Grazing by Karenia brevis on Synechococcus enhances its growth rate and may help to sustain blooms Patricia M. Glibert1,*, JoAnn M. Burkholder2, 1University of Maryland Center for Environmental Science, Horn Point Laboratory, PO Box 775, Cambridge, Maryland 21613, USA2Center for Applied Aquatic Ecology, North Carolina State University, Raleigh, North Carolina 27695, USA
The following research into the sustaining near shore of Red Tide Blooms provides scientific evidence that the Red Tide algae is capable of finding a food source to sustain itself near shore. This research explains how a Red Tide is able to sustain itself near shore for long periods of time.The research entitled, “Grazing by Karenia brevis on Synechococcus enhances its growth rate and may help to sustain blooms”, provides evidence that urea nitrogen run off from land based activities results in the blooms of Synechococcus and as cyanobacterial blooms increase, so too does the potential for Karenia brevis growth to be enhanced. In other words Red Tide blooms are sustained through grazing by Red Tide on the Synechococcus
ABSTRACT: Grazing rates of Karenia brevis Clones CCMP2228 and CCMP2229 were determined in laboratory experiments using Synechococcus sp. Clone CCMP1768 as food. Grazing by K. brevis thus enhances the range of nutritional substrates available to meet its growth requirements, and may play a substantial role in sustaining natural populations in inorganic N-poor waters. With evidence that blooms of Synechococcus can be enhanced due to anthropogenic nutrients, the potential importance of this particulate nutrient source for sustaining red tide blooms in situ is large and may help to resolve the current uncertainty as to how K. brevis blooms are maintained. It can now be hypothesized that as cyanobacterial blooms increase, so too does the potential for Karenia brevis growth to be enhanced and for blooms to be sustained through grazing, especially under the low light conditions associated with bloom self-shading. Recognition of this pathway is at least one step toward reconciling the long-term reported increase in K. brevis blooms (e.g. Brand & Compton 2007) and the tendency for blooms of this species to develop offshore in seemingly oligotrophic waters (e.g. Vargo et al. 2004, 2008)
Grazing by Karenia brevis on Synechococcus enhances its growth rate and may help to sustain blooms
Patricia M. Glibert1,*, JoAnn M. Burkholder2, 1University of Maryland Center for Environmental Science, Horn Point Laboratory, PO Box 775, Cambridge, Maryland 21613, USA2Center for Applied Aquatic Ecology, North Carolina State University, Raleigh, North Carolina 27695, USA
Growth (days to weeks) and uptake rates (hours to days) were assessed. In the growth experiments, K. brevis, previously depleted in nitrogen (N), was grown at 2 light intensities in the presence of varying concentrations of Synechococcus. Under high irradiance (300 µmol photons m–2 s–1; 14 h light: 10 h dark cycle), expo¬nential growth rates approximated those of phototrophic growth without Synechococcus (0.26 to 0.35 d–1). At this irradiance, K. brevis cells in all treatments grew for about 10 d. Under lower irradi¬ance (43 µmol photons m–2 s–1), exponential growth rates of K. brevis cells varied with the enrichment level of Synechococcus, with rates under the highest Synechococcus enrichment level being nearly twice (0.58 d–1) those observed for the high irradiance treatments. Short-term uptake experiments (3 experiments, 2 clones) were done to examine the N-specific rates of grazing of 15N-labeled Syne¬chococcus. N-specific rates of grazing ranged from 9.28 × 10–4 h–1 to 1.22 × 10–2 h–1 and varied with the relative proportion of Synechococcus:K. brevis. These rates represent a range of 0.026 to 2.15 pmol-N K. brevis–1 d–1, or 0.96 to 83.8 Synechococcus K. brevis–1 h–1. Evidence of Synechococcus inside K. brevis was provided by confocal microscopy. Grazing by K. brevis thus enhances the range of nutritional substrates available to meet its growth requirements, and may play a substantial role in sustaining natural populations in inorganic N-poor waters. KEY WORDS: Karenia brevis• Synechococcus• Mixotrophy • Grazing rate • Irradiance • Nitrogen source • Harmful algal bloom • Red tide
INTRODUCTION Karenia brevis is a potentially toxic dinoflagellate that causes red tides in the Gulf of Mexico, and is most prevalent along the southwest Florida shelf (e.g. Vargo et al. 2001, 2004). The toxins associated with these blooms, brevetoxins, have been associated with fish and marine mammal kills and also cause respiratory distress and neurotoxic effects in human consumers of contaminated shellfish (Landsberg 2002). Thus, there is considerable management interest in the causes ofthese blooms. Although there is now consensus that nutrient enrichment from land-based sources has degraded estuarine and coastal marine waters worldwide, leading to alterations in ecosystems and the expansion of many harmful algal blooms (HABs) in many regions (e.g. Glibert et al. 2005, Glibert & Burkholder 2006), the sources of nutrients needed to sustain large K. brevis blooms are still not well understood (Walsh & Steidinger 2001, Brand & Compton 2006, Vargo et al. 2008). While some research has supported the premise that nutrient enrichment can contribute to long-term increases in K. brevis abundance in nearshore areas (Brand & Compton 2007), other studies have reported that blooms of this species generally initiate offshore and are sustained in seemingly oligotrophic waters (e.g. Walsh & Steidinger 2001). Thus, it is unclear whether K. brevis blooms require nearshore nutrient enrichment or can be sustained in relatively nutrient-poor waters (Vargo et al. 2004, 2008, Hu et al. 2006).
Relationships between nutrient export from the coastal zone and the marine ecosystem are complex, involving both direct and indirect pathways mediated through nutrient cycling, trophic interactions, and/or biogeochemical transformations (e.g. Glibert et al. 2005, Glibert & Burkholder 2006). While many HABs are related to inorganic nutrient enrichment, some are apparently not; moreover, inorganic nutrients are not the only cause of bloom stimulation (Glibert et al. 2005). An alternative nutritional strategy for some HABs is mixotrophy, in particular phagotrophy from grazing on algae or other prey (Stoecker et al. 2006). Such mixotrophy is observed in many dinoflagellate species (reviewed by Stoecker 1999, Burkholder et al. 2008). Grazing rates by Karenia brevis have only been published to date in 1 study (Jeong et al. 2005) and there is much to be learned about its ability to graze under a range of conditions. Mixotrophic dinoflagellates ingest a wide variety of prey items including bacteria, other algae and protists, and fish, shellfish, and mammalian tissues (Burkholder et al. 2008). The photosynthetic cyanobacterium Synechococcus has been suggested to be one such food source, due to its ubiquitous nature in both offshore and coastal waters, its potential for high growth rates (Kana & Glibert 1987), and its co-occurrence with many dinoflagellates (e.g. Murrell & Lores 2004). Synechococcus is of particular interest in Florida coastal waters because this species develops large blooms in the Florida Bay region (Phlips & Babylak 1996, Glibert et al. 2004) and is also prevalent along the southwest Florida coast (Heil et al. 2007).
The potential for Synechococcus to be grazed by Karenia brevis under a range of laboratory conditions was explored in the present study. We hypothesized that grazing has the potential to contribute a significant amount of N to sustain growth of K. brevis. The rates of growth with and without this particulate source of nutrients and under varying light regimes were compared to rates under strictly phototrophic growth conditions. The present study provides the first quantitative measurements of grazing over a range of conditions, and shows that grazing can contribute up to 40% of the cellular N requirements h–1 for K. brevis. With evidence that blooms of Synechococcus can be enhanced due to anthropogenic nutrients, the potential importance of this particulate nutrient source for sustaining red tide blooms in situ is large and may help to resolve the current uncertainty as to how K. brevis blooms are maintained.
MATERIALS AND METHODS: Cultures and methods of growth. All cultures were obtained from the Provasoli-Guillard National Culture Collection maintained at the Bigelow Laboratory for Ocean Sciences, West Boothbay Harbor, Maine, USA. The Karenia brevis cultures included Clone CCMP2229, originally isolated from Manasota Key, Florida, and Clone CCMP2228, originally isolated off Sarasota, Florida. Synechococcus strain CCMP1768 was isolated from the Gulf of Mexico. The Karenia brevis cultures were grown in L1 medium at a salinity of 30 (Guillard & Hargraves 1993); the N source and prey concentration varied with the experiment, as described below. Maintenance cultures were grown with NO3– as the N source at a light intensity of 50 µmol photons m–2 s–1 on a 14 h light:10 h dark cycle. The seawater used in media preparation was from the Indian River inlet, Delaware, and was autoclaved prior to nutrient amendment. Cultures were maintained at 20°C in walk-in temperature-controlled rooms. Fluorescent fixtures were positioned to provide a range of light intensities, which varied by experiment. Cultures were not axenic, but aseptic techniques were used to minimize additional bacterial contamination during the growth periods.
Experimental design. Six experiments were conducted, including 2 designed to assess the rate of growth of Karenia brevis under varying conditions, and 4 designed to measure short-term rate processes.
Expt 1: Phototrophic growth of Karenia brevis with NO3–, NH4+ or urea at 3 irradiance levels: Clone CCMP2228 of K. brevis (initial inoculum: 1.3 × 104 cells ml–1) was grown in clean borosilicate test tubes, with 3 N forms (NO3 –, NH4 +, or urea; enriched at concentrations specified for N in L1 media) and at 3 irradiance levels (15, 30, or 115 µmol photons m–2 s–1).Growth was monitored approximately daily for 30 d, as described below (see ‘Analytical protocols’). Five replicates were maintained for each N source and light level. Growth rates were measured for the exponential period.
Expt 2: Mixotrophic growth of Karenia brevis vs. irradiance and Synechococcus availability: Cells of K. brevis Clone CCMP2228, grown on L1 medium (Guillard & Hargraves, 1993) but at the point of N depletion, were used for this experiment. These cells were transferred to 30 clean glass test tubes containing L1 medium minus N. The light exposures for this experiment were 43 and 300 µmol photons m–2 s–1,measured using a Biospherical Instruments QSL-100 with a 4ð probe inserted into the tubes. Initial culture densities of K. brevis were 450 cells ml–1 for the high light treatments and 520 cells ml–1 for the low light treatments. The ambient concentration of NO3– in the Indian River source water was 26 µmol N, and that of DON was 30 µmol N at the start of the experiment; other N substrates were negligible. The experiment was initiated with the addition of 5 different densities of Synechococcus (final concentrations of 0, 6.1 × 103, 2.1 × 104, 6.2 × 104, and 8.6 × 104 cells ml–1). Each light and treatment level was replicated in triplicate. To compensate for the different volumes of Synechococcus that were added (0.1 to 1.0 ml Synechococcus stock culture) and nutrients that would have been carried over with the transfer of Synechococcus, all cultures were brought to a total volume of 50 ml using filtrate from the Synechococcus cultures (filter pore size = 0.22 µm). Growth was assessed approximately daily for 25 d (see ‘Analytical protocols’).
Expt 3: Karenia brevis uptake of urea: Cultures of both clones in late exponential growth phase were used to assess the rate at which K. brevis directly took up dissolved urea. Culture aliquots, previously maintained on NO3–, were exposed over short periods (less than 1 h) to a concentration gradient of 15N-labeled urea, from 0.5 to 20.0 µmol N l–1.Cultures were not N-depleted at the time of urea exposure, but they were urea-depleted. Thus, all enrichment levels were ~100 atom percent (atom %). After incubation, the cultures were filtered onto pre-combusted GF/F filters (nominal pore size = 0.7 µm). The filters were dried and later analyzed by mass spectrometry (see ‘Analytical protocols’).
Expts 4 to 6: Mixotrophy in Karenia brevis. Short-term grazing rates with Synechococcus as prey: Three experiments were conducted using varying proportions of K. brevis and Synechococcus. Two experiments used both clones of K. brevis, while the final experiment used only K. brevis Clone CCMP2229. In each experiment, a monoculture of Synechococcus (late exponential growth phase) was labeled with 10 µmol l–1 15N-urea and incubated for 24 h to ensure uniform isotope labeling. The amount of isotope label in the Synechococcus culture inocula was determined by mass spectrometry. The K. brevis cultures (N-starved, in late exponential growth phase) were gently transferred to new 60 ml culture flasks and an inoculum of the labeled Synechococcus was added. Depending upon the experiment, 3 to 7 different densities of 15N-labeled Synechococcus were used. In all, the treatments yielded ratios of Synechococcus: K. brevis (cell:cell) of 0.36 to 225.8. The flasks were held in a constant temperature incubator for ~24 h on a 14 h light:10 h dark cycle. Following incubation, a 5 ml subsample was removed and preserved in glutaralde hyde (1 % final concentration). The remaining contents of the flasks were filtered through pre-combusted filters (GF/D, nominal pore size = 2.7 µm) that retained the K. brevis cells (~20 µm diameter) and possibly some of the Synechococcus cells (if attached to K. brevis). The filtrate was then re-filtered through a smaller pore pre-combusted filter (GF/F, nominal pore size = 0.7 µm) to retain Synechococcus cells that had not been collected with K. brevis on the first filter. The filters were rinsed with ultra-filtered Indian River water and dried for later analysis by mass spectrometry to determine the amount of 15Nlabel that was in the K. brevis cells versus the Synechococcus cells. Considering all 3 experiments collectively, 30% of the treatments of K. brevis (by strain) and Synechococcus (by density) were tested in duplicate. All 15Nsamples were processed using a Sercon mass spectrometer.
Analytical protocols. In Expts 1 and 2, growth of Karenia brevis and Synechococcus was tracked by changes in fluorescence and calibrated by microscopy and flow cytometry. On an approximately daily basis, 5 ml of each culture were removed, of which 1 ml was transferred to a cuvette and fluorescence emission at 682 nm (chlorophyll) and 560 nm (phycoerythrin) from 495 nm excitation were recorded on a Jobin Yvon Fluoro Max-3 spectrofluorometer. The chlorophyll fluorescence of K. brevis alone in the mixed species cultures was calculated as: chlK.b.= Em682 – 0.19Em560, where chlK.b.is the chlorophyll fluorescence signal from K. brevis, and Em682 and EM560 are the 682 and 560 fluorescence emission signals of the mixed culture. The emission at 560 nm was used as a measure of Synechococcus in these experiments. The factor, 0.19, is the average fluorescence emission ratio (682:560nm) from Synechococcus only. The remaining 4 ml of subsample was preserved in glutaraldehyde (1% final concentration). Of the 540 samples measured by fluorescence, 66 were analyzed by flow cytometry using a Becton Dickinson FACS-caliber flow cytometer to provide an independent measurement of Synechococcus density for cross-confirmation of the fluorescence patterns. Cell numbers were determined by gating by forward and side light scatter. The correlation between fluorometry and flow cytometry readings varied for cells grown at high and low light levels. For high light the correlation was: Fluorometry reading = 5.578x – 23867 (r2 = 0.87, n = 41) For low light, the correlation was: Fluorometry reading = 30.97x – 51935 (r2 = 0.69; n = 25) where x is cell number (cells ml–1) determined by flow cytometry. Selected aliquots were also examined by confocal microscopy to assess whether Synechococcus cells were actually retained within Karenia brevis cells. Autofluorescence of Synechococcus phycobilisomes was used to detect the cyanobacteria inside K. brevis cells. The preserved samples were filtered onto 25 mm GTBP black polycarbonate filters (Millipore, pore size = 3.0 µm). Each filter was placed on a microscope slide, covered by a drop of immersion oil (Cargille type FF) and a cover slip, and examined with a Leica TCS SP1 confocal microscope and a 63×, 1.2 numerical aperture water immersion lens. For these images, the excitation wavelength was 488 nm and the emission windows for K. brevis and Synechococcus pigments were 660 to 710 and 590 to 630 nm, respectively. Images were captured at consecutive intervals of 0.15 µm along the z-axis. Particulate and dissolved nutrients in the cultures were measured using standard methods. Particulate N in filtered samples (GF/D or GF/F) was determined using a Control Equipment CHN analyzer and dissolved nutrients were analyzed using autoanalysis (Lane et al. 2000), except for urea, which was measured using the protocol of Revilla et al. (2005).
Data analysis. Growth rates in Expts 1 and 2 were calculated for specific time intervals (days) within the exponential growth phase using the equation:r = ln(Nt/N0)/t (3) where r is specific growth rate (d–1),Nt is the measure of cells (or units of fluorescence) at time t, N0is that at time 0, and t is the interval of growth in days. The mean growth rates for the time interval when the highest observed rates were sustained (i.e. exponential growth phase) are reported. All growth rates are reported as the means of either 5 (Expt 1) or 3 replicates (Expt 2). For literature comparisons, growth rates were also expressed as divisions d–1 by dividing r by ln 2. Statistical comparisons between groups were based on 1-way ANOVAs; t-tests were used to compare differences within groups. All comparisons were made at a significance level of 0.01. The N-specific uptake rates of urea by Karenia brevis in Expt 3 were calculated as: V= atom % excess/(atom % enrichment × time) (4) where V is the N-specific uptake rate, h–1,atom % excess is the 15Nenrichment of the culture (minus natural 15Nbackground) at the end of the incubation period, atom % enrichment is the initial atom % of the substrate added, and time is the duration of the experiment in hours (e.g. Glibert & Capone 1993). To calculate the N-specific grazing rates of Synechococcus by Karenia brevis in Expts 4 to 6, a version of equation 4 was used. However, in this case, the atom% enrichment was that of the initial Synechococcus culture and the atom% excess was that of the K. brevis culture (GF/D fraction) after incubation. Since Synechococcus was growing during the period of the 24 h incubation, it was becoming isotopically diluted as the initial 15N-labeled urea had been consumed and the cells were using other regenerated forms of N. Hence, an isotope dilution correction was made to account for the exponentially changing enrichment of the Synechococcus cells (e.g. Glibert et al. 1982). To correct the atom % enrichment of the GF/D filters for any contribution by retained Synechococcus cells (or bacteria), the residual amount of Synechococcus on the GF/D filters was determined from the difference of mass of the cells retained on the GF/Fs and the initial culture. The 15Natom% enrichment due to the K. brevis contribution only was then determined according to the following equation:Atom % K. brevis = [(measured atom % GF/D)(massof K. brevis + mass of Synechococcus) – (atom % Synechococcus) (mass of Synechococcus)]/ (mass of K. brevis) (5) where all units of mass are in µmol-N. Grazing rates in terms of amount of N grazed per unit time (N ingestion rates) were calculated by multiplying the N-specific grazing rates by the particulate N (PN) content of the K. brevis culture:N ingestion rate = N-specific grazing rate × PN (6) Lastly, all grazing rates were also corrected for the direct uptake of urea by the cultures by subtracting the urea uptake rates determined above. Grazing rates were also calculated in terms of cells of Synechococcus grazed (Synechococcus K. brevis–1h–1)by converting the N ingestion rate to a cell-specific rate, based on a conversion of 1.65 fmol-N cell–1 of Synechococcus elongatis (Richardson 2004). Clearance rates (CR, µl K. brevis–1h–1)were calculated as described by Jeong et al. (2005) as follows: CR = IR/prey density (7) where IR is the cell-specific ingestion rate (cells Syne‑chococcus K. brevis–1h–1).
RESULTS : Growth experiments Expt 1 was designed to provide the benchmark phototrophic growth rates for one of the clones of Karenia brevis. For all 3 N substrates, growth was poor at the 2 lowest experimental irradiances (31 and 15 µmol photons m–2 s–1;Table 1). Growth at the highest irradiance on all N substrates was significantly higher than growth at the lower irradiance levels. Growth rates on NH4 + and on NO3– at the highest irradiance (115 µmol photons m–2 s–1)were virtually identical at 0.35 to 0.37 d–1 (Table 1), while the average rate of growth on urea at the same irradiance was slightly less, at 0.28 d–1 (Table 1), though not statistically different from rates observed using NH4 + or NO3– as substrates. Expt 2 explored the potential for growth of Karenia brevis on a particulate N source (Synechococcus). In this experiment, the only sources of N were the Synechococcus cells, the slight quantity of inorganic N that was transferred with the original inoculum, any dissolved N that may have been released by the cultures during the growth cycle, and contaminating bacteria. Exponential growth rates for those cultures maintained at the higher irradiance in this experiment (300 µmol photons m–2 s–1)were very similar to those obtained in Expt 1, ranging from 0.26 to 0.35 d–1 (Fig. 1). In contrast, a much wider range of exponential growth rates was observed for those cultures grown at the lower irradiance (43 µmol photons m–2 s–1),0.14 to 0.58 d–1,depending on the Synechococcus enrichment level (Fig. 1). For the low light-grown culture, the highest exponential growth rates corresponded with those of K. brevis cultures that had the highest level of Synechococcus added. At both the highest and 2 lowest levels of Synechococcus enrichment, mean exponential growth rates of K. brevis under the 2 light regimes were significantly different (Fig. 1). The time interval over which the exponential growth was observed also varied in Expt 2, depending on the irradiance and Synechococcus enrichment levels. At the higher irradiance level, all cultures grew rapidly for approximately 10 d, after which the culture growth began to decline (Fig. 2A, C, E, Table 2). There were no significant differences in the average exponential rates of growth, but the absolute cell number attained was related to the amount of Synechococcus added; the lowest Synechococcus level yielded the highest final biomass of Karenia brevis (Fig. 2A). At the lower irradiance, a different pattern was observed. In this case, the culture with the highest Synechococcus concentration initially grew fastest, then growth of the culture with the next highest Synechococcus level increased a few days later, and again a similar increase occurred a few days later with the third highest Synechococcus level, and so on (Fig. 2B, D, F). In contrast to the high light treatments, the maximum biomass of K. brevis attained varied much less with amount of Synechococcus added (e.g. Fig. 2D, F). The exponential rates of growth of Synechococcus in the mixed cultures differed considerably. At the higher irradiance level, all growth rates of Synechococcus were negative (Fig. 3A, C, E). At the lower light intensity, the highest observed growth rates (i.e. exponential growth phase) of all cultures ranged from 0.23 to 0.49 d–1,although the time interval over which the highest growth rate was attained depended on the initial inoculum level (Fig. 3B, D, E, Table 3). Rates of growth of Karenia brevis and Synechococcus were inversely related in the low light treatment: the highest growth rate of K. brevis corresponded with the lowest observed growth rate of Synechococcus.
Urea uptake rates The goal of Expt 3 was to obtain direct rates of urea uptake by Karenia brevis and to use these data to correct the 15Ngrazing experiments for 15Nurea uptake, i.e. as dissolved substrate controls and as controls for any bacterial uptake (urea was the initial 15N-substrate initially used to label the Synechococcus cells). Because these K. brevis cells were not acclimated to growth on urea, this experiment can be viewed asyielding baseline or minimal rates. The rate of urea uptake would be expected to be higher after acclimation when the enzyme systems were optimized for maximal urea uptake. Under these experimental conditions, the N-specific uptake rates (V) of urea for both clones were very low (2 × 10–5 to 1.4 × 10–4 h–1;Fig. 5). Although the data could be fit to a Michaelis-Menten formulation, a linear fit to the data was better.
Grazing rates Expts 4 to 6 were designed to assess short-term grazing rates by Karenia brevis on Synechococcus. N-specific rates of grazing on Synechococcus by K. brevis ranged from 9.28 × 10–4 to 1.22 × 10–2 h–1 and varied with the relative proportion of Synechococcus:K. brevis (Table 4). These rates also represent a range of 0.026 to 2.15 pmol N K. brevis–1d–1,or 0.96 to 83.8 Synechococcus K. brevis–1h–1 (Table 4, Fig. 6). Both clones of K. brevis had comparable rates of N-specific grazing. Expts 4 to 6 used very different cell density ranges of K. brevis and Synechococcus. When the data from all 3 experiments were considered, it was found that as more Synechococcus prey were added relative to K. brevis, the grazing rate of K. brevis (expressed as ranges of K. brevis and Synechococcus. When the data from all 3 experiments were considered, it was found that as more Synechococcus prey were added relative to K. brevis, the grazing rate of K. brevis (expressed as investigated. Shanley & Vargo (1993), in studies of the Wilson clone (reported as Gymnodinium breve), reported increasing rates of growth with increasing irradiance. The highest growth rates of 0.53 divisions d–1 (0.36 d–1)were found when cells were acclimated to the highest experimental growth irradiance of 90 µmol photons m–2 s–1.Maier Brown et al. (2006) also reported growth rates of several K. brevis clones including a Mexico beach clone, the Wilson clone, and a Charlotte Harbor clone; of these, the growth rates of the Charlotte Harbor clone were 0.25 to 0.30 d–1 for salinities above 20, but rates were considerably lower for the other 2 clones. In the present study, growth rates of cultures maintained on Synechococcus (Fig. 1) were in the same range as those grown phototrophically when the cells were maintained at the higher experimental irradiance. At the higher irradiance, growth of Karenia brevis on Synechococcus was not sustained beyond ~10 d (Fig. 2). After all of the Synechococcus cells or other sources of nutrients apparently were consumed, growth of K. brevis declined. However, at the lower irradiance level, maximum growth rates of K. brevis varied considerably. At the 2 lowest concentrations of Synechococcus cells, growth rates were about half of those that were obtained under either phototrophic conditions or in the higher-irradiance mixotrophic conditions. At the highest density of Synechococcus cells in low light, the growth rate was significantly enhanced, up to 0.58 d–1.This growth rate was ~60% higher than the rate at the next highest Synechococcus treatment level. Given the fact that growth rates increased linearly with the Synechococcus levels added and showed no evidence of growth rate saturation, it is possible that even higher growth rates are attainable with higher densities of Synechococcus. In the present study, the potential for heterotrophic bacteria to also contribute to the nutrition of Karenia brevis cannot be discounted (e.g. Seong et al. 2006), nor can the use of dissolved substrates from disruption of the Synechococcus or K. brevis cells. If bacteria were being grazed in these experiments, it would only underscore that mixotrophy can help to sustain growth. However, the observed presence of Synechococcus in the K. brevis cells based on confocal microscopy (Fig. 4) confirms that Synechococcus cells were indeed grazed.
Urea uptake rates Karenia brevis has previously been found to use urea (e.g. Bronk et al. 2004, Sinclair 2008), but the rates of urea uptake for our experiments were very low (Fig. 5). In fact, they are roughly 2 orders of magnitude lower than rates reported by Bronk et al. (2004) and Heil et al. (2007) for natural Florida blooms, but are comparable to rates reported by Sinclair (2008) for cultures grown under low light. Uptake rates of urea by K. brevis from the southwestern Florida shelf, reported by Heil et al. (2007), were comparable to N-specific uptake rates of urea for a related Karenia species, K. mikimotoi, during a bloom off the East China Sea (Li et al. 2009). The higher rates in the field relative to those determined in the present study are likely a consequence of up-regulation of urea metabolism in the field populations when other sources of N were limiting. As noted above, our laboratory experiments were conducted under low urea concentrations, but abundant NO3_. The rates of urea uptake reported here were about 100-fold lower than the grazing rates when compared on the basis of N taken up. Although our reported rates of urea uptake are low compared to other reports of in situ rates when cells were preconditioned with urea, urea uptake and grazing rates may be more comparable under natural conditions. than rates reported by Bronk et al. (2004) and Heil et al. (2007) for natural Florida blooms, but are comparable to rates reported by Sinclair (2008) for cultures grown under low light. Uptake rates of urea by K. brevis from the southwestern Florida shelf, reported by Heil et al. (2007), were comparable to N-specific uptake rates of urea for a related Karenia species, K. mikimotoi, during a bloom off the East China Sea (Li et al. 2009). The higher rates in the field relative to those determined in the present study are likely a consequence of up-regulation of urea metabolism in the field populations when other sources of N were limiting. As noted above, our laboratory experiments were conducted under low urea concentrations, but abundant NO3_. The rates of urea uptake reported here were about 100-fold lower than the grazing rates when compared on the basis of N taken up. Although our reported rates of urea uptake are low compared to other reports of in situ rates when cells were preconditioned with urea, urea uptake and grazing rates may be more comparable under natural conditions. brevis was not reported specifically. The confocal microscopy results also suggest that ingestion of Synechococcus is on the order of a few cells at a time, as no image yielded more than a few fluorescently visible cells inside the K. brevis cells. The N-specific grazing rates for the highest prey: predator ratio were comparable to the specific growth rates reported here when both rates were normalized on a daily basis, suggesting that this source of nutrition can be important under specific environmental conditions. Such a conclusion can be verified using mass balance calculations. The cellular N quota of Synechococcus has been reported to range from 1.65 fmolN cell–1 (Richardson 2004) to 3.5 fmol-N cell–1 (Kana & Glibert 1987) and that of Karenia brevis has been estimated to be 5.4 pmol-N cell–1 (Shanley & Vargo 1993, Sinclair et al. 2006, present study). Thus, it would require roughly 1.5 × 103 to 3.3 × 103 Synechococcus cells to double the biomass of K. brevis if it is assumed (for the basis of mass balance calculations only) that all N were retained for growth. At a K. brevis growth rate of 0.3 d–1 (0.43 divisions d–1),the cellular N quota would be obtained from ingestion of 20 to 40 Synechococcus cells h–1.At a growth rate of 0.58 d–1 (0.84 divisions d–1),as reported here for low-light growth on Synechococcus (Table 2), the cellular N quota would be obtained from ingestion of 50 to 100 Synechococcus K. brevis–1h–1.The rates reported here are in this general range, 0.96 to 83.8 Synechococcus K. brevis–1h–1 (Table 4), indicating that grazing on picoplankton can potentially account for a major fraction of the N demand for growth of K. brevis.
Grazing rates Grazing by Karenia brevis on Synechococcus was previously reported by Jeong et al. (2005), who compared the ability of 18 red tide species to graze Synechoccocus. They reported that, for an initial Synechococcus concentration of 1.25 × 106 cells ml–1,the ingestion rate was 5 Synechococcus K. brevis–1h–1.Our results (0.96 to 83.8 Synechococcus K. brevis–1h–1),using different methodology, compare favorably with, and fall either side of, the values of Jeong et al. (2005). Grazing by K. brevis on Synechococcus has also been observed elsewhere (L. A. Procise & M. M Mulholland unpubl. data), but at rates apparently much lower than those of Jeong et al. (2005) or the present study. The clearance rates reported herein (Table 4) also compare favorably with the range reported by Jeong et al. (2005) for several harmful algal species, although a clearance rate for K.Brevis. No significant differences were found between the 2 Karenia brevis clones studied here. Intraspecific differences in various traits of many clones of dinoflagellates, including N uptake rates by K. brevis, have been observed (e.g. Burkholder & Glibert 2006, Sinclair 2008), but in this case, the ability to feed may be a characteristic of both clones. The clone used by Jeong et al. (2005) was different from the clones used in the present study, further suggesting that feeding may be a common trait of K. brevis. Clonal differences in the growth and N nutrition of Synechococcus spp. are well known. Some clones arrest their cell division upon N limitation, while others continued to divide at slower rates during N depletion (Glibert et al. 1986). Clonal differences in the ability of Synechococcus to take up NH4 + and NO3–have also been shown (Glibert & Ray 1990), but very little is known about the clonal differences in their ability to grow on urea. The extent to which growth and nutrient uptake rates of the Florida clone varies with N growth status is not well understood (but see Richardson 2004), but will affect its ability to maintain populations that may be available as food for Karenia brevis in situ.
Implications for natural blooms The sources of N supporting Karenia brevis have long been in debate for the southwestern Florida shelf and, more generally, the Gulf of Mexico. Blooms have been hypothesized to initiate in response to organic N that becomes available following outbreaks of the cyanobacterium Trichodesmium (e.g. Walsh & Steidinger 2001, Mulholland et al. 2006). Trichodesmium may release a significant fraction of its newly fixed N in the form of dissolved organic nitrogen (DON) (Glibert & Bronk 1994), and estimates of this contribution in natural blooms suggest that DON from Trichodesmi um may be sufficient to support moderately dense (≤105 cells l–1)K. brevis blooms (Mulholland et al. 2006). Other sources of nutrients that may sustain moderate to dense blooms (>105 cells l–1)have been proposed. Vargo et al. (2004) estimated that estuarine (land-based) sources of N and P could help to sustain moderate nearshore blooms of K. brevis. Vargo et al. (2008) evaluated potential dissolved nutrient sources and fluxes and suggested that in addition to estuarine inputs, sources such as release of N from dead and decaying fish, atmospheric deposition, benthic flux, and release from zooplankton may all be important, but no single factor seemed to be sufficient to sustain blooms.
Sinclair et al. (2006) proposed another mechanism for obtaining nutrients. They found that Karenia brevis cells can migrate into the sediments on a daily basis and potentially acquire NO3– originating from the pore waters. They reported N-specific NO3 – uptake rates that were not only higher than the urea uptake rates reported herein, but also were higher than the field measurements reported by Heil et al. (2007), ranging from 0.2 h–1 at a 0.5 µmol N–1 addition to 0.52 h–1 at an addition of 11 µmol N–1.The NO3– uptake rates in the Sinclair et al. (2006) study were also found to vary as a function of time of day and extent to which the cells had been previously exposed to or conditioned on NO3–. Ambient concentrations of NO3– in the water column off the southwestern Florida shelf are typically low (<1 µM N; Heil et al. 2007), and Sinclair et al. (2006) found that those cells that were comparatively NO3 –-depleted for the 12 h prior to descending in the sediments had a higher rate of NO3– uptake.
With the specific rates of N-specific grazing reported here, we can add grazing as a potentially important nutritional strategy of Karenia brevis. The growth experiments demonstrated that in the presence oflimiting dissolved N, K. brevis could grow and survive for a period of at least days to weeks. Indirect evidence that grazing on Synechococcus may be important during blooms comes from a study of the southwestern Florida shelf in 2003 when the phytoplankton community varied in composition along the north–south gradient (Heil et al. 2007). In that study, phytoplankton composition was tracked using pigment signatures, including gyroxanthin diester as an indicator of K. brevis (e.g. Millie et al. 1997) and zeaxanthin as an indicator of Synechococcus (e.g. Kana et al. 1988). The region where K. brevis was detected, off the Peace River, was the region where zeaxanthin was virtually absent despite its presence in surrounding waters (Heil et al. 2007), implying that Synechococcus was controlled by K.brevis grazing. In the same study, urea was found to be significantly associated with the abundance of Synechococcus. It was suggested therein that as anthropogenic sources of N from urea applications in agriculture increase, one possible outcome will be increases in cyanobacterial blooms (Heil et al. 2007).
It can now be hypothesized that as cyanobacterial blooms increase, so too does the potential for Karenia brevis growth to be enhanced and for blooms to be sustained through grazing, especially under the low light conditions associated with bloom self-shading. Recognition of this pathway is at least one step toward reconciling the long-term reported increase in K. brevis blooms (e.g. Brand & Compton 2007) and the tendency for blooms of this species to develop offshore in seemingly oligotrophic waters (e.g. Vargo et al. 2004, 2008). Many important questions remain, including whether the rates reported in the laboratory are comparable to those in the field, whether all clones of Synechococcus are grazed similarly or whether all clones of K. brevis have comparable grazing behavior, whether other algae or bacteria are also grazed and at what rate, how grazing rates may change on a diel basis or with physiological status, and whether toxicity of K. brevis varies with nutrient uptake and feeding strategies.
Grazing by Karenia brevis on Synechococcus enhances its growth rate and may help to sustain blooms
Patricia M. Glibert1,*, JoAnn M. Burkholder2, 1University of Maryland Center for Environmental Science, Horn Point Laboratory, PO Box 775, Cambridge, Maryland 21613, USA2Center for Applied Aquatic Ecology, North Carolina State University, Raleigh, North Carolina 27695, USA