, Shunshan Duan1, Christopher J. Gobler2,**

1Institute of Hydrobiology, Jinan University, Guangzhou 510632, PR China2School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York 11794-5000, USA

3Key Laboratory of Marine Ecology and Environmental Sciences, Institute of Oceanology, Chinese Academy of Science, Qingdao 266071, PR China

ABSTRACT: Diatoms within the genus Pseudo-nitzschia can form near-monospecific blooms inboth natural and iron-fertilized high-nutrient, low-chlorophyll (HNLC) regions and can havedetrimental impacts on marine ecosystems. Here, we demonstrate the ability of P. pungens iso-lated from the South China Sea and 2 strains of P. multiseries isolated from the Bay of Fundy, Canada,to produce extracellular compounds capable of lysing and/or inhibiting the growth of multiplephytoplankton species. Since the allelopathic activity was found in both P. multiseries, which pro-duces domoic acid (DA), and P. pungens, which produces little if any DA, the allelopathic effectsof Pseudo-nitzschia spp. seem to be unrelated to this compound. Allelopathic inhibition of otherphytoplankton was documented during exponential and stationary growth phases of Pseudo-nitzschia, and the strongest allelopathic effects were obtained from sonicated cultures, suggestingthat the sudden release of allelochemicals via processes such as cell lysis or zooplankton grazingmay have the strongest effect in an ecosystem setting. Differences in the responses of target spe-cies to Pseudo-nitzschia spp. suggest these algae may produce multiple compounds that vary intheir allelopathic potency and composition as a function of species, strain, growth stage, and per-haps other factors. Collectively, these results suggest that the allelopathy may affect competitionbetween Pseudo-nitzschia spp. and other phytoplankton and may play an important role in theformation and persistence of natural and iron-fertilized blooms.

KEY WORDS: Allelopathy · Pseudo-nitzschia multiseries · Pseudo-nitzschia pungens · Harmfulalgal bloom (HAB) · Competition · Phytoplankton

Resale or republication not permitted without written consent of the publisher

Aquat Microb Ecol 74: 29–41, 2015

with anthropogenic nutrient loading, whereas innear-shore, coastal regions, blooms have been re -lated to upwelling events (as reviewed by Andersonet al. 2008, Heisler et al. 2008). Open ocean iron-enrichment experiments within high-nutrient, low-chlorophyll (HNLC) zones have frequently beenshown to stimulate the growth of Pseudo-nitzschiaspp. (Coale et al. 1996, 2004, Boyd et al. 2000, Landryet al. 2000, Gervais et al. 2002, Peloquin & Smith2006). For example, Trick et al. (2010) demonstratedthat Pseudo-nitzschia spp. within HNLC regions canproduce high level of DA in response to iron fertiliza-tion. In both natural and fertilized HNLC blooms,Pseudo-nitzschia can become near-monospecificamong larger phytoplankton, sometimes accountingfor >90% of the micro-phytoplankton (Boyd et al.2007, Trick et al. 2010). One factor that could pro-mote such competitive dominance over other phyto-plankton could be the release of allelochemicals.

Since large amounts of extracellular DA can beproduced by Pseudo-nitzschia (Bates et al. 1991,Mal donado et al. 2002) and since Pseudo-nitzschiacan dominate phytoplankton communities as DAaccumulates (Trick et al. 2010), it is plausible that thiscompound acts as an allelochemical to inhibit thegrowth of other algae. Prior studies investigating theeffects of DA on other phytoplankton, however, havedemonstrated that even high levels of this compounddid not appreciably alter the growth of a wide arrayof phytoplankton, including diatoms, prymnesio-phytes, euglenophytes, dinophytes, raphidophytes,prasinophytes, and cryptophytes (Windust 1992,Lund holm et al. 2005). Lundholm et al. (2005) de -monstrated that the growth of Chrysochromulinaericina was reduced when co-cultured with P. multi-series but concluded this effect was due to elevatedpH rather than DA or allelochemicals.

Harmful algae can display a wide range of toxicityand/or noxious effects among strains and field popu-lations (Burkholder & Glibert 2009). For example,studies of the toxigenic dinoflagellate Alexandriumtamarense have documented a large intra-populationclonal variability in allelopathic potency (Tillmann etal. 2009, Hattenrath-Lehmann & Gobler 2011). Thereare >37 Pseudo-nitzschia species (Trai ner et al.2012), and even monospecific diatom blooms can becomposed of a great diversity of strains with differentphysiological characteristics (Rynearson & Armbrust2000). Differences exist in the production of DA byPseudo-nitzschia species and strains (Trainer et al.2012), although other aspects of physiological diver-sity among Pseudo-nitzschia spp. have not beencomprehensively assessed. Furthermore, many spe-

cies of diatoms have been shown to have allelopathicproperties due to the production of compoundsbesides DA, such as oxylipins and aldehydes (Ianora& Miralto 2010). Given the physiological diversityamong Pseudo-nitzschia spp. regarding productionof secondary metabolites such as DA, the ability ofdiatoms to be allelopathic, and the ability of Pseudo-nitzschia populations to bloom to the exclusion ofother phytoplankton, the extent to which Pseudo-nitzschia may be allelopathic remains an open question.

Here, we present a study investigating the poten-tial allelopathic effects of 2 species of Pseudo-nitzschia (P. pungens and P. multiseries) on 5 phyto-plankton species. We explored how cell density,growth stage, and target species influenced allelo-pathic effects. We further investigated allelopathicmechanisms of action using filtrate and sonicatedextracts of Pseudo-nitzschia cultures.

MATERIALS AND METHODS

Cultures and culturing conditions

The effects of P. pungens (PP2) and P. multiseries(2 strains: CLNN16 and CLNN21) on other phyto-plankton were investigated via co-culturing. Clonalcultures of Pseudo-nitzschia were obtained by pipet-ting a single cell under an inverted microscope frombloom water from the South China Sea (P. pungens,strain No. PP2) and Bay of Fundy, Canada (P. multi-series, strain No. CLNN16 and CLNN21). Informa-tion regarding the genetic confirmation of these 2Pseudo-nitzschia species has been previously in -cluded in Tang et al. (2010). Quantification of DA incultures has demonstrated that PP2 did not produceappreciable levels of DA (below detection limit),while CLNN16 and CLNN21 were both confirmedDA producers. Cultures were grown in sterile GSemedium with a salinity of 32.5 PSU, made with auto-claved and 0.2 µm filtered seawater (Doblin et al.1999). Cultures were grown at 21°C in an incubatorwith a 12 h light:12 h dark cycle, illuminated by abank of fluorescent lights providing a light intensityof ~100 µmol quanta m−2 s−1.

Five target phytoplankton species were used in thisstudy, including 2 species of dinoflagellates (Aka -shiwo sanguinea AS2 and Prorocentrum minimumCCMP696), a raphidophyte (Chattonella marinaChatM1), a haptophyte (Phaeocystis globosa), and acryptophyte (Rhodomonas salina CCMP1319). TheCCMP cultures were obtained from the Provasoli-

30

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp.

Guillard National Center for Culture of MarinePhytoplankton (Maine, USA), while C. marina Chat -M1 isolated from Singapore coastal waters waskindly provided by M. J. Holmes at the National Uni-versity of Singapore. A. sanguinea AS2 was isolatedby Y. Tang from Chesapeake Bay (Virginia, USA),and P. globosa was isolated from the South China Seaby N Xu. All the cultures were maintained under thesame conditions as used for Pseudo-nitzschia.

All cultures used for experiments were in early ormid-exponential growth phase and had high levels ofnitrate, phosphate, and silicate present. In combina-tion with additions of GSe medium to mono- (control)and co-cultures in all experiments, ambient nitrateand phosphate concentrations (measured spectro -photometrically; Parsons et al. 1984) always re mainedabove 100 and 20 µM, respectively, and thus re -mained above the half-saturation constants fornitrate and phosphate for nearly all phytoplankton(Smayda 1997). All experiments presented here wereshort (≤3 d) and contained relatively low levels ofalgal biomass, 2 precautions that kept pH levels lowduring experiments (between 7.4 and 8.3). Duringeach experiment, the differences in pH values be -tween controls and treatments were always

Aquat Microb Ecol 74: 29–41, 2015

licate 10 ml test tubes containing 90 or 75% 0.22 µmfiltrate of P. pungens or P. multiseries cultures en -riched with stock solutions of nutrient to levels of thefull strength GSe medium. The monospecific culturesof A. sanguinea AS2 and R. salina CCMP1319 dilutedwith GSe medium were used as controls (identicalcell density as in treatment). Test tubes were incu-bated for 72 h under conditions as described above,after which cultures were preserved with Lugol’ssolution (final concentration: 2%), and cell densitieswere enumerated.

Allelopathic effects of sonicated extracts of Pseudo-nitzschia cultures

To better understand the nature of the allelopathiceffect of Pseudo-nitzschia spp. on other microalgae,experiments were conducted in which different components and concentration gradients of Pseudo-nitzschia spp. cultures were manipulated. Specifi-cally, cultures of P. pungens PP2 (cell density: 287 700cells ml−1 in exponential phase and 225 330 cells ml−1

in stationary phase) and P. multiseries CLNN21 (celldensity: 208 100 cells ml−1 in exponential phase and213 300 cells ml−1 in stationary phase) were lysed viasonication with a high power sonicator (UltrasonicPower). The lysis of cells was confirmed microscopi-cally. Half of the sonicated culture was then filteredthrough a 0.22 µm polycarbonate membrane to createa cell-free treatment, while the other half was usedunamended. The cultures of A. sanguinea AS2 (finalcell density: 100 cells ml−1) or R. salina CCMP1319 (fi-nal cell density: 500 cells ml−1) were inoculated intotriplicate test tubes containing 10 ml whole or filteredsonicated cultures of P. pungens or P. multiseries en-riched with nutrients of GSe medium. The percent-ages of sonicated cultures used during experimentswere 90, 75, 50, and 25% for filtered treatments and90 and 75% for non-filtered treatments. The culturesof A. sanguinea AS2 and R. salina CCMP1319 dilutedwith GSe medium to the same final AS2 or CCMP -1319 cell density as above were used as controls. Alltest tubes were incubated for 72 h and then preservedwith Lugol’s solution (final concentration: 2%) forenumeration of cell densities.

Statistics

Statistical analyses were performed using SPSS17.0. Significant differences in final cell densities ofthe target species among treatments and controls

were assessed with 1-way ANOVAs. In experimentswith multiple dilutions of Pseudo-nitzschia, finalcell densities of the target species followed a sig-moidal declining pattern when plotted against log- transformed Pseudo-nitzschia cell concentrations. Assuch, estimates of EC50, i.e. the Pseudo-nitzschia cellconcentration yielding a 50% decline in the targetspecies, were determined by fitting the data points tothe following equation using the non-linear fit:

Nfinal = Ncontrol/(1 + (x/EC50)h)

where Nfinal was the final cell concentration of targetspecies in treatments, Ncontrol was the final cell con-centration of target species in controls, x was the log-transformed cell concentration of Pseudo-nitzschiaspp., and h was the fit constant. Results are presentedas EC50 values (cells ml−1) with 95% confidence intervals.

RESULTS

Effects of Pseudo-nitzschia pungens on multiple phytoplankton species

Pseudo-nitzschia pungens significantly reducedthe cell densities of 3 of the 5 target species duringco-culturing experiments, compared to their respectivecontrols: the dinoflagellate Akashiwo sanguinea, thecryptophyte Rhodomonas salina, and the raphido-phyte Chattonella marina (p < 0.05 for each; Fig. 1a).While A. sanguinea cell densities were re duced by>60% during the incubation, reductions in the densi-ties of R. salina and C. marina were smaller (~20%and ~10%, respectively). The growth rates of the 3target algae also decreased compared to their re -spective controls, among which negative growth wasobserved in A. sanguinea and R. salina (Fig. 1b). Frag-mental algal cells of A. sanguinea and R. salina wereobserved when co-cultured with P. pungens, indica-ting that target algae were lysed. While the growth-inhibiting effect of P. pungens on the 3 target specieswas not density-dependent (p > 0.05, post hoc pair-wise comparison of ANOVA; Fig. 1a), P. pungensgrew rapidly when co-cultured with A. sanguinea, R.salina and C. marina, regardless of initial cell densi-ties (μ > 1.5 d−1), and thus, final cell densities weresimilar among treatments. In contrast to these sensi-tive phytoplankton species, the armored dinoflagel-late Prorocentrum minimum and the haptophytePhaeocystis globosa were unaffected or even pro-moted (P. globosa) by co-culturing with P. pungens.Initial pH values of all cultures were ~7.5. At the end

32

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp.

of the experiment, pH levels in R. salina and P. glo-bosa treatments with the highest cell densitiesincreased slightly (8.1 and 7.7 respectively), whereasthe pH levels of the other treatments were ~7.5.

Effects of different Pseudo-nitzschia strains

Co-culture experiments using 2 strains of P. multi-series (CLNN16 and CLNN21) and the target speciesA. sanguinea revealed that strain CLNN16 could in -hibit the growth of A. sanguinea (p < 0.05; Fig. 2a,b),whereas the inhibition effects of CLNN21 were onlyobserved in treatments with the highest cell densities

(p > 0.05; Fig. 2a,b). The strain CLNN16 (30 to 40%reduction in A. sanguinea cell densities; p < 0.05; Fig.2a) appeared significantly more potent than CLNN21(0 to 10% reduction; p > 0.05; Fig. 2a) with approxi-mately the same cell volumes. For both strains of P.multiseries, the inhibitory effects at their high celldensities were stronger than at the low and mediumcell densities (Fig. 2a,b). Obvious lytic effects werealso microscopically observed in the target alga A.sanguinea. These results indicate that the allelo-pathic effects of Pseudo-nitzschia spp. were not spe-cies-specific, although different algal species orstrains varied in the strength of their allelopathiceffects. Initial pH levels were ~7.5 in all treatments.At the end of experiments, pH in the CLNN21 treat-

33

Fig. 1. Effects of Pseudo-nitzschia pungens on the growth of5 phytoplankton species. The initial cell densities of P. pun-gens were 220 (low [L]), 2200 (moderate [M]), and 13 400(high [H]) cells ml−1. Initial cell densities in treatments andcontrols for Akashiwo sanguinea AS2, Prorocentrum mini-mum CCMP696, Chattonella marina ChatM1, Phaeocystisglobosa, and Rhodomonas salina CCMP1319 were 100, 600,1400, 285 000, and 70 000 cells ml−1, respectively. Results wereexpressed as triplicate mean ± 1 SD. Significant re duction in(a) cell density or (b) growth rate relative to the control is

denoted with *p < 0.05 or **p < 0.01

Fig. 2. Effects of 2 strains of P. multiseries — CLNN16 andCLNN21 — on the growth of A. sanguinea (AS2). The ini-tial cell densities of CLNN16 and CLNN21 were 220 (low[L]), 2200 (moderate [M]), and 22 000 (high [H]) cells ml−1.The initial cell densities of AS2 in treatments and controlswere ~100 cells ml−1. Results were expressed as triplicatemean ± 1 SD. Significant reduction in (a) cell density or(b) growth rate relative to the control is denoted with

*p < 0.05 or **p < 0.01

Aquat Microb Ecol 74: 29–41, 2015

ment with the highest cell density was 8.3, which washigher than that in the CLNN16 treatment (7.7).

Effects of filtrate of Pseudo-nitzschia spp. cultures

Filtrate (i.e. cell-free culture medium) of bothP. pungens PP2 and P. multiseries CLNN21 cultureshad significant allelopathic effects on algal targetA. sanguinea (p < 0.01; Fig. 3a,c), while only PP2showed growth-inhibiting effects on R. salina (p <0.05; Fig. 3b). Consistent with whole cell assays,A. sanguinea (up to 50% reduction in cell densities)was more sensitive than R. salina (0 to 10% reduc-tion; Fig. 3). Filtrate from stationary phase culturesexhibited more potent effects than filtrate from expo-

nential phase cultures. For example, while exponen-tial phase P. pungens filtrate reduced A. sanguineadensities by 10 to 20%, stationary phase filtrateyielded reductions of ~50% (p < 0.001; Fig. 3a). ForP. multiseries, filtrate of the exponential phase hadno effect on A. sanguinea, whereas filtrate of the sta-tionary phase caused 30 to 50% reduction in celldensity, compared to the control (p < 0.05; Fig. 3c).Likewise, filtrate of P. pungens within the exponen-tial phase did not alter R. salina densities, whereascultures in the stationary phase caused significantreductions (p < 0.05; Fig. 3b). These filtrate experi-ments demonstrated that growth-inhibiting effects ofP. pungens and P. multiseries were not dependent oncell contact or actively growing populations andwere most potent during the stationary phase. Initial

34

Fig. 3. Effects of filtrate (0.22 µm) of (a,b) P. pungens and (c,d) P. multiseries from the exponential phase and stationary phaseon the growth of (a,c) A. sanguinea and (b,d) R. salina. F-90% and F-75% indicate that the percentages of the filtrates were90 and 75% by volume. The initial cell densities of P. pungens PP2 were 132 750 cells ml−1 in the exponential phase and174 500 cells ml−1 in the stationary phase, while those of P. multiseries CLNN21 were 144 200 cells ml−1 in the exponentialphase and 335 500 cells ml−1 in the stationary phase. The initial cell densities for A. sanguinea AS2 and R. salina CCMP1319were 100 and 500 cells ml−1, res pectively. Results were expressed as triplicate mean ±1 SD. Significant reduction in cell

density from control is denoted with *p < 0.05 or **p < 0.01

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp.

pH levels were between 7.4 and 7.7 in all treatments.During the experiment, the pH levels were consis-tently below 8.0, with final pH levels between 7.7and 7.9.

Effects of sonicated extracts

Administration of sonicated extracts of Pseudo-nitzschia spp. cultures caused highly significant inhi-bition of the growth of both A. sanguinea andR. salina (Fig. 4). The whole sonicated extracts and

the filtrate of sonicated extracts exhibited dose-dependent effects, and in most cases exponentialphase cultures were more potent than stationaryphase cultures (p < 0.05, post hoc pairwise compari-son of ANOVA) (Fig. 4). One exception to this trendwas observed for the filtered and sonicated culturesof P. pungens in stationary phase, which were moreeffective against A. sanguinea than those in expo-nential phase. The filtrate of sonicated extracts wasgenerally more potent than the unfiltered sonicatedcultures (Fig. 4). The most dramatic results wereobtained from the administration of sonicated extract

35

Fig. 4 Effects of sonicated extracts of (a,b) P. pungens and (c,d) P. multiseries from the exponential phase and stationary phaseon the growth of (a,c) A. sanguinea and (b,d) R. salina. The letters S and SF indicate the whole sonicated extracts and the son-icated and filtered (0.22 µm) extracts, respectively. Percentages 90, 75, 50, and 25% indicate the dilution gradients by volume.The initial cell densities for P. pungens PP2 were 287 700 cells ml−1 in the exponential phase and 225 330 cells ml−1 in the sta-tionary phase, and for P. multiseries CLNN21 were 208 100 cells ml−1 in the exponential phase and 213 300 cells ml−1 in the sta-tionary phase. The initial cell densities for A. sanguinea AS2 and R. salina CCMP1319 were 100 and 500 cells ml−1. Results

were expressed as triplicate mean ± 1 SD

Aquat Microb Ecol 74: 29–41, 2015

of the exponential phase of P. multiseries to A. san-guinea, which yielded a complete elimination whenadministered at or above 75% of the culture volume(p < 0.001; Fig. 4c). Even at lower dosages (25 to50%), 60 to 90% of A. sanguinea cells perished (p <0.001; Fig. 4c). For R. salina, 65−75% and 15−65%reductions in cell density were observed when ex -posed to the filtrate of sonicated P. multiseries cul-tures in exponential and stationary phases, respec-tively (p < 0.05; Fig. 4d). The whole (unfiltered)sonicated extract of P. multiseries in exponentialphase reduced R. salina cell densities by 30 to 40%(p < 0.001), while the whole sonicated extract from stationary phase cultures had no significant effect(Fig. 4d). Sonicated extracts of P. pungens culturesalso significantly reduced A. sanguinea cell densi-ties, with 90% filtrate of the sonicated culture in thestationary phase having the strongest effect (95%reduction; p < 0.001) and 25% filtrate of the soni-cated culture in the exponential phase having theweakest effect (15% reduction; p < 0.001; Fig. 4a).Sonicated extracts of P. pungens cultures in the sta-tionary phase had no significant effect on R. salina,while the sonicated extract in the exponential phasereduced R. salina cell densities by 30 to 75% (p <0.001; Fig. 4b). Initial pH levels were between 7.4and 7.7 in all treatments. At the end of experiments,pH levels ranged from 7.7 to 7.9.

Using the results from the experiments with sonicated extracts, EC50 values of Pseudo-nitzschiawere calculated and were significantly different foreach donor/target combination, ranging from 5658(P. multiseries/A. sanguinea) up to 286 754 cells ml−1

(P. pungens/R. salina) (Fig. 5). EC50 values for A. san-guinea were much lower than for R. salina. Based onthe lowest EC50 values for combinations of Pseudo-nitzschia spp. and target algae, both Pseudo-nitzschiaspecies were similarly effective in inhi biting thegrowth of R. salina, but P. multiseries exhibited aninhibiting effect against A. sanguinea about 10-foldstronger than P. pungens (Fig. 5). Taking into accounta smaller cell size of P. multiseries (width: 3 to 4 µmvs. 4 to 5 µm, length: 30 to 40 µm), the growth-inhibit-ing effect of P. multiseries was generally strongerthan that of P. pungens. Cultures of P. multiserieswithin the exponential growth phase more effec-tively lysed both target species than stationary phaseones, whereas A. sanguinea was less affected and R.salina was more affected by exponential phase cul-tures of P. pungens compared to those in the station-ary phase (Fig. 5).

DISCUSSION

Characteristics of the allelopathy of Pseudo-nitzschia spp.

This study demonstrated the ability of P. pungens(isolated from South China Sea) and P. multiseries(2 strains: CLNN16 and CLNN21, isolated from theBay of Fundy, Canada) to produce extracellular com-pounds capable of lysing and/or inhibiting the growthof 3 target phytoplankton species. While the com-pounds involved have yet to be identified, they maybe generally classified as allelochemicals, secondarymetabolites that act directly on target species (com-petitors or predators).

It has been postulated that physiological factorslinked to nutritional status or growth stage can con-tribute to the variation in the production of allelo-chemicals (Tillmann et al. 2008). Elevated pH haspreviously been considered to be a source of ‘toxicity’of the prymnesiophyte Chrysochromulina polylepison target phytoplankton because this species canraise pH values above 9.0 when it grows to elevateddensities (Schmidt & Hansen 2001). In this study, ashort experimental period (3 d) minimized the accu-mulation of algal biomass and resulted in only minorfluctuation of pH level. Further, the pH measure-ments of mono- and co-cultures remained between7.4 and 8.3, well below levels shown to cause delete-rious effects on marine protists (Schmidt & Hansen2001, Pedersen & Hansen 2003) and within the rangenormally found in target algal cultures. Moreover,more potent allelopathic effects were observed when

36

P. multiseries (e)P. multiseries (s) P. pungens (e)

P. pungens (s)

00 2 4 6 8 10 12 14

5

10

15

20

25

30

35

R. s

alin

a E

C50

(x10

4 ce

lls m

l–1)

A. sanguinea EC50 (×104 cells ml–1)

Fig. 5. A comparison of EC50 values obtained from sonicatedextract experiments with A. sanguinea (x-axis) and R. salina(y-axis). Each data point represents a mean (n = 3) ± 1 SDfor 1 Pseudo-nitzschia strain in exponential phase (e) or

stationary phase (s)

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp.

sonicated filtrates were employed compared to co-cultures that had more algal biomass and higher pH.Thus, elevated pH in Pseudo-nitzschia cultures can-not account for the growth-inhibiting effects ob -served in this study. Nutrients were also unlikely tohave influenced results because additions of nutri-ents to mono- (control) and co-cultures (treatments)in all experiments led to ambient nitrate and phos-phate concentrations always remaining above 100and 20 µM, respectively, and thus above the half- saturation constants for these nutrients for mostphytoplankton (Smayda 1997).

Since extracellular lytic activity was documented inboth P. multiseries, which produces DA, and P pun-gens, which produces little if any DA, the observedallelopathic effects of Pseudo-nitzschia were unlikelyrelated to this compound (Windust 1992, Lundholmet al. 2005). A similar phenomenon has also beenreported in the toxic dinoflagellates Karenia brevisand Alexandrium spp. that exhibit strong allelo-pathic effects unrelated to brevetoxins and saxitox-ins, respectively (Tillmann et al. 2009, Prince et al.2010, Hattenrath-Lehmann & Gobler 2011). As de -scribed for Alexandrium spp. which show inter-clonal variability in lytic potency (Prince et al. 2008,Tillmann et al. 2009, Hattenrath-Lehmann & Gobler2011), Pseudo-nitzschia species and strains variedwith regard to their growth inhibitory effects onother phytoplankton. For example, the allelopathicpo ten cy of P. multiseries CLNN16 was stronger thanthat of P. multiseries CLNN21. Furthermore, 2 otherstrains of P. multiseries (CL-195 also from the Bay ofFundy and OKPm013-2 from Japan) displayed onlymild effects on other phytoplankton that were attrib-uted to high pH (Lundholm et al. 2005). Consistentwith these observations, significant geographicalpop ulation differentiation has been documentedwithin 1 P. pungens variant (P. pungens var. pun-gens) be cause of restricted gene flow among distinctgeographical populations (Casteleyn et al. 2010).Collectively, these findings suggest that blooms ofPseudo-nitzschia are likely to vary in their allelo-pathic capabilities and that allelopathic potency islikely to vary as a function of species, strain, andgrowth stage.

Our filtrate experiments demonstrated that growth-inhibiting effects of P. pungens and P. multiserieswere not dependent on cell contact or actively grow-ing populations but rather that allelochemical(s) canbe released to the ambient environments to act ontarget algae. Stationary phase filtrates of P. pungensshowed stronger growth-inhibiting effects on 2 tar-get algae than the exponential phase filtrates did.

Similarly, the stationary-phase P. multiseries culturecaused >50% decrease in cell density of Akashiwosanguinea, but no significant inhibition effects wereobserved in its exponential-phase culture. Allelo-chemical(s) produced by Pseudo-nitzschia spp. maygradually accumulate in culture medium and attainhigher concentrations in stationary-phase cultureswith longer growth period and higher cell density. Asimilar laboratory study with Isochrysis galbana foundthat the growth-inhibiting effects increased continu-ously from exponential to decline phase (Sun et al.2012), which was consistent with our findings. How-ever, another study reported that the allelopathicactivity of the expo nential-phase Nodularia spumi-gena on target algae Thalassiosira weissflogii andRhodomonas sp. was stronger than that of the sta-tionary-phase culture (Suik kanen et al. 2004). Theobvious variation in the strength of allelopathy withgrowth stage in different algal species suggests thatthe attributes of microalgal allelopathy are species-specific. Although higher cell densities were used infiltrate experiments, exponential-phase filtrates ofP. pungens did not exhibit the growth-inhibitingeffects on Rhodomonas salina (Fig. 3) that were ob -served in co-culture experiments (Fig. 1a,b). Onepossible explanation is related to the stability ofpotential allelochemicals. A recent study of the lyticactivity of Prymnesium parvum has demonstratedthat the extracellular toxins in the supernatant arehighly unstable, and the lytic activity of the intracell-ular toxins, when released by sonication, is not ashigh as that of the extracellular toxins (Blossom et al.2014). Similarly, the extracellular allelochemicals ofPseudo-nitzschia spp. were likely somewhat unsta-ble and thus degraded in the filtrate experiment butpersisted when whole cells were present. Anotherfactor that may partly explain the weaker effects ofthe filtrate compared to whole cells is that some alle-lochemicals may be retained on the filter used (poresize: 0.22 µm). Tillmann et al. (2008) also found thatallelopathic effects of culture filtrate (

Aquat Microb Ecol 74: 29–41, 2015

tion of culture was more significantly inhibitory toA. sanguinea and R. salina than whole cells and culture filtrate. The amplified effects of sonicatedPseudo-nitzschia spp. may be due to the immediaterelease of an intracellular pool of allelochemicals tothe culture medium. While cell densities in some ofthese experiments were higher than in the co-cultureexperiments, the differences were smaller comparedto the difference in the growth-inhibiting effects be -tween the experiments, suggesting this differencewould not account for the stronger effect. In addition,the growth-inhibiting effects in all treatments withsonicated extracts and filtrate of sonicated extractswere dose-dependent, an observation consistent withother allelopathic studies (Tillmann 2003, Gra néli &Salomon 2010). Consistent with other ex periments,the allelopathic strength of sonicated materials differed among Pseudo-nitzschia species, suggestingthat the composition and concentration of allelo-pathic compounds might differ among Pseudo-nitzschia species and strains. The different responsesof the target species (A. sanguinea and R. salina) toeach species of Pseudo-nitzschia at different growthstages further suggested that target algae differ intheir sensitivity to the same allelochemicals. Alterna-tively, allelochemicals may be chemically diverse — acocktail of compounds rather than a single analogue(Tillmann et al. 2008), targeting multiple cellularsites. A recent report demonstrated that K. brevisproduces both unstable, polar, organic allelopathicmolecules as well as a suite of less polar and morestable compounds that are moderately allelopathic totarget species (Prince et al. 2010). Differences inPseudo-nitzschia allelopathic potency may be simi-larly related to the composition of the allelochemicalsproduced by different Pseudo-nitzschia species orstrains at different growth stages and originated fromdifferent sub-cellular locations as well as to their differing modes of action on target species.

In sonication experiments, stationary-phase fil-trates of sonicated P. pungens cultures displayedstronger growth-inhibiting effects than exponential-phase treatments, whereas opposite results wereobserved with whole sonicated extracts. In addition,stronger growth-inhibiting effects were also found inthe exponential phase of P. multiseries on A. san-guinea and of both Pseudo-nitzschia species on R.salina. Clearly, sonication caused the release of intra-cellular allelochemicals that had a stronger effect onthe target algae than filtrate of the cultures, suggest-ing that the majority of allelochemicals are storedintra-cellularly. Since different experimental meth-ods and complex environmental factors may influ-

ence the allelopathic activity of phytoplankton, fur-ther research is needed to quantify and understandthe composition of allelochemicals produced byPseudo-nitzschia spp.

Although a direct comparison of target algae sensi-tivity to allelochemicals is difficult because of differ-ences in cell concentrations, cell size, and thus in surface area/volume ratios, the present study indi-cated that the unarmored dinoflagellate A. san-guinea was highly sensitive to P. pungens while thearmored dino flagellate Prorocentrum minimum wasnot. Among different classes of phytoplankton, dia -toms are considered more resistant to allelochemicalsubstances than ciliates and flagellates (Tillmann etal. 2008). Our observation of these 2 dinoflagellatessuggests that armored species may be more resistantto the allelochemical substances produced by P. pun-gens than unarmoured species because their cellu-lose plates may provide partial protection from allelochemicals. Clearly, a more comprehensiveinvestigation of armored and unarmored dinoflagel-lates will be required to definitively address thisissue. Regardless, the differing sensitivities of targetspecies to Pseudo-nitzschia allelochemicals indicatethe potential for these compounds to both promoteblooms of Pseudo-nitzschia and shape phytoplanktoncommunity composition (Fistarol et al. 2003, Prince etal. 2008, Tang & Gobler 2010, Hattenrath-Lehmann& Gobler 2011).

Relative cell concentrations, physiological status oftarget species, and the plankton community compo-sition may be important factors that modulate allelo-pathic effects (Poulson et al. 2010). In the presentstudy, Phaeocystis globosa was not significantly in -hibited by P. pungens, which might be due to the initial cell densities of P. globosa being much higherthan P. pungens since a higher algal cell density pro-vided more surface area to adsorb and thus dilutedallelochemical(s) (Tang & Gobler 2010). The possibil-ity that P. globosa may be resistant to the allelochem-icals, however, cannot be excluded. Further study ofthese allelopathic effects under a range of cell con-centrations may be useful to more fully evaluate therole of allelopathy in interspecies competition be -tween Pseudo-nitzschia and other phytoplankton.

Many species of diatoms have been shown to haveallelopathic properties due to the production of com-pounds such as fatty acids and aldehydes (Yamasakiet al. 2007, Ianora & Miralto 2010). Furthermore,Ianora & Miralto (2010) suggested a possible con -nection between the production of species-specificoxylipins by P. delicatissima and low hatching suc-cess and apoptosis in the offspring of the copepod

38

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp.

Calanus helgolandicus. Whether oxylipins are com-mon among Pseudo-nitzschia species and whetherthese compounds are responsible for allelopathiceffects on co-occurring phytoplankton are unknown.

Ecological implications of the allelopathic effects

In the present study, reduced growth rates and celldensities were documented in some target algae co-cultured with Pseudo-nitzschia species. Anotherstudy dealing with allelopathy of C. polylepis indi-cated that the harmful effect was observed as an ini-tial decrease in growth rate of the tested algae, fol-lowed by a decline in their population numbers(Schmidt & Hansen 2001). While the duration of ourexperiments did not permit a fine scale temporalexamination of growth rates, we did observe a signif-icant reduction in both the growth rate and biomassaccumulation in target algae sensitive to Pseudo-nitzschia allelochemicals. Thus, the allelopathic ef -fects should be a powerful chemical weapon in com-petition with co-occurring phytoplankton species.Allelopathy has also been hypothesized to play a rolein species succession (Keating 1977), the formation ofharmful algal blooms (Smayda 1997), and the estab-lishment of invasive species (Figueredo et al. 2007).Allelopathic effects are thought to be most relevantat high cell densities typical of algal blooms within anecosystem setting (Jonsson et al. 2009). The potentialeffects of allelochemicals on early bloom develop-ment, when cell concentrations are lower, remainless understood. Species of Pseudo-nitzschia formdense blooms with cell densities ranging from 106 to108 cells l−1 (Trainer et al. 2012). Our co-culture ex -periments indicated that even at lower cell densities(105 cells l−1), P. pungens still inhibited the growth ofmultiple phytoplankton species. Experiments con-ducted with filtrates of cultures demonstrated thatPseudo-nitzschia species exuded allelochemicalsthat remained active over the course of short-termexperiments. In experiments conducted with the son-icated cultures of Pseudo-nitzschia within exponen-tial and stationary growth phases, stronger allelo-pathic effects were observed in the exponentialgrowth phase. Considering all of these re sults, sev-eral characteristics of the source and fate of thePseudo-nitzschia allelochemicals are apparent: Theyare synthesized during active growth, stored intra-cellularly, and are slowly released during exponen-tial growth but released more rapidly during the sta-tionary phase, as cells of poor physiological state leakinternal contents. Sonicated cultures exhibited the

most potent allelopathy, perhaps mimicking fieldpopulations that are lysed or grazed and immediatelyleak potent allelochemicals. Regardless, these resultssupport the hypothesis that allelopathy may be a keystrategy in interspecies competition between Pseudo-nitzschia spp. and other phytoplankton, which in turnmay influence the formation and persistence of blooms.Indeed, Pseudo-nitzschia spp. is a highly adaptablealgal group that can bloom regularly in coastal andoff-shore waters (Trainer et al. 2012). Allelopathymay be a mechanism by which this species is able todominate phytoplankton communities in both natu-ral and iron-fertilized HNLC blooms (Boyd et al.2007, Trick et al. 2010).

Presently, most allelopathic compounds remainunidentified. Possible modes of action include oxida-tive damage, loss of competitor motility, inhibition ofphotosynthesis, inhibition of enzymes, and mem-brane damage (reviewed by Legrand et al. 2003). Forexample, Prince et al. (2008) reported that K. brevismay form nearly monospecific blooms by loweringthe photosynthetic efficiency of competitor speciesand increasing competitor membrane permeability,eventually resulting in competitor growth suppres-sion or death. Although the specific mode of action ofthe allelochemical compounds in Pseudo-nitzschiaagainst target species is presently unknown, therapid lytic action in some treatments (e.g. sonicatedextracts) suggests the allelochemicals may have tar-geted the structure and function of the cell mem-brane and/or the cytoskeleton of the target cell.Alternatively, the allelochemicals may initiate pro-grammed cell death in target species. Our resultsrevealed that the allelopathic effects of sonicated cul-tures were stronger than the filtrate, which suggeststhat specific biological factors such as presence ofcompetitors, physiological status, cell concentrations,and zooplankton grazing may regulate the exudationof the allelochemicals in an ecosystem setting. Fur therinvestigation into mechanisms of allelo chemical exu-dation in Pseudo-nitzschia spp. would certainly ad -vance the understanding of Pseudo-nitzschia bloomecology.

CONCLUSIONS

Two species of Pseudo-nitzschia (P. pungens andP. multiseries) were found to produce extracellularcompounds capable of lysing and/or inhibiting thegrowth of multiple co-occurring phytoplankton spe-cies at low cell densities (105 cells l−1). Allelochemi-cals of Pseudo-nitzschia seemed to be mainly stored

39

Aquat Microb Ecol 74: 29–41, 2015

intracellularly and slowly released during the expo-nential phase and more rapidly in the stationaryphase or during rapid cell disruption (i.e. sonication)that may mimic zooplankton grazing or other meansof cell damage within an ecosystem setting. Theseresults support the hypothesis that allelopathy maybe a mechanism by which this species is able to out-compete other phytoplankton and form blooms.

Acknowledgements. The authors extend their gratitude toElyse Walker and Jen Goleski for phytoplankton culturingand technical assistance. The manuscript benefited greatlyfrom comments and suggestions provided by several anony-mous reviewers and the editor Dr. Hugh MacIntyre. Weacknowledge the financial support from the National Natu-ral Science Foundation of China (NSFC) (Grant No.U1133003, 40776078), the National High-tech R&D Programof China (Grant No. 2013AA065805), and the New TamarindFoundation.

LITERATURE CITED

Anderson DM, Burkholder JM, Cochlan WP, Glibert PM andothers (2008) Harmful algal blooms and eutrophication: examining linkages from selected coastal regions of theUnited States. Harmful Algae 8: 39−53

Bargu S, Powell CL, Coale SL, Busman M, Doucette GJ, Sil-ver MW (2002) Krill: a potential vector for domoic acid inmarine food webs. Mar Ecol Prog Ser 237: 209−216

Bates SS, Bird CJ, de Freitas ASW, Foxall R and others(1989) Pennate diatom Nitzschia pungens as the primarysource of domoic acid, a toxin in shellfish from easternPrince Edward Island, Canada. Can J Fish Aquat Sci 46: 1203−1215

Bates SS, de Freitas ASW, Milley JE, Pocklington R, Quil-liam MA, Smith JC, Worms J (1991) Controls on domoicacid production by the diatom Nitzschia pungens f. mul-tiseries in culture: nutrients and irradiance. Can J FishAquat Sci 48: 1136−1144

Blossom HE, Andersen NG, Rasmussen SA, Hansen PJ(2014) Stability of the intra-and extracellular toxins ofPrymnesium parvum using a microalgal bioassay. Harm-ful Algae 32: 11−21

Boyd PW, Watson AJ, Law CS, Abraham ER and others(2000) A mesoscale phytoplankton bloom in the polarSouthern Ocean stimulated by iron fertilization. Nature407: 695−702

Boyd PW, Jickells T, Law CS, Blain S and others (2007)Mesoscale iron enrichment experiments 1993−2005: synthesis and future directions. Science 315: 612−617

Burkholder JM, Glibert PM (2009) The importance of intra-specific variability in harmful algae: preface to a collec-tion of topical papers. Harmful Algae 8: 744−745

Casteleyn G, Leliaert F, Backeljau T, Debeer AE and others(2010) Limits to gene flow in a cosmopolitan marineplanktonic diatom. Proc Natl Acad Sci USA 107: 12952−12957

Coale KH, Johnson KS, Fitzwater SE, Gordon RM and others(1996) A massive phytoplankton bloom induced by anecosystem-scale iron fertilization experiment in theequatorial Pacific Ocean. Nature 383: 495−501

Coale KH, Johnson KS, Chavez FP, Buesseler KO and others(2004) Southern Ocean iron enrichment experiment: carbon cycling in high- and low-Si waters. Science 304: 408−414

Doblin MA, Blackburn SI, Hallegraeff GM (1999) Growthand biomass stimulation of the toxic dinoflagellateGymnodinium catenatum (Graham) by dissolved organicsubstances. J Exp Mar Biol Ecol 236: 33−47

Figueredo CC, Giani A, Bird DF (2007) Does allelopathycontribute to Cylindrospermopsis raciborskii (cyano -bacteria) bloom occurrence and geographic expansion?J Phycol 43: 256−265

Fistarol GO, Legrand C, Granéli E (2003) Allelopathic effectof Prymnesium parvum on a natural plankton commu-nity. Mar Ecol Prog Ser 255: 115−125

Gervais F, Riebesell U, Gorbunov MY (2002) Changes in pri-mary productivity and chlorophyll a in response to ironfertilization in the Southern Polar Frontal Zone. LimnolOceanogr 47: 1324−1335

Granéli E, Salomon PS (2010) Factors influencing allelo -pathy and toxicity in Prymnesium parvum. J Am WaterResour Assoc 46: 108−120

Hattenrath-Lehmann TK, Gobler CJ (2011) Allelopathicinhibition of competing phytoplankton by North Amer -ican strains of the toxic dinoflagellate, Alexandriumfundyense: evidence from field experiments, laboratoryexperiments, and bloom events. Harmful Algae 11: 106−116

Heisler J, Glibert P, Burkholder J, Anderson DM and others(2008) Eutrophication and harmful algal blooms: a scien-tific consensus. Harmful Algae 8: 3−13

Ianora A, Miralto A (2010) Toxigenic effects of diatoms ongrazers, phytoplankton and other microbes: a review.Ecotoxicology 19: 493−511

Jonsson PR, Pavia H, Toth G (2009) Formation of harmfulalgal blooms cannot be explained by allelopathic inter-actions. Proc Natl Acad Sci USA 106: 11177−11182

Keating KI (1977) Allelopathic influence on blue-greenbloom sequence in a eutrophic lake. Science 196: 885−887

Kotaki Y, Koike K, Yoshida M, Thuoc CV and others (2000)Domoic acid production in Nitzschia sp. isolated from ashrimp-culture pond in Do Son, Vietnam. J Phycol 36: 1057−1060

Landry MR, Ondrusek ME, Tanner SJ, Brown SL and others(2000) Biological response to iron fertilization in the east-ern equatorial Pacific (IronEx II). I. Microplankton com-munity abundances and biomass. Mar Ecol Prog Ser 201: 27−42

Legrand C, Rengefors K, Fistarol GO, Granéli E (2003)Allelopathy in phytoplankton — biochemical, ecologicaland evolutionary aspects. Phycologia 42: 406−419

Lundholm N, Hansen PJ, Kotaki Y (2005) Lack of allelo-pathic effects of the domoic acid-producing marinediatom Pseudo-nitzschia multiseries. Mar Ecol Prog Ser288: 21−33

Maldonado MT, Hughes MP, Rue EL, Wells ML (2002)The effect of Fe and Cu on growth and domoic acid pro-duction by Pseudo-nitzschia multiseries and Pseudo-nitzschia australis. Limnol Oceanogr 47: 515−526

Martin JL, Haya K, Burridge LE, Wildish DJ (1990) Nitzschiapseudodelicatissima − a source of domoic acid in the Bayof Fundy, eastern Canada. Mar Ecol Prog Ser 67: 177−182

Parsons TR, Maita Y, Lalli CM 1984. A manual of chemicaland biological methods for seawater analysis. PergamonPress, Oxford

40

http://dx.doi.org/10.3354/meps067177http://dx.doi.org/10.4319/lo.2002.47.2.0515http://dx.doi.org/10.3354/meps288021http://dx.doi.org/10.2216/i0031-8884-42-4-406.1http://dx.doi.org/10.3354/meps201027http://dx.doi.org/10.1046/j.1529-8817.2000.99209.xhttp://dx.doi.org/10.1126/science.196.4292.885http://dx.doi.org/10.1073/pnas.0900964106http://dx.doi.org/10.1007/s10646-009-0434-yhttp://dx.doi.org/10.1016/j.hal.2008.08.006http://dx.doi.org/10.1016/j.hal.2011.08.005http://dx.doi.org/10.1111/j.1752-1688.2009.00395.xhttp://dx.doi.org/10.4319/lo.2002.47.5.1324http://dx.doi.org/10.3354/meps255115http://dx.doi.org/10.1111/j.1529-8817.2007.00333.xhttp://dx.doi.org/10.1016/S0022-0981(98)00193-2http://dx.doi.org/10.1126/science.1089778http://dx.doi.org/10.1038/383495a0http://dx.doi.org/10.1073/pnas.1001380107http://dx.doi.org/10.1016/j.hal.2009.03.006http://dx.doi.org/10.1126/science.1131669http://dx.doi.org/10.1038/35037500http://dx.doi.org/10.1016/j.hal.2013.11.006http://dx.doi.org/10.1139/f91-137http://dx.doi.org/10.1139/f89-156http://dx.doi.org/10.3354/meps237209http://dx.doi.org/10.1016/j.hal.2008.08.017

Xu et al.: Allelopathic effects of Pseudo-nitzscha spp. 41

Pedersen MF, Hansen PJ (2003) Effects of high pH on thegrowth and survival of six marine heterotrophic protists.Mar Ecol Prog Ser 260: 33−41

Peloquin JA, Smith WO (2006) The role of phytoplanktonsize on photochemical recovery during the southernocean iron experiment. J Phycol 42: 1016−1027

Poulson KL, Sieg RD, Prince EK, Kubanek J (2010) Allelo-pathic compounds of a red tide dinoflagellate have spe-cies-specific and context-dependent impacts on phyto-plankton. Mar Ecol Prog Ser 416: 69−78

Prince EK, Myers TL, Kubanek J (2008) Effects of harmfulalgal blooms on competitors: allelopathic mechanisms ofthe red tide dinoflagellate Karenia brevis. LimnolOceanogr 53: 531−541

Prince EK, Poulson KL, Myers TL, Sieg RD, Kubanek J(2010) Characterization of allelopathic compoundsfrom the red tide dinoflagellate Karenia brevis. HarmfulAlgae 10: 39−48

Qiu X, Yamasaki Y, Shimasaki Y, Gunjikake H and others(2011) Growth interactions between the raphidophyteChattonella antiqua and the dinoflagellate Akashiwosanguinea. Harmful Algae 11: 81−87

Rynearson TA, Armbrust EV (2000) DNA fingerprintingreveals extensive genetic diversity in a field popu -lation of the centric diatom Ditylum brightwellii. LimnolOceanogr 45: 1329−1340

Schmidt LE, Hansen PJ (2001) Allelopathy in the prymnesio-phyte Chrysochromulina polylepis: effect of cell concen-tration, growth phase and pH. Mar Ecol Prog Ser 216: 67−81

Scholin CA, Gulland F, Doucette GJ, Benson S and others(2000) Mortality of sea lions along the central Californiacoast linked to a toxic diatom bloom. Nature 403: 80−84

Smayda TJ (1997) Harmful algal blooms: their ecophysiol-ogy and general relevance to phytoplankton blooms inthe sea. Limnol Oceanogr 42: 1137−1153

Stonik IV, Orlova TY, Shevchenko OG (2001) Morphologyand ecology of the species of the genus Pseudo-nitzschia(Bacillariophyta) from Peter the Great Bay, Sea of Japan.Russ J Mar Biol 27: 362−366

Subba Rao DV, Quilliam MA, Pocklington R (1988) Domoicacid — a neurotoxic amino acid produced by the marinediatom Nitzschia pungens in culture. Can J Fish AquatSci 45: 2076−2079

Suikkanen S, Fistarol GO, Granéli E (2004) Allelopathiceffects of the Baltic cyanobacteria Nodularia spumigena,Aphanizomenon flos-aquae and Anabaena lemmerman-nii on algal monocultures. J Exp Mar Biol Ecol 308: 85−101

Sun YY, Xu SZ, Li WH, Zhang J, Wang CH (2012) Antialgalsubstances from Isochrysis galbana and its effects on thegrowth of Isochrysis galbana and six species of feedmicroalgae. Inf Tech Agric Eng 134: 211−223

Tang YZ, Gobler CJ (2010) Allelopathic effects of Cochlo-dinium polykrikoides isolates and blooms from the estu-aries of Long Island, New York, on co-occurring phyto-plankton. Mar Ecol Prog Ser 406: 19−31

Tang YZ, Koch F, Gobler CJ (2010) Most harmful algalbloom species are vitamin B1 and B12 auxotrophs. ProcNatl Acad Sci USA 107: 20756−20761

Tillmann U (2003) Kill and eat your predator: a winningstrategy of the planktonic flagellate Prymnesium par -vum. Aquat Microb Ecol 32: 73−84

Tillmann U, Alpermann T, John U, Cembella A (2008) Alle-lochemical interactions and short-term effects of thedinoflagellate Alexandrium on selected photoautotro -phic and heterotrophic protists. Harmful Algae 7: 52−64

Tillmann U, Alpermann TL, Da Purificação RC, Krock B,Cembella A (2009) Inter-population clonal variability inallelochemical potency of the toxigenic dinoflagellateAlexandrium tamarense. Harmful Algae 8: 759−769

Trainer VL, Bates SS, Lundholm N, Thessen AE, Adams NG,Cochlan WP, Trick CG (2012) Pseudo-nitzschia physio-logical ecology, phylogeny, toxicity, monitoring andimpacts on ecosystem health. Harmful Algae 14: 271−300

Trick CG, Bill BD, Cochlan WP, Wells ML, Trainer VL, Pick-ell LD (2010) Iron enrichment stimulates toxic diatomproduction in high-nitrate, low-chlorophyll areas. ProcNatl Acad Sci USA 107: 5887−5892

Windust AJ 1992. The response of bacteria, microalgae andzooplankton to the diatom Nitzschia pungens f. multi-series and its toxic metabolite domoic acid. MSc thesis,Dalhousie University, Halifax

Yamasaki Y, Nagasoe S, Matsubara T, Shikata T, ShimasakiY, Oshima Y, Honjo T (2007) Allelopathic interactionsbetween the bacillariophyte Skeletonema costatum andthe raphidophyte Heterosigma akashiwo. Mar Ecol ProgSer 339: 83−92

Editorial responsibility: Hugh MacIntyre, Halifax, Nova Scotia, Canada

Submitted: January 13, 2014; Accepted: October 3, 2014Proofs received from author(s): December 9, 2014

http://dx.doi.org/10.3354/meps339083http://dx.doi.org/10.1073/pnas.0910579107http://dx.doi.org/10.1016/j.hal.2011.10.025http://dx.doi.org/10.1016/j.hal.2009.03.005http://dx.doi.org/10.1016/j.hal.2007.05.009http://dx.doi.org/10.3354/ame032073http://dx.doi.org/10.1073/pnas.1009566107http://dx.doi.org/10.3354/meps08537http://dx.doi.org/10.1007/978-3-642-27537-1_27http://dx.doi.org/10.1016/j.jembe.2004.02.012http://dx.doi.org/10.1139/f88-241http://dx.doi.org/10.1023/A%3A1013701403968http://dx.doi.org/10.4319/lo.1997.42.5_part_2.1137http://dx.doi.org/10.1038/47481http://dx.doi.org/10.3354/meps216067http://dx.doi.org/10.4319/lo.2000.45.6.1329http://dx.doi.org/10.1016/j.hal.2011.08.001http://dx.doi.org/10.1016/j.hal.2010.06.003http://dx.doi.org/10.4319/lo.2008.53.2.0531http://dx.doi.org/10.3354/meps08788http://dx.doi.org/10.1111/j.1529-8817.2006.00266.xhttp://dx.doi.org/10.3354/meps260033

cite43: cite5: cite56: cite14: cite42: cite3: cite27: cite55: cite41: cite39: cite12: cite40: cite25: cite53: cite38: cite11: cite24: cite52: cite37: cite23: cite51: cite36: cite6: cite49: cite22: cite50: cite35: cite4: cite48: cite21: cite34: cite19: cite47: cite62: cite20: cite33: cite18: cite61: cite59: cite32: cite45: cite58: cite31: cite16: cite9: cite29: cite44: cite7: cite57: