Complex interactions between nutrient enrichment and zooplankton in regulating estuarine phytoplankton assemblages: Microcosm experiments informed by an environmental dataset

Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

Contents lists available at ScienceDirect

Journal of Experimental Marine Biology and Ecology journal homepage: www.elsevier.com/locate/jembe

Complex interactions between nutrient enrichment and zooplankton in regulating estuarine phytoplankton assemblages: Microcosm experiments informed by an environmental dataset Megan B. Rothenberger a,⁎, Alyssa J. Calomeni b a b

Biology Department, Lafayette College, Kunkel Hall, Easton, PA 18042, USA School of Agricultural, Forest and Environmental Sciences, Clemson University, G17 Lehotsky Hall, Clemson, SC 29634, USA

a r t i c l e

i n f o

Article history: Received 13 February 2015 Received in revised form 22 March 2016 Accepted 23 March 2016 Available online xxxx Keywords: Raritan Bay Microcosm Phytoplankton Nutrient enrichment Zooplankton grazing Eutrophication

a b s t r a c t The interactive effects of nutrients and zooplankton grazing on phytoplankton assemblage composition and bloom formation are not well understood, especially in coastal ecosystems. Therefore, the objective of this study was to evaluate phytoplankton assemblage responses to changing environmental conditions in Raritan Bay, an estuary between the states of New York and New Jersey with a long history of cultural eutrophication and harmful algal blooms (HABs). Environmental monitoring of water quality and plankton species composition (monthly data collected from April 2010–April 2013), multivariate ordination techniques, and microcosm experiments were integrated to achieve this objective. Multivariate analysis of the monitoring dataset led to the generation of a number of hypotheses regarding phytoplankton composition, individual nuisance species, environmental factors, and zooplankton composition. The field observations were then supplemented with a series of controlled microcosm experiments designed to test those hypotheses. In particular, the effects and interaction of varying mesozooplankton abundance (sieved, unsieved and enriched) and enrichment with two nutrients (nitrate and dissolved iron) on spring and summer phytoplankton assemblages were evaluated. The environmental monitoring data and the results of the microcosm experiments both indicate that Si:N ratios are important factors governing phytoplankton dynamics in Raritan Bay and that effects of nutrient enrichment on phytoplankton species composition are magnified when mesozooplankton abundance is low. The microcosm experiments, however, led to unexpected results regarding the influence of iron on phytoplankton assemblages. Iron did not emerge as one of the environmental parameters related to spring and summer phytoplankton species composition in the multivariate analysis of field data. Yet, enrichment with iron in microcosms resulted in significant increases in diatoms, and enrichment with both iron and nitrate resulted in significant increases in dinoflagellates and HAB taxa, including Heterocapsa triquetra (Ehrenberg) Stein and Dinophysis spp. This finding suggests that small pulses of iron, in combination with low Si:N ratios (b1) and low mesozooplankton abundance, would shift phytoplankton assemblages in Raritan Bay toward greater dinoflagellate dominance. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Cultural eutrophication, the nutrient over-enrichment of surface waters by human activity, has been described as one of the most pressing problems of the 21st century (Turner and Rabalais, 2003). More than 60% of coastal waters in the United States are exhibiting the symptoms of eutrophication which include dense, and sometimes toxic, algal blooms, depletion of dissolved oxygen (DO), and loss of marine biodiversity (National Research Council [NRC], 2000). Therefore, it is not surprising that eutrophication has represented and defined one of the central themes of coastal research since the 1970s when scientists first began expressing serious concern about the impact of nutrient ⁎ Corresponding author. E-mail address: [email protected] (M.B. Rothenberger).

http://dx.doi.org/10.1016/j.jembe.2016.03.015 0022-0981/© 2016 Elsevier B.V. All rights reserved.

enrichment on the structure and function of coastal ecosystems (Nixon, 1995). Cloern (2001) described three phases in the development of our conceptual model of the cultural eutrophication problem in coastal ecosystems. Most of the early research conducted during Phase I of the conceptual model emphasized the relationship between nutrient loading, increasing phytoplankton biomass, and enhanced depletion of oxygen from bottom waters. Phase II, the contemporary phase of the conceptual model, recognizes that the processes that determine primary productivity and food web structure vary considerably from one system to the next. As a result, Phase II emphasizes system-specific responses to nutrient enrichment and includes research aimed at clarifying the way that nutrient loading, nutrient ratios, and habitat characteristics interact to affect the abundance and composition of the phytoplankton and overall food web structure. There is no question that a number of studies in different coastal ecosystems have

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

provided important Phase II insights (e.g., Smayda, 1989; Iriogen et al., 2005; Burkholder et al., 2006; Mitra and Flynn, 2006; Sunda et al., 2006). Cloern (2001) described a third phase of the eutrophication model that includes consideration of the interaction between nutrient enrichment and other stressors (e.g., invasive species, climate change) and application of our scientific understanding of eutrophication to develop restoration strategies. Certain pieces of the Phase II model still remain poorly understood despite our common understanding of the eutrophication concept and the vast knowledge gained over the past four decades. One part of the Phase II model that needs further clarification is the origin and development of harmful algal blooms. Coastal scientists generally agree that increased nutrient enrichment is promoting the development of many HABs (Heisler et al., 2008), but the environmental processes that select for blooms of certain species are still not well understood, especially in estuaries. Reviews of HAB research indicate that both the quantity and quality of the nutrient pool can contribute to the success of individual bloom species and that both chronic and episodic nutrient delivery can promote HABs (Anderson et al., 2002; Glibert and Burkholder, 2006; Heisler et al., 2008). Furthermore, the ultimate success of a given species, and its response to nutrient enrichment, depends upon the rate at which zooplankton grazers are consuming that species and constraining its production. Studies have shown that selective feeding by zooplankton on certain species can alter the structure of a phytoplankton assemblage that developed under a specific nutrient regime (e.g., Porter, 1977; Turner and Tester, 1989; Gobler et al., 2002; Iriogen et al., 2005). Although this research has permitted generalizations to be drawn about the role of nutrients and grazers in generating HABs, more Phase II eutrophication studies are needed to provide system- and season-specific insights about how “bottom up” (nutrient concentrations and ratios) and “top down” (zooplankton abundance and composition) factors interact to create change in phytoplankton species composition and lead to bloom formation. Therefore, the overall goal of this study was to combine environmental monitoring with microcosm experiments to investigate the combined effects of nutrient enrichment and zooplankton grazers on the phytoplankton species composition in Raritan Bay, a eutrophic estuary located at the southern portion of Lower New York Bay between the states of New York and New Jersey. Three years of data on water quality and plankton species composition (i.e., monthly sampling from April–November) were collected from six sites within Raritan Bay. Ordination techniques, which have been described as a group of methods for data reduction leading to hypothesis generation (Kent and Coker, 1992; McCune and Grace, 2002), were then used to examine complex relationships among nutrients, zooplankton composition, and phytoplankton composition. Ordination of the environmental dataset led to the following hypotheses regarding plankton dynamics in Raritan Bay: 1) abundance of HAB species is associated with reduced mesozooplankton grazing pressure, 2) river discharge and Si:N ratios, particularly during spring months, are positively related to diatom abundance and negatively related to abundance of flagellates, including certain HAB species (e.g., Heterosigma akashiwo (Hada) Hada ex Hada et Chihara, Prorocentrum minimum (Pavillard) Schiller, Pfiesteria-like species), and 3) iron is not an important predictor of phytoplankton species composition in this system. A series of microcosm experiments, driven by these hypotheses, were designed to examine the effect and interaction of three zooplankton treatments (i.e., sieved, unsieved and enriched) and enrichment with two nutrients (nitrate and dissolved iron) on spring and summer phytoplankton assemblages. Although microcosms have been criticized for yielding misleading results when they exclude important features of ecosystems (e.g., secondary consumers, seasonal effects; Carpenter, 1996), they are among the few tools available for studying the systemic responses of phytoplankton to nutrient enrichment in the presence of grazers (Cloern, 2001). Microcosm experiments are often more constructive when they are based on results of larger, field

63

monitoring programs, as they can be used to test hypotheses and study the effects of disturbance (Carpenter, 1996). Because this system has been the site of an ongoing monitoring program since 2010, results of experimental manipulations can be related back to patterns observed under natural conditions. 2. Methods 2.1. Study site The Raritan Bay system (New York–New Jersey, U.S.A.; Fig. 1) is both ecologically and economically significant, supporting numerous recreational and commercial fisheries and providing habitat for waterfowl, shellfish, and marine, estuarine, and anadromous fish (Kane and Kerlinger, 1994). The water quality and ecological integrity of this system, however, have long been threatened by a dense human population and urban and industrial overdevelopment in the watershed. By the turn of the 20th Century, shoreline development, discharges of industrial waste, and inadequate waste disposal methods led to water contamination, beach closures, reduced fish abundance and diversity, and collapse of the oyster industry (Environmental Protection Agency [EPA], 2007). By the early 1960s, the rich nutrient supplies and sluggish circulation of the Raritan Bay were contributing to extremely dense phytoplankton populations and low summer DO concentrations (Jeffries, 1962). Although a number of studies in this system have demonstrated some improvements to water quality (e.g., increases in oxygen and decreases in fecal coliforms) following passage of the Clean Water Act in 1972 and improvements to sewage treatment (Brosnan and O'Shea, 1996; O'Shea and Brosnan, 2000), more recent water quality data indicate that the system is still polluted (United States Environmental Protection Agency (US EPA), 2007; Rothenberger et al., 2014). In fact, full utilization of these waters for commercial fishing and recreational activities is restricted by poor water quality (US EPA, 2007). The water and sediments are contaminated with a variety of pollutants such as heavy metals, polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAH), and pesticides (Breteler, 1984). Nitrate and soluble reactive phosphorus (SRP) concentrations (grand mean 12 μM for nitrate and 0.5 μM for SRP) in Raritan Bay continue to indicate eutrophic conditions and are as much as 50 and 20 times higher, respectively, than concentrations reported 50 years ago (Jeffries, 1962; Rothenberger et al., 2014). The Raritan Bay also continues to exhibit numerous symptoms of eutrophication, including high algal biomass, high turbidity, seasonal hypoxia, fish kills, and blooms of potentially harmful phytoplankton species (Reid et al., 2002; Rothenberger et al., 2014). Chronic phytoplankton blooms of various species, including the dinoflagellates Ceratium tripos (Müller) Nitzsch, Prorocentrum micans Ehrenberg, and Heterocapsa rotundata (Lohmann) Hansen, and the raphidophyte H. akashiwo (formerly referred to as Olisthodiscus luteus Carter, Smayda, 1998; Olsen and Mahoney, 2001), have occurred in the Hudson Raritan Estuary (HRE) and New Jersey coastal waters for over three decades (Gastrich, 2000). These blooms have been associated with moderate bather discomfort and/or illness, diminished aesthetic value of beaches, hypoxia, and fish mortality (Gastrich, 2000). 2.2. Environmental monitoring and ordination Six sites in Raritan Bay (Fig. 1) were sampled on a monthly basis from April through November. The water quality samples collected and analyzed to inform the microcosm experiments described below were gathered from April 2010 through April 2013. Water temperature, salinity, pH, and dissolved oxygen (DO) were determined using a YSI 6820 V2 multiparameter meter, which was calibrated prior to each sample period. Water transparency was assessed using Secchi depth data (Wetzel and Likens, 2001). A vertical polycarbonate Van Dorn water sampler (2.2 L) was thoroughly rinsed with site water and used

64

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

Fig. 1. The Raritan River Basin and Bay, showing the six sampling stations used for monitoring of water quality trends and plankton composition. Monthly monitoring began in April 2010 and is ongoing. Microcosms were incubated near site 6.

to collect water samples near the surface (~ 0.5 m), at mid-depth (~1.5 m), and low-depth (~3.0 m) for biological and chemical analyses. Samples for nutrient analyses were poured into acid-cleaned bottles, maintained in darkness on ice for transport to the laboratory, and refrigerated or frozen as appropriate until analysis (EPA, 1993). Phytoplankton samples were poured into amber bottles and preserved with acidic Lugol's solution on site. Fresh, unpreserved phytoplankton samples were also brought back to the laboratory to aid microscopic identification. Zooplankton samples were collected at a depth of 1.5 m using a 12-liter Schindler-Patalas trap and preserved with 4% buffered formalin. All plankton samples were held at 4 °C until analysis. Nutrient analyses were performed using standard methods (American Public Health Association, APHA et al., 2012). Concentrations of all nutrients were determined colorimetrically with a HACH (Loveland, CO) DR/2500 spectrophotometer. Accuracy checks, including the use of reagent blanks and standard solution adjusts, followed HACH procedures (Hach Manual for DR/2500 spectrophotometer). NO− 3 was determined within 12 h of collection using the method of cadmium reduction, and the Nessler method was used within 24 h for determination of ammonium (NH+ 4 ). Samples for analysis of SRP were first filtered

in the field with a Whatman Puradisc 0.45 μm syringe filter and analyzed within 48 h using the ascorbic acid-molybdate blue method, and SiO2 was determined using the silicomolybdate method. The 1,1,10 phenanthroline method was used to determine ferrous iron (hereinafter just called iron or dissolved iron). Samples for analysis of dissolved iron were analyzed as soon as possible after collection (i.e., within 8 h) to prevent oxidation of ferrous to ferric iron. Samples for analysis of total iron were preserved with concentrated nitric acid to adjust the pH to ~ 2 and sent to Penn State Agricultural Analytical Services laboratory for analysis via inductively coupled plasma-atomic emission spectrometry as per EPA method 200.7. Phytoplankton counts were conducted using a Palmer Maloney chamber (400 ×) under a Leica DM 1000 phase contrast microscope. Cells were identified to the lowest taxon possible with light microscopy, and taxa abundances, including colonial and filamentous forms, were quantified by enumerating single cells. Many identifications were taken to genus or species level, but for some flagellates, individual genera were difficult to consistently differentiate from one another in preserved samples and were mostly identified to the class level (i.e., Cryptophyceae, Prymnesiophyceae,

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

Prasinophyceae, Raphidophyceae). Cryptoperidiniopsoids, Pfiesteria piscicida Steidinger and Burkholder, Pfiesteria shumwayae Glasgow and Burkholder (following Marshall et al., 2006), and other physically similar taxa were classified as “Pfiesteria-like” under light microscopy (Seaborn et al., 2006). For enumeration of zooplankton, a subsample of ~ 10% of the total sample was examined (Goswami, 2004). Counts were conducted using a Sedgewick Rafter counting cell (100 ×), also with a Leica DM 1000 phase contrast microscope. Identifications were taken to genus or species level, and juvenile or nauplii stages of zooplankton were counted separately from their adult stage because of differential feeding patterns (Johnson and Allen, 2005) and difficulty in identifying them to genus level under light microscopy. Phytoplankton and zooplankton data were compiled into separate matrices of cell number by site and sample date, resulting in a final matrix of 78 columns of phytoplankton cell counts and 40 columns of zooplankton counts by 144 sample rows (3 yr × 8 mo × 6 sites). To reduce the “noise” (variability) in the plankton datasets and enhance detection of assemblage patterns in relation to environmental parameters, rare taxa (defined as present in b5% of samples) were removed prior to analysis (McCune and Grace, 2002), and abundances of the remaining 41 phytoplankton species and 32 zooplankton species were log transformed after 1 was added as a constant. Every plankton sample used in the analysis had a corresponding suite of physical and chemical measurements (described above) that could be related to ordinations by species composition. Precipitation data were also acquired from the New Jersey Weather & Climate Network, and precipitation was determined for each sample by summing the total amount of rainfall at the New Brunswick station (i.e., the nearest municipal area) for the week prior to the sampling date. Raritan River discharge rates were acquired from United States Geological Survey (USGS) at Bound Brook, which is approximately 20 km upstream from the upper edge of the study area at site 1. Season (April–May = spring; June–September = summer; October–November = autumn), site, and year were all included in the Nonmetric Multidimensional Scaling (NMDS) analyses (described below) as categorical variables. Ordination techniques, which order samples along axes expressing the main trends or gradients in the data, were used to investigate potential environmental predictors of phytoplankton and zooplankton assemblage patterns. Non-Metric Multidimensional Scaling (NMDS), considered the most effective ordination method for ecological data (McCune and Grace, 2002), was used to establish seasonal and site-tosite differences in plankton assemblage structure, and to interpret those differences in terms of environmental conditions. All NMDS ordinations were carried out using PC-ORD version 5.31 software (MjM software 2010). A one-way ANOVA (p b 0.05) was also used to compare bloom and non-bloom conditions in Raritan Bay. A bloom was defined as 2 standard deviations above the multiyear mean (4.5 × 104 cells ml−1). Bloom samples were further differentiated into those dominated by diatoms (relative abundance of diatoms was N 50%) and those dominated by other taxa. Approximately 90% of the non-diatom blooms were composed of mixed flagellates (chlamydomonad-like species, Cryptomonas spp., Euglena spp., H. rotundata, H. akashiwo, and P. micans were most abundant). Because zooplankton abundance, nitrate concentration, and Si:N ratio were strongly related to phytoplankton species composition in the ordination of the full dataset, these parameters were compared between bloom and non-bloom samples. Although iron was not related to phytoplankton species composition in the analysis of the full dataset or in the analysis of spring and summer samples, iron was included in this study for a number of reasons. First, it emerged as a parameter related to Raritan Bay phytoplankton species composition in autumn (Rothenberger et al., 2014). In addition, iron supply controls many aspects of algal physiology, including photosynthesis, nitrate assimilation, and N2-fixation (Sunda, 2006). A number of studies indicate that, because different phytoplankton species have different iron

65

requirements, iron supply can be important in regulating phytoplankton productivity and species composition in both oceanic and coastal ecosystems (Brand et al., 1983; Sunda and Huntsman, 1995; Hutchins and Bruland, 1998). The effect of iron on phytoplankton species composition is much less clear in the coastal environment, yet episodic pulses of iron have been associated with red tide flagellate blooms in coastal ecosystems (Glover, 1978; Yamochi, 1983; Doucette and Harrison, 1991). Because the relative impact of iron supply on coastal phytoplankton and HAB species is still largely unknown (Boyer and Brand, 1998; Sunda, 2006), the concentration of iron was compared between bloom and non-bloom conditions in Raritan Bay. Results of both the ordination and ANOVA analyses informed the design of the microcosm experiments described below. 2.3. Microcosm experiments Microcosms are artificial, simplified ecosystems that can be used to predict the response of natural phytoplankton assemblages to specific actions under controlled conditions (Beyers and Odum, 1993). Three different sets of microcosm experiments were used to assess the effect of grazers and nutrients on phytoplankton biomass and composition. The first two experiments were designed to assess the effect of mesozooplankton grazers on phytoplankton composition. Mesozooplankton are defined here as the size fraction of zooplankton between 160–2000 μm in maximum length (Sieburth, 1978). In Raritan Bay, this group may include copepods, polychaete larvae, barnacle nauplii, cladocerans, and some rotifers (e.g., Asplanchna spp.). Although microzooplankton grazing (i.e., grazing by heterotrophic and mixotrophic organisms b 160 μm in size) is generally accepted as being the main predatory pressure on planktonic primary production (Calbet and Landry, 2004; Iriogen et al., 2005; Calbet, 2008), mesozooplankton are also an important influence on phytoplankton composition and food web structure (Calbet, 2008). In addition, multiyear monitoring of phytoplankton-zooplankton relationships in Raritan Bay indicated an inverse relationship between mesozooplankton and certain bloom-forming taxa, such as H. akashiwo, P. minimum (Pavillard) Schiller, and Pfiesteria-like dinoflagellates (Rothenberger et al., 2014). The third experiment was designed to assess the effects and interactions of nitrate, iron, and mesozooplankton. All three microcosm experiments were carried out in late spring and summer because multiyear monitoring data indicated that phytoplankton biomass generally peaks in late spring and the majority of the identified bloom-forming species are most abundant during summer months (Rothenberger et al., 2014). Experimental manipulations to assess the effect of mesozooplankton grazers on phytoplankton biomass and composition were initiated on 9 July 2010 and 4 September 2010. These experiments consisted of two treatments and a control, each with three replicates. For the sieved treatment, mesozooplankton grazers (N 160 μm) were removed (i.e., 0 organisms per L−1), and, for the enriched treatment, total zooplankton density was increased to approximately three times the natural concentration (i.e., ~3 × 102 organisms per L−1) and mesozooplankton density was approximately doubled (i.e., ~ 30 organisms per L−1, Table 1). For unsieved control containers, total zooplankton density (i.e., ~ 1 × 102 organisms per L− 1) and mesozooplankton density (i.e., ~15 organisms per L−1) were maintained (Table 1). Natural plankton assemblages and water for the microcosms were collected at site 6 (Fig. 1) using a vertical polycarbonate Van Dorn water sampler (2.2 L) that had been thoroughly rinsed with bay water prior to collection. Sample water was either poured directly into acidcleaned buckets (i.e., for unsieved and enriched treatments) or through a ~ 160 μm net into a third acid-cleaned bucket (i.e., for sieved treatments without mesozooplankton). Because they cannot be removed by sieving as they are similar in size to phytoplankton, microzooplankton were still present in sieved treatments (Table 2).

66

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

Table 1 Comparison of in situ conditions at site 6 with initial treatments effects for enclosure experiments conducted in 2010, which were designed to test the effect of grazers alone. Values are replicate means (SEs), except for in situ zooplankton densities which were sampled at a depth of 1.5 m at site 6 on the days of the enclosure experiments. All nutrient concentrations are in μM, and zooplankton densities are the number of organisms per L−1. July 2010

September 2010

Parameter

In Situ

Sieved

Unsieved

Enriched

In situ

Sieved

Unsieved

Enriched

Nitrate Silicon Si:N SRP Iron Total zooplankton Microzooplankton Mesozooplankton

5.38 (2.22) 17.1 (7.31) 3.57 (0.80) 0.38 (0.02) 0.30 (0.05) 113 101 12

5.16 (1.07) 15.2 (2.33) 3.34 (0.91) 0.57 (0.01) 0.52 (0.02) 50.6 (3.53) 50.6 (3.53) 0.00 (0.00)

7.26 (2.09) 20.0 (2.88) 2.84 (0.15) 0.42 (0.02) 0.90 (0.01) 106 (3.18) 94.3 (3.53) 11.0 (1.15)

6.45 (1.25) 13.9 (1.61) 2.11 (0.03) 0.46 (0.01) 1.23 (0.10) 295 (14.3) 276 (10.0) 18.3 (4.33)

7.14 (4.52) 17.0 (6.42) 1.98 (0.33) 0.37 (0.03) 1.25 (0.20) 252 244 8

5.38 (2.34) 18.3 (1.66) 5.50 (2.53) 0.35 (0.04) 1.25 (0.78) 207 (7.69) 207 (7.69) 0.00 (0.00)

8.06 (3.23) 19.9 (3.84) 5.31 (3.55) 0.56 (0.06) 1.97 (0.27) 214 (6.00) 198.7 (6.36) 15.3 (0.67)

7.53 (4.40) 14.4 (1.11) 4.0 (2.17) 0.32 (0.04) 0.78 (0.16) 316 (4.93) 281 (5.36) 34.7 (1.67)

For the enriched treatment, additional zooplankton were collected with a Schindler-Patalas trap lowered 1.5 m below the water surface and transferred into the “enriched” bucket. Microcosms were 1-L transparent polyethylene containers that were acid-cleaned and rinsed with bay water before experiments to eliminate impurities. Containers were sequentially filled from the respective buckets a third at a time to improve initial homogeneity among containers (Carter et al., 2005). The order of filling the containers was random, and an air bubble was left in the containers to aid in mixing by wave action. All samples were weighted with a 15 mL centrifuge tube filled with steel shot, attached to separate ropes in random order, and suspended from a yellow surface buoy that was anchored to the sea floor on a 40 kg concrete block. The surface buoy and weights allowed the bottles to remain at ~ 1 m depth, regardless of the tide. The microcosms were incubated in situ for one day near site 6 of our monitoring study out of the way of boat traffic. Phytoplankton and zooplankton abundance were determined before and after the first two experiments using the methods described above in the monitoring section. The third enclosure experiment was initiated on 17 May 2013. It was arranged in a 23 full-factorial design with eight manipulation types: no added nutrients/ambient zooplankton density, added nitrate/ambient zooplankton, added iron/ambient zooplankton, added iron and nitrate/ ambient zooplankton, no added nutrients/mesozooplankton absent (water collected was sieved through a ~ 160 μg net to remove mesozooplankton grazers), added nitrate/mesozooplankton absent, added iron/mesozooplankton absent, added nitrate and iron/ mesozooplankton absent. Each treatment consisted of three replicates. Nitrogen was added as sodium nitrate (NaNO3) to increase nitrate concentrations by approximately 10 μM, and iron was added as ferric sulfate (Fe2(SO4)3) to increase dissolved iron concentrations by approximately 1 μM. The nitrogen additions lowered Si:N ratios from ~ 3 to b 1 (Table 3). Nutrients were added to containers, alone or in combination, on day 0 as a single pulse. Samples were collected and microcosms were prepared using the same methods as described above for the first two experiments except that, for logistical reasons, the third experiment

Table 2 A comparison of “bloom” and “non-bloom” conditions in Raritan Bay as indicated by multiyear monitoring dataset. A bloom is defined as 2 standard deviations above the multiyear mean (4.5 × 104 cells ml−1). Values are means (SEs). For each comparison, ANOVA F-statistics and P-values are shown (df = 2). All nutrient concentrations are in μM, and zooplankton densities are the number of organisms per L−1. Parameter

No bloom

Diatom bloom

Non-diatom bloom

P

F

Microzooplankton Mesozooplankton Nitrate Si:N Iron

137 (15) 42 (3) 13 (0.9) 5 (0.6) 2 (0.2)

378 (63) 196 (21) 6 (0.8) 10 (3) 4 (0.8)

1084 (182) 9 (1) 24 (4) 1 (0.4) 0.9 (0.4)

b0.001 b0.001 b0.001 0.01 b0.001

79 110 11 5 8

concluded after seven days. Phytoplankton and zooplankton abundance, SiO2, nitrate, dissolved iron, and SRP were determined before and after the experiments using the methods described above.

2.4. Microcosm statistical analyses For the enclosure experiments conducted in 2010, a one-way ANOVA (p b 0.05) was used to compare the effect of zooplankton abundance on change in phytoplankton abundance and species richness between sieved treatments (n = 3), unsieved control treatments (n = 3), and enriched grazer treatments (n = 3). Change was calculated as final abundance/initial abundance or final richness/initial richness. To evaluate compositional changes in phytoplankton among the three treatments, changes in abundance of cyanobacteria, diatoms, dinoflagellates, other flagellates, and taxa causing HABs were also used as dependent variables. HAB taxa are those that have been identified as capable of causing harmful algal blooms (i.e., they cause harm either through toxin production, cell physical structure, or accumulated biomass; Anderson et al., 2002). If a significant effect was determined using a one-way ANOVA, then Tukey's HSD (honest significant difference) tests were used to make post hoc comparisons among the three zooplankton treatments. For the enclosure experiment conducted in 2012, a three-way ANOVA was used with the independent factors being experimental nitrate level (ambient and + 10 μM), experimental iron level (ambient and + 1 μM), and mesozooplankton density (sieved and unsieved). The dependent variables were change in total phytoplankton abundance, species richness, diatom abundance, dinoflagellate abundance, and the abundance of HAB taxa. Cyanobacteria and other flagellates were very low in abundance in the initial samples from these microcosms and therefore were not included as dependent variables for this experiment. Data were square root transformed prior to analysis to fulfill the requirements of normality (tested by Shapiro–Wilk's W test) and homogeneity of variances. In addition to using a three-way ANOVA to examine the main effects and interactions of mesozooplankton abundance, nitrate concentration, and iron concentration on phytoplankton abundance and taxonomic groups, indicator species analysis (Dufrêne and Legendre, 1997) was performed on the phytoplankton dataset (29 taxa; log-transformed change in abundance) by microcosm matrix (8 treatments × 3 replicates = 24 microcosms) in order to examine compositional changes at the species and genus level. Phytoplankton indicator taxa (abundant and consistently “faithful” to a particular group; significant index value p ≤ 0.05) were determined by running indicator species analysis in PC-ORD version 5.31 software separately by mesozooplankton abundance (sieved and unsieved), nitrate concentration (groups = ambient nitrate and enriched nitrate), and iron concentration (groups = ambient iron and enriched iron).

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

67

Table 3 Comparison of in situ conditions at site 6 with initial treatments effects for the enclosure experiment conducted in May 2013, which was designed to assess the effects and interactions of both nutrients and mesozooplankton grazers. Values are replicate means (SEs), except for in situ zooplankton densities which were sampled at a depth of 1.5 m at site 6 on the days of the enclosure experiments. All nutrient concentrations are in μM, and zooplankton densities are the number of organisms per L−1. Parameter

In situ

U, None

U, Nitrate

U, Iron

U, Both

S, None

S, Nitrate

S, Iron

S, Both

Nitrate Silicon Si:N SRP Iron Total Zooplankton Microzooplankton Mesozooplankton

10.8 (2.90) 26.6 (6.93) 2.57 (0.56) 0.37 (0.05) 1.07 (0.10) 83 56 27

9.98 (0.68) 29.2 (1.96) 2.93 (0.01) 0.34 (0.05) 1.01 (0.24) 119 (6.67) 82.7 (3.38) 36.7 (8.45)

24.3 (3.13) 19.3 (0.88) 0.84 (0.16) 0.26 (0.04) 0.78 (0.12) 65.7 (33.7) 44.3 (22.8) 21.3 (10.8)

9.31 (0.41) 37.5 (1.22) 4.05 (0.27 0.22 (0.04) 3.22 (0.53) 71.7 (5.78) 56.7 (7.80) 15.0 (2.08)

26.3 (3.00) 20.3 (2.49) 0.77 (0.01) 0.40 (0.04) 1.96 (0.11) 79.7 (7.13) 64.7 (6.64) 15.0 (1.53)

11.7 (1.93) 84.7 (5.36) 7.80 (1.01) 0.15 (0.03) 1.13 (0.12) 56.6 (4.70) 56.6 (4.70) 0.00 (0.00)

27.7 (6.23) 19.3 (1.93) 0.77 (0.19) 0.34 (0.04) 0.89 (0.34) 74.0 (10.3) 74.0 (10.3) 0.00 (0.00)

13.3 (2.33) 40.7 (2.19) 3.23 (0.55) 0.35 (0.06) 2.32 (0.21) 67.3 (4.41) 67.3 (4.41) 0.00 (0.00)

19.7 (3.93) 11.5 (1.38) 0.67 (0.23) 0.42 (0.04) 2.56 (0.68) 59.0 (6.43) 58.0 (7.23) 1.00 (1.00)

3. Results 3.1. Monitoring results Multivariate analysis of monitoring data from Raritan Bay indicates that abundance of most diatoms and the dinoflagellates, H. rotundata and Heterocapsa triquetra, is greatest during spring months when precipitation, river discharge, Si:N ratios, and mesozooplankton concentrations are high (Fig. 2). In fact, most spring diatom blooms occurred when Si:N ratios were around 10 (Table 2). The largest phytoplankton blooms occurred during late summer when salinity was high and Si:N ratios and mesozooplankton abundance were low (Fig. 2). These summer blooms were composed primarily of flagellates, including the potentially harmful H. akashiwo, P. minimum, and Pfiesteria-like dinoflagellates (Fig. 2), and they corresponded with significantly lower mesozooplankton abundance (i.e., b 10 organisms per L− 1) and Si:N ratios (i.e., ≤1, Table 2). Although iron did not emerge in the ordination results as an environmental factor significantly related to phytoplankton species composition during spring and summer in Raritan Bay, a comparison of bloom and non-bloom conditions indicates that iron concentrations are significantly higher during diatom blooms (Table 2).

(237 cfs) for July and 5.7 m3 s− 1 (201 cfs) for September, which is below average (monthly averages measured by United States Geological Survey at Bound Brook ~ 20 km upstream from study area are 18.9 m3 s−1 [667 cfs] for July and 21.0 m3 s−1 [742 cfs] for September). Average ambient conditions during summer 2010 at site 6 were 22 °C, 9.0 mg DO L−1, 6 μM nitrate, 0.4 μM SRP, and 1.1 × 102 zooplankton individuals per L−1 (Table 1). Zooplankton assemblages were dominated by rotifers and copepods in July and by rotifers, copepods and ciliates in September (Fig. 3). These conditions are typical of this system during summer months (Rothenberger et al., 2014), but salinity (mean 34) was higher than average and most likely reflected reduced river discharge rates associated with drought conditions (Rothenberger et al., 2014). In contrast, during the enclosure experiment conducted in May 2013, precipitation totals and river discharge were near average (Office of the New Jersey State Climatologist [ONJSC], 2013). Average ambient conditions during May 2013 at site 6 were 14 °C, a salinity of 27, 10.0 mg DO L−1, 10 μM nitrate, a Si:N ratio of 2, 0.4 μM SRP, 1 μM dissolved iron, and ~1 × 102 zooplankton per L−1 (Table 3). Zooplankton assemblages were dominated by copepods during the third enclosure experiment (Fig. 3). Conditions in May 2013 were near average for this system during late spring (Rothenberger et al., 2014).

3.2. Environmental conditions during microcosm experiments 3.3. Effect of zooplankton grazers alone During the enclosure experiments conducted in July and September 2010, drought conditions were affecting the coastal counties of the Raritan River basin (Office of the New Jersey State Climatologist [ONJSC], 2010). As a result, Raritan River discharge rates were 6.7 m3 s− 1

The mesozooplankton that were removed from sieved containers consisted primarily of polychaete larvae and the copepods Acartia spp. and Eurytemora spp. (Fig. 3). In the first two experiments, increasing

Fig. 2. Ordination of all samples by phytoplankton species composition showing seasonal differences in environmental conditions and phytoplankton species composition. Data from the three years of sampling were averaged so that each point in the diagram represents monthly mean abundance (n = 4) of all phytoplankton species from a given site and month. (A) Vectors indicate strength and direction of environmental gradients (r2 cutoff value = 0.15; Si:N = dissolved Si:N ratio, SRP = soluble reactive phosphorus) (B) Vectors indicate strength and direction of phytoplankton species gradients (r2 cutoff value = 0.40). Abbreviated taxa are Ditylum brightwellii, Hetercapsa rotundata, Asterionellopsis glacialis, Heterocapsa triquetra, Leptocylindrus danicus, Prorocentrum minimum, Leptocylindrus minimus, Cylindrotheca closterium, and Heterosigma akashiwo.

68

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

were large centric and chain-forming diatoms (i.e., Coscinodiscus spp. and Leptocylindrus minimus), Cylindrotheca closterium (Ehrenberg), and several dinoflagellates, including the HAB taxa H. triquetra, Karlodinium spp., and Scrippsiella trochoidea (Stein) Loeblich III (Table 7). There was also a significant effect of nitrate addition (i.e., Si:N ratios) and iron addition on phytoplankton abundance and composition (Table 6). There was a significantly greater increase in total phytoplankton abundance and diatom abundance at ambient nitrate concentrations (p b 0.01) when Si:N ratios were also higher (Fig. 4A and C), but dinoflagellates and HAB taxa increased to a greater extent when nitrate was enriched and Si:N ratios were b1(p b 0.05 for HAB taxa) (Fig. 4D and E). Various pennate diatoms and green flagellates in the genus Pyramimonas were indicators of ambient nitrate concentrations, whereas centric and chain-forming diatoms were indicators of enriched nitrate concentrations (Table 7). Potentially harmful dinoflagellates in the genus Dinophysis also emerged as indicators of enriched nitrate concentrations and lowered Si:N (Table 7). Iron addition also resulted in a significantly greater increase in total phytoplankton abundance and the abundance of dinoflagellates and HAB taxa (p b 0.0001) (Table 6, Fig. 4A, D, and E). Again, various diatoms and Pyramimonas spp. were indicators of ambient iron concentrations and centric diatoms and Dinophysis spp. were indicators of enriched iron concentrations (Table 7). Results of the three-way ANOVA also indicate that interactions among the experimental factors produced significant changes in phytoplankton abundance and composition (Table 6). There was a significant two-way interaction of zooplankton abundance and nitrate concentration on total phytoplankton abundance and diatom abundance (p b 0.01) (Table 6). The pattern of this interaction was that total phytoplankton and diatom abundance increased to a greater extent in microcosms with ambient nitrate concentrations, and this effect was stronger when mesozooplankton were removed from microcosms (Fig. 4A and C). There was also a significant two-way interaction of zooplankton abundance and iron concentration on total phytoplankton abundance (p b 0.001) (Table 6). When microcosms contained ambient concentrations of zooplankton, the difference in the change in phytoplankton abundance between ambient iron and iron enriched microcosms was small (Fig. 4A, D, and E). When mesozooplankton were removed, the addition of iron resulted in a significantly greater increase in total phytoplankton abundance (Fig. 4A). There was also a significant interaction of nitrate and iron concentration on phytoplankton composition (Table 6), but the results of this interaction differed depending on the algal group (Fig. 4). For example, the change in diatom abundance was similar in ambient and enriched nitrate treatments when iron concentrations were lower (Fig. 4C). Pennate diatoms, particularly the chain-forming Asterionellopsis glacialis (Castracane) Round, Navicula spp., and Nitzschia spp., dominated microcosms with ambient nitrate and iron concentrations and higher Si:N ratios. All diatoms increased to a greater extent in microcosms in which nitrate was at ambient levels and iron was enriched (Fig. 4C). Enriched iron, but not nitrate concentrations, was associated with a significant increase in centric and chain-forming diatoms, including Coscinodiscus spp., Skeletonema costatum (Greville) P.T. Cleve, and Thalassiosira gravida P.T. Cleve. On the other hand, dinoflagellates and HAB taxa, including H. triquetra and Dinophysis spp., increased to a

Fig. 3. Comparison of in situ (IS) abundance of zooplankton groups compared with abundance of zooplankton groups in sieved (S), unsieved (U), and enriched (E) microcosms during experimental trials in July 2010 (J), September 2010 (S), and May 2013 (M).

mesozooplankton abundance caused a significant decrease in total phytoplankton abundance (Tables 4 and 5). Although mesozooplankton abundance did not have a statistically significant effect on phytoplankton species richness (Table 4), the number of species in the microcosms did decrease in the ambient and enriched treatments (Table 5). The change in abundance of cyanobacteria, diatoms, and HAB taxa was positive (i.e., final abundance was greater than initial abundance), but the magnitude of the positive change for each of these algal groups decreased with increasing mesozooplankton abundance (Table 5). Statistical analyses indicated that mesozooplankton abundance did not have a significant effect on the abundance of cyanobacteria, diatoms, dinoflagellates, or HAB taxa (Table 4). After incubation, treatments with higher grazer densities had significantly lower abundance of other flagellates, which included chlorophytes, cryptomonads, euglenoids, haptophytes, and raphidophytes (Tables 4 and 5). 3.4. Effects and interactions of both grazers and nutrients There was a significant effect of mesozooplankton abundance on phytoplankton abundance and composition (Table 6). There was a significantly greater increase in total phytoplankton abundance when mesozooplankton were absent from the microcosms (p b 0.0001) (Fig. 4A). Similar to the first two experiments, the mesozooplankton that were removed from sieved containers in the third experiment consisted primarily of polychaete larvae and the copepods (Fig. 3). Zooplankton abundance did not have a significant effect on phytoplankton species richness or diatom abundance, but the removal of mesozooplankton from the microcosms resulted in a significantly greater increase in dinoflagellates and HAB taxa (Table 6, Fig. 4D and E). Indicator species analysis suggested that the best indicators of lowered zooplankton abundance, or taxa that were consistently abundant in microcosms in which mesozooplankton grazers were removed, Table 4 Summary of one-way ANOVA. Source

df Abundance MS

Zooplankton abundance 2

F

Richness p

MS

F

p

Cyanobacteria

Diatoms

MS

MS

F

p

F

p

Dinoflagellates

Other flagellates

HAB taxa

MS

MS

MS

F

p

F

p

F

p

1.11 15.90 0.000⁎ 0.17 2.79 0.093 2.51 2.31 0.133 3.78 1.18 0.334 0.06 0.31 0.739 4.75 5.97 0.012⁎ 227 1.42 0.273

The independent factor is zooplankton abundance with three levels (i.e., mesozooplankton absent, at ambient levels, or enriched). Dependent variables are change in total phytoplankton abundance, species richness, cyanobacteria abundance, diatom abundance, dinoflagellate abundance, abundance of other flagellates, and abundance of harmful algal bloom (HAB) taxa. ⁎ p b 0.05.

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

69

Table 5 Mean (n = 6; replicates for experiments 1 and 2 combined) ± 1SE change in total phytoplankton abundance (cells ml−1), species richness, cyanobacteria abundance, diatom abundance, dinoflagellate abundance, abundance of other flagellates, and abundance of harmful algal bloom (HAB) taxa in experimental treatments over seven days. Change calculated as final abundance/initial abundance or final richness/initial richness. Treatment

Total abundance⁎

Species richness

Cyanobacteria abundance

Diatom abundance

Dinoflagellate abundance

Abundance other flagellates⁎

HAB taxa abundance

Sieved Unsieved Enriched

1.7 ± 0.2a 1.4 ± 0.1a 0.9 ± 0.1b

1.0 ± 0.2 0.7 ± 0.1 0.8 ± 0.1

2.4 ± 0.6 1.2 ± 0.1 1.3 ± 0.3

3.0 ± 0.4 2.6 ± 1.7 1.5 ± 0.5

0.6 ± 0.2 0.5 ± 0.2 0.7 ± 0.2

1.9 ± 0.6a 2.1 ± 0.2a 0.5 ± 0.2 b

2.2 ± 0.8 1.2 ± 0.2 1.1 ± 0.4

⁎ Significant effect determined using one-way ANOVA. Different letter superscripts indicate that means are significantly different (one-way ANOVA followed by Tukey's HSD).

greater extent in microcosms in which both nitrate and iron were enriched (Fig. 4D and E). Finally, there was a significant three-way interaction on all of the dependent variables measured (Table 6). For each of the dependent variables, the pattern that resulted from nutrient enrichment was magnified when mesozooplankton were removed from the microcosms (Fig. 4). At ambient zooplankton concentrations, nutrient enrichment had little effect on phytoplankton abundance and composition (Fig. 4). When mesozooplankton are removed, however, enriched concentrations of both nutrients resulted in a significant increase in dinoflagellates and HAB taxa whereas enriching iron but not nitrate concentrations resulted in a significant increase in diatoms (Fig. 4C, D, and E). 4. Discussion Our water quality monitoring study in Raritan Bay represents one of the few available studies, and the first for the Raritan Bay ecosystem, to include accurate and consistent data over a 4-year period (i.e., data collection began in April 2010 and is ongoing) on phytoplankton assemblage composition in relation to both nutrients and zooplankton composition. Although our multiyear dataset cannot yet be considered “long-term” (i.e., 10–40 years), it has been valuable both for documenting seasonal, spatial, and temporal change in this system and for generating and analyzing testable hypotheses. The microcosm experiments described here were driven by these hypotheses and verified by the results of the larger field program. The first hypothesis that microcosm experiments were designed to test is that top-down control effects of mesozooplankton grazers on phytoplankton abundance and composition are significant in Raritan Bay. Data from our multi-year monitoring study indicate that blooms of potentially harmful flagellates correspond with low mesozooplankton concentrations (i.e., b 10 organisms per L−1), and results of the microcosm experiments support this hypothesis. All three microcosm experiments generally indicate that phytoplankton increased to a significantly greater extent when mesozooplankton were absent from the microcosms, but results from the first set of experiments seem to reflect the importance of microzooplankton grazing for controlling abundance of certain

phytoplankton groups (as in Iriogen et al., 2005). For example, manipulations of mesozooplankton grazers did not have a significant effect on diatom abundance, and the abundance of flagellates was only significantly reduced when both microzooplankton and mesozooplankton abundance was increased in enriched microcosms. Other studies indicate that the microzooplankton, including heterotrophic dinoflagellates, are the primary consumers of diatoms and small flagellates (Calbet, 2008). In the third experiment, however, mesozooplankton abundance was either maintained at ambient concentrations or removed, and microzooplankton abundance was relatively consistent across treatments. Again, phytoplankton abundance increased during the incubation period in each microcosm, but the net increase in phytoplankton abundance was significantly greater when mesozooplankton were removed from microcosms. Furthermore, the interaction between lowered grazer abundance and enriched nutrients had an interesting effect on phytoplankton assemblages. First, the pattern that resulted from nutrient enrichment (described below) was magnified when mesozooplankton were removed from the microcosms. This finding may indicate that mesozooplankton grazing can mask the effects of nutrient loadings on phytoplankton community structure. When mesozooplankton were present in the microcosms, the effect of enriching nitrate and iron on phytoplankton abundance and composition was minimal. When mesozooplankton were removed to mimic situations where high predation by fish, ctenophores, and scyphomedusae occurs (Granéli and Turner, 2002), nutrient enrichment resulted in significant changes in the phytoplankton assemblage. In other words, across treatments the nutrient pool apparently selected for growth of a particular group of species, but in the absence of strong top-down control exerted by grazers, the relationship between nutrients and phytoplankton abundance and composition was more apparent (Reynolds, 1984; Sterner, 1989; Cottingham and Schindler, 2000; Glibert and Burkholder, 2006). Our hypotheses regarding the relationship between nutrients and phytoplankton abundance and composition were also based upon results of the multiyear monitoring study. Multivariate analysis indicated that river discharge and Si:N ratios, particularly during spring months, are positively related to diatom abundance and negatively related to abundance of flagellates, including certain HAB species. Therefore, the

Table 6 Summary of three-way ANOVA. Source

df

Abundance

Richness

Diatoms

Dinoflagellates

MS

F

p

MS

F

p

MS

F

p

MS

F

p

MS

F

p

Zooplankton (Z) Nitrate (N) Iron (I) Z∗N Z∗I N∗I Z∗N∗I Residual

1 1 1 1 1 1 1 16

2.34 0.64 1.11 0.29 0.80 0.68 0.18 0.03

78.00 21.33 37.00 9.67 26.67 22.67 6.00

0.000⁎ 0.000⁎ 0.000⁎ 0.007⁎ 0.000⁎ 0.000⁎ 0.026⁎

0 0.02 0.01 0.01 0.04 0.02 0.19 0.01

0 2.00 1.00 1.00 4.00 2.00 19.00

1.000 0.176 0.332 0.332 0.063 0.176 0.000⁎

0.08 3.45 0.42 2.24 0.00 2.12 2.67 0.24

0.33 14.38 1.75 9.33 0.00 8.83 11.13

0.574 0.002⁎ 0.204 0.008⁎

5.42 0.10 1.24 0.03 0.04 0.72 0.70 0.03

180.67 3.33 41.33 1.00 1.33 24.00 23.33

0.000⁎ 0.087 0.000⁎

4.70 0.26 1.00 0.00 0.00 1.14 1.10 0.04

117.50 6.50 25.00 0.00 0.00 28.50 27.50

0.000⁎ 0.021⁎ 0.000⁎

1.000 0.009⁎ 0.004⁎

HAB taxa

0.332 0.265 0.000⁎ 0.000⁎

1.000 1.000 0.000⁎ 0.000⁎

Independent factors are zooplankton abundance, nitrate concentration, and iron concentration, each with two levels. Dependent variables are change in total phytoplankton abundance, species richness, diatom abundance, dinoflagellate abundance, and abundance of harmful algal bloom (HAB) taxa. ⁎ p b 0.05.

70

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

Fig. 4. Mean (n = 3) change in total phytoplankton abundance (A), species richness (B), diatom abundance (C), dinoflagellates abundance (D), and abundance of HAB taxa (E) in experimental treatments over seven days. Change calculated as final abundance/initial abundance or final richness/initial richness.

second hypothesis tested using microcosms is that Si:N ratios have an important influence on phytoplankton species composition in Raritan Bay. An increase in nitrate inputs and decrease in Si:N ratios (i.e., Si:N b 1) were expected to shift phytoplankton assemblages toward dominance by non-diatoms. The third experiment addressed this hypothesis through the addition of nitrate without manipulating Si concentration, which was meant to mimic episodic anthropogenic

nutrient loading. Again, results of the microcosm experiments support this hypothesis. The nutrient pool in these microcosm experiments influenced whether diatoms or dinoflagellates became dominant, especially when grazer abundance was lowered. Ambient nitrate and iron concentrations and higher Si:N ratios were associated with increases in the abundance of pennate diatoms, whereas enriched iron, but not nitrate, was

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

71

Table 7 Results of indicator species analysis (n = 24) for zooplankton abundance, nitrate concentration, and iron concentration. A significance level of p ≤ 0.05 was used to determine the indicator species for each group. Ambient (control)

Treatment (absent zooplankton, enriched nutrients)

Zooplankton abundance

Asterionellopsis glacialis, Thalassiosira gravida

Nitrate concentration Iron concentration

Asterionellopsis glacialis, Navicula spp., Nitzschia spp., Pyramimonas spp. Asterionellopsis glacialis, Ditylum brightwellii, Nitzschia spp., Pyramimonas spp.

Coscinodiscus spp., Cylindrotheca clausterium, Gymnodinium spp., Heterocapsa triquetra, Karlodinium spp., Leptocylindrus minimus, Scrippsiella trochoidea Dinophysis spp., Leptocylindrus minimus, Rhizosolenia spp. Coscinodiscus spp., Dinophysis spp., Thalassiosira gravida

associated with a significant increase in centric and chain-forming diatoms. The diatom L. minimus, however, emerged as an indicator of enriched nitrate concentrations. When mesozooplankton were removed, enriched concentrations of both nitrate and iron and lowered Si:N ratios were associated with increases in dinoflagellates, including the HAB taxa H. triquetra and Dinophysis spp. These findings are consistent with spatial and temporal patterns that have emerged through ordination of our multiyear monitoring dataset. The phytoplankton taxa that emerged as indicators of lower nitrate and iron concentrations are generally smaller in size with greater surface-to-volume ratios. Previous laboratory and field studies have demonstrated that small size is generally associated with higher maximum growth rates (Sarthou et al., 2005) and lower halfsaturation constants for uptake of both nitrate (Eppley et al., 1969) and iron (Timmermens et al., 2004). Therefore, small cells or cells with high surface-to-volume ratios with greater efficiency in nutrient uptake may have had a competitive advantage when nitrate and iron concentrations were lower in the microcosms. Si availability was also likely to have been regulating the abundance of diatoms within microcosms. Mean Si:N ratios were significantly lower in microcosms that were enriched with nitrate (M = 0.7 ± 0.05) compared with those that were not enriched with nitrate (M = 4 ± 1.2, p = 0.03). This may explain why enrichment with iron but not nitrate led to a significant increase in most diatoms but enrichment with both iron and nitrate, and therefore a reduction in Si:N ratios, led to a significant increase in dinoflagellates. As diatom growth rates declined with lowered Si:N ratios, dinoflagellates, including potentially toxic Dinophysis spp., became more important in the assemblage. These results correspond with the nutrient ratio hypothesis that increased loading of N relative to Si would lead to a proportional shift away from a diatom-dominated assemblage toward dominance by flagellates (Smayda, 1980, 1989). These results are also consistent with patterns described in other coastal systems (Dortch et al., 2001; Smayda et al., 2004). In fact, Dinophysis spp., a group of species capable of causing diarrhetic shellfish poisoning (DSP; Yasumoto et al., 1980) and most abundant in Raritan Bay in late spring according to our monitoring dataset, appeared as an indicator species of both enriched nitrate and iron. Similar shifts in phytoplankton species composition have been measured in both experimental and field monitoring studies. In enclosure experiments, diatom abundance and productivity are significantly higher in treatments amended with N, P, and Si than in treatments amended only with N and P (Egge and Aksnes, 1992; Domingues et al., 2011). For example, Domingues et al. (2011) documented an increase in the potentially harmful dinoflagellate Kryptoperidinium foliaceum (F. Stein) Lindemann in microcosms enriched with nitrate in the absence of Si. Long-term field studies in eutrophic estuaries have also documented a decline in the relative abundance of diatoms and an increase in the relative abundance of flagellates as N and P (but not Si) concentrations increase (Conley et al., 1993). The diatom L. minimus did not seem to follow the overall pattern for diatoms in our microcosm experiments in Raritan Bay. One possible explanation is that Leptocylindrus has a higher half-saturation constant (Ks) for uptake of nitrate than other Raritan Bay diatoms and therefore reproduces more effectively at higher nitrate concentrations (Eppley et al., 1969). In addition, studies suggest that Leptocylindrus has a

significantly lower level of silicification than some of the other seasonally abundant diatoms in Raritan Bay (i.e., A. glacialis, Coscinodiscus spp., S. costatum, and Thalassiosira spp.; Rousseau et al., 2002). This may explain why L. minimus was more abundant, and therefore an indicator species, of enriched nitrate and lowered Si:N ratios. A similar pattern has been found for Rhizosolenia pungens Cleve-Euler, another weakly silicified diatom, in Comau Fjord (Iriarte et al., 2013). Because R. pungens appears to be adapted to lower silicic acid and higher nitrogen environmental conditions, Iriarte et al. (2013) suggest that this diatom be considered a functional phytoplankton species indicative of coastal waters impacted by human activities. Finally, the hypothesis that iron is not an important predictor of phytoplankton species composition in this system was not supported by the results of this microcosm study. This result is interesting given the different opinions about the role of iron supply in regulating coastal phytoplankton dynamics. It is well known that iron additions in openocean environments stimulate phytoplankton productivity and bloom potential (e.g., Martin and Fitzwater, 1988; Coale et al., 1996). By comparison, it has been suggested that iron is unlikely to exert any major regulation on phytoplankton dynamics in coastal ecosystems such as Raritan Bay since the amount of iron supplied by terrestrial soils is relatively high (e.g., Martin et al., 1989; Reynolds, 2006). On the other hand, episodic pulses of iron from land drainage or sediment resuspension have been implicated in red tide dinoflagellate blooms (Doucette and Harrison, 1991) and blooms of H. akashiwo (Yamochi, 1983). Other studies suggest that nitrate stimulation of coastal phytoplankton production is enhanced in the presence of iron and that this synergistic interaction reflects the essential role of iron in nitrate reduction (Paerl, 1997). In this study, additions of iron to microcosms led to significant increases in overall phytoplankton abundance, and, as described above, interactions between iron and nitrate enrichment determined whether diatoms or dinoflagellates increased in abundance. Our monitoring dataset indicates that late spring and early summer phytoplankton assemblages are dominated by certain diatoms (i.e., A. glacialis, Coscinodiscus spp., and Leptocylindrus) and, to a much lesser extent, dinoflagellates (i.e., Dinophysis spp., H. triquetra, H. rotundata). This microcosm study suggests that episodic pulses of both nitrate and iron, either from land drainage, sediment resuspension or heavy rainfall, would shift spring and summer phytoplankton assemblages in Raritan Bay toward greater dinoflagellate dominance. Because other microcosm enrichment studies demonstrate considerable seasonal variation in phytoplankton species responses to nutrient enrichment (Carter et al., 2005; Domingues et al., 2011; Iriarte et al., 2013), it would be valuable to assess the effects and interactions of nutrients and zooplankton grazers on phytoplankton assemblages during other times of the year and under varying climatic conditions. This study suggests that the relative impact of microzooplankton and mesozooplankton on the phytoplankton assemblage is seasonally variable in Raritan Bay (i.e., mesozooplankton appeared to have a greater impact in spring and microzooplankton in late summer). Finally, microcosm experiments are limited by the in situ conditions of the ecosystem. For example, H. akashiwo has formed the largest blooms in Raritan Bay during our monitoring study; however, this species was not documented in samples during microcosm experiments. As a result, it was not possible to assess the interacting effects of nutrients and grazers on this important bloom-former in Raritan Bay.

72

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73

5. Conclusions While it is often difficult to extrapolate experimental results to system-level scenarios due to the complexity of the forces governing phytoplankton dynamics in estuaries, our approach – combining routine field monitoring (Rothenberger et al., 2014) with controlled microcosm experiments (this study) – synthesizes evidence from a variety of sources. Our small-volume, short-term nutrient addition experiments indicate that shifts in phytoplankton species composition following nutrient enrichment were magnified when mesozooplankton were removed from microcosms; enrichment with iron only resulted in significant increases in diatoms; and enrichment with both iron and nitrate resulted in significant increases in dinoflagellates and the centric diatom L. minimus. The results of the microcosm experiments mostly confirm the patterns that have emerged through ordination of our multiyear monitoring dataset. Blooms of potentially harmful flagellates have coincided with periods of lowered mesozooplankton biomass, and nitrate concentrations and Si:N ratios are important factors governing phytoplankton dynamics in Raritan Bay. Lower nitrate concentrations and higher Si:N ratios are correlated with increased abundance of diatoms such as A. glacialis and Thalassiosira spp., and increased nitrate and lower Si:N ratios are correlated with increased abundance of flagellates. One surprising result of this study is the significant effect of iron on coastal phytoplankton assemblages. Based on the results of our multivariate analysis of the field dataset, iron did not seem to be an important factor explaining variation in phytoplankton structure during spring. In contrast, microcosm results suggest that small pulses of iron can affect phytoplankton growth and community structure, even in nutrient-enriched estuaries, and this finding warrants further examination. In sum, these results demonstrate that top-down and bottom-up factors interact in complex ways to alter phytoplankton community structure and both must be considered when attempting to use monitoring data to predict a HAB. Acknowledgements We thank Captain Mick Trzaska, captain and owner of the CRT II, for transporting us to our sampling sites and for technical support in the field. We thank Phil Auerbach for technical support in the laboratory and with field equipment, and Jeffrey Hollander, Raphael Cuomo, Danielle Sobol, Carly Feiro, Shane Foye, Alexander Pong, Ryan Hughes, Elizabeth Cole, and Joshua Hitchings for assistance with water sampling and nutrient analyses. We thank an anonymous reviewer for counsel on the manuscript. Funding support for this research was provided by the Lafayette College Department of Biology. [SS] References American Public Health Association (APHA), American Water Works Association, Water Environment Federation, 2012. Standard Methods for the Examination of Water and Wastewater. 22nd ed. APHA. Anderson, D.M., Glibert, P.M., Burkholder, J.M., 2002. Harmful algal blooms and eutrophication: nutrient sources, composition, and consequences. Estuaries 25, 704–726. Beyers, R.J., Odum, H.T., 1993. Ecological Microcosms. Springer-Verlag, New York, NY. Boyer, G.L., Brand, L.E., 1998. Trace elements and harmful algal blooms. In: Anderson, D.M., Cembella, A.D., Hallegraeff, G.M. (Eds.), Physiological Ecology of Harmful Algal Blooms. Springer-Verlag, Berlin, pp. 489–508. Brand, L.E., Sunda, W.G., Guillard, R.R.L., 1983. Limitation of marine phytoplankton reproductive rates by zinc, manganese, and iron. Limnol. Oceanogr. 28, 1182–1198. Breteler, J. (Ed.), 1984. Chemical Pollution of the Hudson–Raritan Estuary. National Oceanic and Atmospheric Administration, Rockville, MD. Brosnan, T.M., O'Shea, M.L., 1996. Long-term improvements in water quality due to sewage abatement in the lower Hudson River. Estuaries 19, 890–900. Burkholder, J.M., Dickey, D.A., Kinder, C.A., Reed, R.E., Mallin, M.A., McIver, M.R., Cahoon, L.B., Melia, G., Brownie, C., Smith, J., Deamer, N., Springer, J., Glasgow Jr., H., Toms, D., 2006. Comprehensive trend analysis of nutrients and related variables in a large eutrophic estuary: a decadal study of anthropogenic and climatic influences. Limnol. Oceanogr. 51, 463–487. Calbet, A., 2008. The trophic roles of microzooplankton in marine systems. ICES J. Mar. Sci. 65, 325–331.

Calbet, A., Landry, M., 2004. Phytoplankton growth, microzooplankton grazing, and carbon cycling in marine systems. Limnol. Oceanogr. 49, 51–57. Carpenter, S., 1996. Microcosm experiments have limited relevance for community and ecosystem ecology. Ecology 77, 677–680. Carter, C.M., Ross, A.H., Schiel, D.R., Howard-Williams, C., Hayden, B., 2005. In situ microcosm experiments on the influence of nitrate and light on phytoplankton community composition. J. Exp. Mar. Biol. Ecol. 326, 1–13. Cloern, J.E., 2001. Our evolving conceptual model of the coastal eutrophication problem. Mar. Ecol. Prog. Ser. 210, 223–253. Coale, K.H., Johnson, K.S., Fitzwater, S.E., Gordon, R.M., Tanner, S., Chavez, F.P., Ferioli, L., Sakamoto, C., Rogers, P., Millero, F., Steinberg, P., Nightingale, P., Cooper, D., Cochlan, W.P., Landry, M.R., Constantinou, J., Rollwagen, G., Trasvina, A., Kudela, R., 1996. A massive phytoplankton bloom induced by an ecosystem-scale iron fertilization experiment in the equatorial Pacific Ocean. Nature 383, 495–501. Conley, D.J., Schelske, C.L., Stoermer, E.F., 1993. Modification of the biogeochemical cycle of silica with eutrophication. Mar. Ecol. Prog. Ser. 101, 179–192. Cottingham, K.L., Schindler, D.E., 2000. Effects of grazer community structure on phytoplankton response to nutrient pulses. Ecology 81, 183–200. Domingues, R.B., Anselmo, T.P., Barbosa, A.B., Sommer, U., Galvão, H.M., 2011. Nutrient limitation of phytoplankton growth in the freshwater tidal zone of a turbid, Mediterranean estuary. Estuar. Coast. Shelf Sci. 91, 282–297. Dortch, Q., Rabalais, N., Turner, R.E., Qureshi, N.A., 2001. Impacts of changing Si/N ratios and phytoplankton species composition. In: Rabalais, N., Turner, R.E. (Eds.), Coastal Hypoxia: Consequences for Living Resources and Ecosystems. American Geophysical Union, Washington D.C., pp. 37–48. Doucette, G.J., Harrison, P.J., 1991. Aspects of iron and nitrogen nutrition in the red tide dinoflagellate Gymnodinium sanguineum I. Effects of iron depletion and nitrogen source on biochemical composition. Mar. Biol. 110, 165–173. Dufrêne, M., Legendre, P., 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecol. Monogr. 67, 345–366. Egge, J.K., Aksnes, D.L., 1992. Silicate as regulating nutrient in phytoplankton competition. Mar. Ecol. Prog. Ser. 83, 281–289. Eppley, R.W., Rogers, J.N., McCarthy, J.J., 1969. Half-saturation constants for uptake of nitrate and ammonium by marine phytoplankton. Limnol. Oceanogr. 14, 912–920. Gastrich, M.D., 2000. Harmful Algal Blooms in Coastal Waters of New Jersey. New Jersey Department of Environmental Protection. Glibert, P.M., Burkholder, J.M., 2006. The complex relationships between increasing fertilization of the Earth, coastal eutrophication and proliferation of harmful algal blooms. In: Granéli, E., Turner, J.T. (Eds.), Ecology of Harmful Algae. Springer-Verlag, Berlin, pp. 203–214. Glover, H., 1978. Iron in Maine coastal waters; seasonal variation and its apparent correlation with a dinoflagellate bloom. Limnol. Oceanogr. 23, 534–537. Gobler, C.J., Renaghan, M.J., Buck, N.J., 2002. Impacts of nutrients and grazing mortality on the abundance of Aureococcus anophagefferens during a New York brown tide bloom. Limnol. Oceanogr. 47, 129–141. Goswami, S.C., 2004. Zooplankton Methodology, Collection & Identification: A Field Manual. National Institute of Oceanography, Dona Paula. Granéli, E., Turner, J.T., 2002. Top down regulation in ctenophore–copepod–ciliate–diatom–phytoflagellate communities in coastal waters: a mesocosm study. Mar. Ecol. Prog. Ser. 239, 57–68. Heisler, J., Gilbert, P.M., Burkholder, J.M., Anderson, D.M., Cochlan, W., Dennison, W.C., Dortch, Q., Gobler, C.J., Hell, C.A., Humphries, E., Lewitus, A., Magnien, R., Marshall, H.G., Sellner, K., Stockwell, D.A., Stoecker, D.K., Suddleson, M., 2008. Eutrophication and harmful algal blooms: a scientific consensus. Harmful Algae 8, 3–13. Hutchins, D.A., Bruland, K.W., 1998. Iron-limited diatom growth and Si:N uptake ratios in a coastal upwelling regime. Nature 393, 561–564. Iriarte, J.L., Pantoja, S., González, H.E., Silva, G., Paves, H., Labbé, P., Rebolledo, L., Van Ardelan, M., Häussermann, V., 2013. Assessing the micro-phytoplankton response to nitrate in Comau Fjord (42°S) in Patagonia (Chile), using a microcosms approach. Environ. Monit. Assess. 185, 5055–5070. Iriogen, X., Flynn, K.J., Harris, R.P., 2005. Phytoplankton blooms: a ‘loophole’ in microzooplankton grazing impact. J. Plankton Res. 27, 33–321. Jeffries, H.P., 1962. Environmental characteristics of Raritan Bay, a polluted estuary. Limnol. Oceanogr. 7, 21–31. Johnson, W.S., Allen, D.M., 2005. Zooplankton of the Atlantic Gulf Coasts: A Guide to Their Identification and Ecology. Johns Hopkins University Press, Baltimore, MD. Kane, R., Kerlinger, P., 1994. Raritan Bay Wildlife Habitat Report with Recommendations for Conservation. New Jersey Audubon Society, Bernardsville, NJ. Kent, M., Coker, P., 1992. Vegetation Description and Analysis: A Practical Approach. John Wiley and Sons. Marshall, H.G., Hargraves, P.E., Burkholder, J.M., Parrow, M.W., Elbrächter, M., Allen, E.H., Knowleton, V.M., Rublee, P.A., Hynes, W.L., Egerton, T.A., Remington, D.L., Wyatt, K.B., Lewitus, A.J., Henrich, V.C., 2006. Taxonomy of Pfiesteria (Dinophyceae). Harmful Algae 5, 481–496. Martin, J.H., Fitzwater, S.E., 1988. Iron deficiency limits phytoplankton growth in the north-east Pacific subarctic. Nature 331, 341–343. Martin, J.H., Gordon, R.M., Fitzwater, S.E., Broenkow, W.W., 1989. VERTEX: phytoplankton/iron studies in the Gulf of Alaska. Deep Sea Res. A 36, 649–680. McCune, B., Grace, J.B., 2002. Analysis of ecological communities. MjM Software Design. Mitra, A., Flynn, K.J., 2006. Promotion of harmful algal blooms by zooplankton predatory activity. Biol. Lett. 2, 194–197. National Research Council, 2000. Clean Coastal Waters — Understanding and Reducing the Effects of Nutrient Pollution. National Academy Press, Washington, DC. Nixon, S.W., 1995. Coastal marine eutrophication: a definition, social causes, and future concerns. Ophelia 41, 199–219.

M.B. Rothenberger, A.J. Calomeni / Journal of Experimental Marine Biology and Ecology 480 (2016) 62–73 O'Shea, M.L., Brosnan, T.M., 2000. Trends in indicators of eutrophication in Western Long Island Sound and the Hudson–Raritan Estuary. Estuaries 23, 877–901. Office of the New Jersey State Climatologist, 2010–2013. Past Climate Summaries. http:// climate.rutgers.edu/stateclim/?section=menu&%20target=climate_summaries (Accessed 5 October 2014). Olsen, P., Mahoney, J., 2001. Phytoplankton in the Barnegat Bay-Little Egg Harbor estuarine system: species composition and picoplankton bloom development. J. Coast. Res. Spec. Issue 32, 115–143. Paerl, H.W., 1997. Coastal eutrophication and harmful algal blooms: importance of atmospheric deposition and groundwater as “new” nitrogen and other nutrient sources. Limnol. Oceanogr. 42, 1154–1165. Porter, K.G., 1977. The plant–animal interface in freshwater ecosystems. Am. Sci. 65, 159–170. Reid, R.N., Olsen, P.S., Mahoney, J., 2002. A Compilation of Reported Fish Kills in the Hudson–Raritan Estuary during 1982 through 2001. U.S. Department of Commerce, Northeast Fisheries Science Center (Ref. Doc. 02–09, 16 pp.). Reynolds, C.S., 1984. The Ecology of Freshwater Plankton. Cambridge University Press, New York. Reynolds, C.S., 2006. Ecology of Phytoplankton. Cambridge University Press, Cambridge. Rothenberger, M.B., Swaffield, T., Calomeni, A.J., Cabrey, C.D., 2014. Multivariate analysis of water quality and plankton assemblages in an urban estuary. Estuar. Coasts 37, 695–711. Rousseau, V., Leynaert, A., Daoud, N., Lancelot, C., 2002. Diatom succession, silicification and silicic acid availability in Belgian coastal waters (Southern North Sea). Mar. Ecol. Prog. Ser. 236, 61–73. Sarthou, G., Timmermans, K.R., Blain, S., Tréguer, P., 2005. Growth physiology and fate of diatoms in the ocean: a review. J. Sea Res. 53, 25–42. Seaborn, D.W., Tengs, T., Cerbin, S., Kokocinski, M., Marshall, H.G., 2006. A group of dinoflagellates similar to Pfiesteria as defined by morphology and genetic analysis. Harmful Algae 5, 1–8. Sieburth, J., 1978. Pelagic ecosystem structure: heterotrophic compartments of the plankton and their relationship to plankton size fractions. Limnol. Oceanogr. 23, 1256–1263. Smayda, T.J., 1980. Phytoplankton species succession. In: Morris, I. (Ed.), The Physiological Ecology of Phytoplankton. Blackwell, Oxford, pp. 493–570. Smayda, T.J., 1989. Primary production and the global epidemic of phytoplankton blooms in the sea: a linkage? In: Cosper, E.M., Bricelj, V.M., Carpenter, E.J. (Eds.), Novel Phytoplankton Blooms. Springer-Verlag, Berlin, pp. 449–483

73

Smayda, T.J., 1998. Ecophysiology and bloom dynamics of Heterosigma akashiwo (Raphidophyceae). In: Anderson, D.M., Cembella, A.D., Hallegraeff, G.M. (Eds.), Physiological Ecology of Harmful Algal Blooms. Springer-Verlag, Berlin, pp. 113–131. Smayda, T.J., Borkman, D.G., Beaugrand, G., Belgrano, A.G., 2004. Ecological effects of climate variation in the North Atlantic: phytoplankton. In: Stenseth, N.C., Ottersen, G., Hurrell, J., Belgrano, A.G. (Eds.), Ecological Effects of Climate Variations in the North Atlantic. Oxford University Press, Oxford, pp. 49–54. Sterner, R.W., 1989. The role of grazers in phytoplankton succession. In: Sommer, U. (Ed.), Plankton Ecology. Springer-Verlag, Berlin, pp. 107–170. Sunda, W.G., 2006. Trace metals and harmful algal blooms. In: Granéli, E., Turner, J.T. (Eds.), Ecology of Harmful Algae. Springer-Verlag, Berlin, pp. 203–214. Sunda, W.G., Huntsman, S.A., 1995. Iron uptake and growth limitation in oceanic and coastal phytoplankton. Mar. Chem. 50, 189–206. Sunda, W.G., Granéli, E., Gobler, C.J., 2006. Positive feedback and the development and persistence of ecosystem disruptive algal blooms. J. Phycol. 42, 963–974. Timmermens, K.R., van der Wagt, B., de Baar, H.J.W., 2004. Growth rates, half-saturation constants, and silicate, nitrate, and phosphate depletion in relation to iron availability of four large, open-ocean diatoms from the Southern Ocean. Limnol. Oceanogr. 49, 2141–2151. Turner, R.E., Rabalais, N.N., 2003. Linking landscape and water quality in the Mississippi River Basin for 200 years. Bioscience 53, 563–572. Turner, J.T., Tester, P.A., 1989. Toxic marine phytoplankton, zooplankton grazers, and pelagic food webs. Limnol. Oceanogr. 42, 1203–1214. United States Environmental Protection Agency (EPA), 1993. Methods for determination of inorganic substances in environmental samples. EPA Office of Research and Development, Document EPA/600/R-93/100. United States Environmental Protection Agency (EPA), 2007. Coastal Condition, New York/New Jersey Harbor Estuary Program. EPA Office of Water, National Estuary Program Coastal Condition ReportCoastal Condition. Harbor Estuary Program, New York/New Jersey. Wetzel, R.G., Likens, G.E., 2001. Limnological Analyses. 2nd ed. WB Saunders, New York. Yamochi, S., 1983. Mechanisms for outbreak of Heterosigma akashiwo red tide in Osaka Bay, Japan. J. Oceanogr. Soc. Jpn 39, 310–316. Yasumoto, T., Oshima, Y., Sugawara, W., Fukuyo, Y., Oguri, H., Igarashi, T., Fujita, N., 1980. Identification of Dinophysis fortii as the causative agent of diarrhetic shellfish poisoning. Bull. Jpn. Soc. Sci. Fish. 46, 1405–1411.