- Split View
-
Views
-
Cite
Cite
David Aldridge, Duncan A. Purdie, Mikhail V. Zubkov, Growth and survival of Neoceratium hexacanthum and Neoceratium candelabrum under simulated nutrient-depleted conditions, Journal of Plankton Research, Volume 36, Issue 2, March/April 2014, Pages 439–449, https://doi.org/10.1093/plankt/fbt098
- Share Icon Share
Abstract
The dinoflagellate Neoceratium is commonly observed in oceanic waters, depleted in major inorganic nutrients such as nitrogen and phosphorus. Using culture isolates, we investigated whether two Neoceratium species (N. hexacanthum and N. candelabrum) can grow phototrophically at low nutrient concentrations found in surface waters of oligotrophic subtropical gyres (OSGs). No phototrophic growth (indicated by changes in cell numbers, the presence of dividing cells or cellular protein increase) was observed when N. hexacanthum and N. candelabrum were grown in low nutrient seawater. In separate experiments, to determine survival time under oligotrophic nutrient conditions, 68% of N. hexacanthum cells were able to re-establish growth after spending 1–10 days in North Atlantic gyre seawater; 40% recovered after 11–20 days and only 3% recovered after 21–30 days. The longest period any single cell survived, and then went on to divide, was 26 days. These findings demonstrate that Neoceratium cells could remain viable for >3 weeks in surface waters of OSGs, but to sustain their growth nutrients must be obtained periodically from an alternative source: via ephemeral upwelling of nutrient-rich waters, phagotrophy and/or movement to and from the nutricline.
INTRODUCTION
In oligotrophic subtropical gyres (OSGs), globally important ecosystems that cover over 30% of the surface of the Earth, primary productivity is limited by low standing stocks of nutrients, usually nitrogen (Moore et al., 2013). The low nutrient concentrations found here are predominantly sustained via microbial regeneration in the surface mixed layer, rather than by inputs from below the nutricline (Karl, 2002). Organisms that have low sinking velocities and high nutrient uptake rates are better able to cope with low nutrient conditions, compared with larger cells (Fuhrman et al., 1989; Zubkov et al., 2007), which explains why these environments are typically dominated by prokaryotes and small eukaryotic algae (Cho and Azam, 1988; Zubkov et al., 2000; Zubkov and Tarran, 2008; Hartmann et al., 2012).
Dinoflagellates in the genus Neoceratium (cell volume: 49 000–331 000 µm3 cell−1, Takahashi and Bienfang, 1983) are commonly found, although not dominant, in OSGs (Graham 1941; Graham and Bronikovsky 1944; Weiler 1980). In the North Pacific central gyre, for example, where Neoceratium is estimated to contribute <1% of total phytoplankton carbon, concentrations range from 0.8 to 3 cells L−1 (Weiler, 1980; Matrai, 1986). Although cell concentrations are significantly lower in the central gyre than in coastal waters of the Pacific Ocean (typically 4–14%) (Weiler, 1980; Matrai, 1986), growth rates (based on the frequency of dividing cells) are relatively high based on one in situ study in the North Pacific subtropical gyre: 26–38% of the maximum rates observed for this genus (Weiler, 1980). This is curious, as large cells, with low surface area to volume ratios, should be at a disadvantage due to their relatively low nutrient affinities (Aksnes and Egge, 1991).
The presence of Neoceratium in OSGs may reflect an ability of cells to optimally position themselves in the water column for nutrients and light (Weiler and Karl, 1979; Heaney and Furnass, 1980; Taylor et al., 1988; Baek et al., 2009), or supplement their nutrient demand via phagotrophy (Bockstahler and Coats, 1993; Li et al., 1996; Smalley et al., 1999, 2003, 2012; Smalley and Coats, 2002). However, it is also possible that Neoceratium do not require an alternative source of nutrients: cells may be able to survive and grow at the low nutrient concentrations typically found in OSGs. Physiological adaptations to low nutrient concentrations may allow cells to survive and grow in nutrient impoverished surface waters of OSGs. For example, low half saturation constants for nitrogen and phosphorus and specific characteristics for nutrient uptake, such as luxury consumption, have been suggested to give Neoceratium an advantage over other algal species growing at low nutrient concentrations (Baek, et al., 2008), and the production of the enzyme alkaline phosphatase may allow for the exploitation of dissolved organic phosphorus pools (Mackey et al., 2012; Girault et al., 2013). These adaptations may account for the previous observation of Neoceratium fusus growing in low nutrient culture medium (Baek et al., 2007; NO3− ≤1.0 µM and PO43− ≤0.1 µM).
Here, we use cultures of two Neoceratium spp. (N. hexacanthum, and N. candelabrum) commonly found in OSGs, to determine if (and for how long) cells can survive and grow at nutrient concentrations typically found in surface waters of these environments. Cell growth, under these conditions, was investigated by monitoring changes in cell numbers, dividing cells, protein per cell and Fv/Fm in low nutrient seawater (LNSW) compared with nutrient replete seawater (RSW). Changes in other cell growth properties (survival time, viability, time-lag of growth response and growth rate) were determined by monitoring cells in embryo dishes, exposed to North Atlantic gyre seawater (NAGSW) for varying periods of time before addition of nutrients.
METHOD
Maintenance of Neoceratium cultures
Neoceratium strains (N. hexacanthum, strain number P10B2; N. candelabrum, strain number P37C2) were obtained from the Gulf of Villefranche (Point B, 0–80 m), in The Mediterranean Sea between 2007 and 2008. They were cultured in K/5 medium (Keller et al., 1987), minus silicate, made from NAGSW (Table I), collected from the surface of the central gyre in October 2009 during Atlantic Meridional Transect (AMT) 19, onboard the Royal Research Ship (R.R.S.) James Cook. All cultures were maintained in 100 mL borosilicate Erlenmeyer flasks and kept at 18°C in a temperature controlled incubator with a 12:12 h light/dark photocycle, with a photon flux density of 60 μmol quanta m−2 s−1.
Experimental seawater . | Nitrate concentration (μM) . | Phosphate concentration (μM) . |
---|---|---|
Low nutrient seawater (LNSW) | 0.58 | 0.04 |
Nutrient replete seawater (RSW) | 185 | 1.40 |
Surface North Atlantic gyre seawater (NAGSW) | <0.02a | <0.02a |
Experimental seawater . | Nitrate concentration (μM) . | Phosphate concentration (μM) . |
---|---|---|
Low nutrient seawater (LNSW) | 0.58 | 0.04 |
Nutrient replete seawater (RSW) | 185 | 1.40 |
Surface North Atlantic gyre seawater (NAGSW) | <0.02a | <0.02a |
All experimental seawater was made from NAGSW; the only amendments made were to concentrations of nitrate and phosphate.
aMeasurements that were below the detection limit of the nutrient analyser.
Experimental seawater . | Nitrate concentration (μM) . | Phosphate concentration (μM) . |
---|---|---|
Low nutrient seawater (LNSW) | 0.58 | 0.04 |
Nutrient replete seawater (RSW) | 185 | 1.40 |
Surface North Atlantic gyre seawater (NAGSW) | <0.02a | <0.02a |
Experimental seawater . | Nitrate concentration (μM) . | Phosphate concentration (μM) . |
---|---|---|
Low nutrient seawater (LNSW) | 0.58 | 0.04 |
Nutrient replete seawater (RSW) | 185 | 1.40 |
Surface North Atlantic gyre seawater (NAGSW) | <0.02a | <0.02a |
All experimental seawater was made from NAGSW; the only amendments made were to concentrations of nitrate and phosphate.
aMeasurements that were below the detection limit of the nutrient analyser.
Isolation into experimental flasks
Prior to all experiments, Neoceratium were isolated from their culture medium into experimental medium using the following method. The concentration of cells in the stock culture (late exponential/stationary growth phase) was determined in order to calculate the volume required to achieve desired experimental concentrations (∼10–30 cells mL−1). Prior to transfer to experimental Erlenmeyer flasks (acid washed with HCL and autoclaved), this predetermined volume of stock culture was collected on a 20 μm nylon net filter (NY20; Millipore) and rinsed with filtered NAGSW, using a pipette, in order to remove excess nutrients. Cells were rinsed from the filter into one of the experimental flasks, and then evenly divided (by volume) between all the flasks. All experiments were performed in triplicate.
Experiments were conducted in 100 mL Erlenmeyer flasks, using 50 mL of LNSW or RSW (Table I). Experiments were incubated at 18°C on a 12:12 h light/dark cycle at a photon flux density of 60 μmol quanta m−2 s−1. Every 4 days (starting on day 1) sub-samples from each experimental treatment were taken, from which: cell numbers, number of dividing cells, estimated protein per cell, photosynthetic efficiency of photosystem II (PSII; Fv/Fm) and dissolved inorganic nutrient concentrations (NO3− + NO2−, and PO43−; only for the N. candelabrum experiment) were determined.
Cell numbers
Cell numbers were determined by enumeration of 1 mL of experimental medium in a 1 mL plastic Sedgewick–Rafter chamber, under an inverted microscope. In addition to cell numbers, observations on motility were made and the number of dividing cells was recorded.
Protein content of cells
1.5 mL of experimental medium was frozen (−80°C) in 1.5 mL cryovials for analysis at a later date. The average protein content of Neoceratium cells was measured using the bicinchoninic acid (BCA) method (Smith et al., 1985), using bovine serum albumin (fraction 5) as a standard. BCA reagents were added in an 8:1 ratio (reagents: sample), and samples were left at room temperature for 24 h, prior to analysis on a Nanodrop 1000 Spectrophotometer. Data values more than two standard deviations from the mean were omitted from the final analysis (N. hexacanthum: 1 of 54 samples; N. candelabrum: 2 of 54 samples).
Fast repetition rate fluorometry (FRRf)
Fast repetition rate fluorometry (FRRf) was used to assess the maximum PSII photochemical efficiency (Fv/Fm), which typically decreases under periods of ‘stressful’ growth such as nutrient starvation (Kolber and Falkowski, 1993; Kolber et al. 1998). Two millilitres of each experimental medium were transferred to a 15 mL darkened Falcon tube for at least 1 h prior to FRRf analysis, allowing the primary receptor molecules of PSII (quinones) to be fully oxidized and ready to receive electrons (Suggett et al., 2009). FRRf measurements were made using the ‘standard’ protocols described by Suggett et al. (Suggett et al., 2003) and Moore et al. (Moore et al., 2005). A Chelsea Instruments Fasttracka FRR fluorometer was programmed to deliver single photochemical turnover saturation of PSII from 100 flashlets of 1.1μs at 1 μs intervals. Fluorescence transients generated were then fitted to the Kolber–Prasil–Falkowski (Kolber et al., 1998) model to yield values of Fv/Fm (Moore et al. 2006).
Nutrient measurements
Dissolved inorganic nutrients (nitrate + nitrite and phosphate) were analysed from subsamples taken from LNSW experimental flasks (N. candelabrum experiments only). Ten-millilitre subsamples were filtered through 0.2 μm Millipore filters into 15 mL Falcon tubes; subsamples were the result of combining 3.5 mL subsamples from triplicate experimental flasks; this was done in order to provide sufficient volume for analysis. Samples were stored at −80°C prior to analysis within 1 month of collection. Nitrate and phosphate measurements were analysed using a Skalar SanPlus Autoanalyser according to Sanders and Jickells (Sanders and Jickells, 2000).
Changes in cell growth properties over 30 days of exposure to NAGSW
Neoceratium hexacanthum cells (exponential growth phase) were reverse filtered in order to remove excess nutrients from the culture medium. Following this, approximately 10 cells (average: 9; range: 5–14) were transferred to 30 embryo dishes containing 2.5 mL of NAGSW, and covered with a glass lid to prevent evaporation (the mass of each dish was monitored to ensure no evaporation occurred). Exact cell numbers were confirmed, and every day for 30 days (starting on day 1) nutrients were added (0.5 mL of K-medium) to a single dish. Following nutrient addition cell numbers and dividing cells were monitored every 2 days, allowing for the estimation of the following.
Maximum viability time
The longest period that any single cell survived and went on to divide.
Changes in growth rate
Growth rate was estimated, after initial time-lags in growth response following nutrient addition, by fitting an exponential growth model [Stirling model; y = y0+ a(exp(b × x)−1)/b] to the data using SigmaPlot (Systat Software Inc.) software where ‘x’ refers to x-axis values (experiment day number), ‘y’ refers to y-axis values (cell number), and ‘a’ and ‘b’ are coefficients provided by the software. The slope of the line was then used to estimate growth rate in divisions per day (day−1). Several different exponential models were tested for how well they fitted the data. The Sterling model was used as it had the highest R2 values.
Change in the percentage of viable cells (cells capable of growth)
For each embryo dish, on the day of nutrient addition, we attempted to calculate the number of viable cells present that contributed to the increase in cell numbers. To do this, we removed the first two data points from each of the growth curves and then fitted the Stirling growth model (discussed above) to the data in order to estimate how many cells were viable on day 1 (y = 0). We decided to remove the first two data points as this was found to provide the best compromise between removing the influence of early data points and reducing the error of the estimate of y = 0. For dishes where no growth was observed, the number of viable cells was taken to be zero. By combining data together into three groups (days 1–10, days 11–20 and days 21–30), and comparing the number of viable cells from each group to the total number of cells upon nutrient addition, it was possible to estimate the percentage of viable cells during each of these three periods.
Changes in time-lag of growth response
The growth time-lag for each dish was estimated from the number of days it took, after nutrient addition, for the first sign of growth to occur: either an increase in cell number, or the presence of dividing cells.
RESULTS
Growth at nutrient concentrations typical of OSGs: N. hexacanthum
In experiments on N. hexacanthum, in LNSW, cell numbers showed little variation over 21 days (Fig. 1A: range, 17.7 to 20 cells mL−1). Cell numbers increased in RSW up until day 13, from 28 to 49 cells mL−1; after this point cell numbers decreased from 49 to 33 cells mL−1. The percentage of dividing cells observed broadly mirrored the above pattern (Fig. 1C): no dividing cells were observed in LNSW; in RSW dividing cells increased from 0 to 4.8% between days 1 and 9, before decreasing to 0.7% by day 17. Motile cells were observed throughout the 21-day experiment in RSW, but no motile cells were observed after day 17 in LNSW.
Average protein content per cell (Fig. 1E) remained stable in LNSW between days 1 and 13 (8.4–9.1 ng cell−1), before decreasing to 4.3 ng cell−1 by day 21. In RSW, average protein per cell fluctuated between 6.9 and 11.3 ng cell−1, but was similar on days 1 and 21 (8.7 and 9.9 ng cell−1 respectively).
There was little difference in Fv/Fm values between the two nutrient treatments, where a small decrease occurred over the 21 days: from 0.274 to 0.245 in LNSW, and from 0.307 to 0.235 in RSW (Fig. 1G).
Growth at nutrient concentrations typical of OSGs: N. candelabrum
In experiments on N. candelabrum, in LNSW, cell numbers (Fig. 1B) showed a small decrease over 21 days, from 21 to 15 cells mL−1, whereas cell numbers increased in RSW over the course of the whole 21-day period, from 16 to 156 cells mL−1. The percentage of dividing cells (Fig. 1D) mirrored the above pattern: the percentage of dividing cells on day 1 was roughly similar across both scenarios (9 and 12%). Beyond day 5, no dividing cells were observed in LNSW; in RSW, there was an increase in dividing cells, between days 1 and 5, from 12 to 19%, followed by a decrease to 2% by day 21. Motile cells were observed throughout the experiment under both nutrient scenarios.
The average protein content per cell (Fig. 1F) remained relatively stable in both nutrient scenarios between days 1 and 21, with some indication of a downward trend over time: in the LNSW scenario protein concentrations were 5.5 ng cell−1 on day 1 and 2.7 ng cell−1 on day 21; in the RSW scenario, protein concentrations were 6.4 ng cell−1 on day 1 and 4.0 ng cell−1 on day 21.
Fv/Fm values (Fig. 1H) closely mirrored changes in cell numbers (Fig. 1B). In LNSW values stayed relatively constant up until day 9 (range, 0.089–0.116); this was followed by a decrease, to 0.014, by day 21. In RSW, there was a rapid increase between days 1 and 9 (from 0.107 to 0.325), followed by a gradual increase until day 17 (to 0.394), before falling to 0.319 by day 21.
In LNSW, PO43− concentrations remained between the detection limit of <0.02 and 0.04 μM. NO3− concentrations ranged between 0.41 and 0.86 μM.
Changes in cell growth properties over 30 days of exposure to NAGSW
In almost all embryo dishes, cell numbers either remained the same or decreased before nutrients were added (Fig. 2). The longest time any single cell survived in NAGSW, and then went on to re-grow after addition of nutrients, was 26 days (Fig. 2). Motile cells were observed after 28 days, but none of these cells was able to re-establish growth when nutrients were added.
The percentage of viable cells decreased over the course of the experiment (Fig. 3A): between 1 and 10 days, the percentage of viable cells was 68% (71/105 cells); this dropped to 40% between days 11 and 20 (44/110 cells), and 3% (2/64 cells) between days 21 and 30.
The time-lag between nutrient addition and the first sign of growth appearing generally increased over the course of the experiment (Fig. 3B): the time-lag was 2 days up until day 10; from day 11 onwards, with the exception of day 13, the time-lag was always greater than 2 days (between 4 and 8 days).
Once growth did commence, the growth rate of cells appeared to be unaffected by the length of exposure to NAGSW (Fig. 3C). The mean growth rate (once time-lag had been accounted for) observed was 0.16 day−1 (range: 0.07–0.31 day−1), and for most embryo dishes observed (12 out of 17) the growth rate was within a fairly narrow range of 0.11–0.19 day−1. Although the lowest growth rate (0.07 day−1) was observed on day 26, the growth rate just 1 day before (day 25) was close to the mean growth rate observed (0.15 day−1).
DISCUSSION
Our observations, on N. hexacanthum and N. candelabrum, revealed that ontogenetic growth (cell division) did not occur at the low nutrient concentrations used here (starting concentrations: NO3−, 0.58 μM; PO43−, 0.04 μM), and cells became more nutrient stressed with time, as shown by decreasing Fv/Fm values. Fv/Fm worked differently as an indicator of ‘nutrient stress’ in both species; the reduction in Fv.Fm was far more pronounced in N. candelabrum. We do not have an explanation for why this occurred, but considering all the parameters that were measured it appears that the ‘health’ of both species did suffer under low nutrient concentrations. In addition to the absence of ontogenetic growth, we can eliminate the possibility that cells were slowly increasing in size (somatic growth), and therefore potentially dividing on timescales longer than the experimental period (21 days), as average protein content per cell roughly halved over the course of the experiment in both species. This suggests that Neoceratium could not survive in OSGs if they were entirely dependent on the nutrient concentrations found in surface waters (above the nutricline), much less achieve the high growth rates (0.09–0.13 day−1) that have been previously observed in the North Pacific subtropical gyre (Weiler, 1980).
N:P ratios were higher in the RSW than in LNSW (132:1 compared with 15:1), which may have enhanced the growth rate in the RSW treatment; Neoceratium has previously been shown to grow faster at higher N:P ratios (Baek et al., 2008). However, these differences in ratios would not account for why there was no growth at all in the LNSW, as there is no evidence that low N:P ratios prevent Neoceratium from growing (Baek et al., 2008). Increases in swimming speed, sinking rate and turbulent shear can all reduce nutrient diffusion limitation in larger cells to some extent (Chisholm, 1992). In our experimental set-up using culture flasks, sinking rate and turbulent shear would have been negligible compared with the natural environment. However, it has been pointed out that these two factors could, at most, scarcely double nutrient uptake rates in a cell with a diameter of 200 μm (Chisholm, 1992). Therefore, it seems unlikely that this factor would account for growth of Neoceratium in OSG surface waters. We cannot rule out the possibility that, in the natural environment, Neoceratium exploit nutrient micropatches in OSG surface waters (Lehman and Scavia, 1982), where Neoceratium cells are likely to be able to take up nutrients at the same rate as smaller algal cells based on recent findings that phytoplankton uptake rates are not dependent on cell size (Marañón et al., 2013). The uptake of dissolved organic nutrients may also provide an additional source of phosphorus and nitrogen to Neoceratium in OSGs. As these fractions are thought to turn over on time periods of minutes to days (Bronk et al., 2007), it is unlikely that the concentrations of these nutrients, in our experimental media, were representative of concentrations found in the natural environment. It is known that Neoceratium produce the enzyme alkaline phosphatase, which may allow for the exploitation of dissolved organic phosphorus pools (Girault et al., 2013; Mackey et al., 2012). Until recently, it was believed that the dissolved organic nitrogen pool was largely refractory and, therefore, unavailable to phytoplankton. However, a review by Bronk (Bronk et al., 2007) presents evidence that phytoplankton, and particularly some dinoflagellate species, may be able to access this pool, especially some of the more labile fractions such as urea, dissolved free amino acids and nucleic acids. In fact, data presented by Baek et al. (Baek et al., 2008) suggest that Neoceratium are able to use urea as a nitrogen source. It is clearly important to identify which dissolved organic nitrogen sources Neoceratium, and other phytoplankton, are able to utilize in order to determine what role they may play in controlling autotrophic production within OSGs.
Results from experiments on N. hexacanthum in embryo dishes containing NAGSW, where growth did not occur until the addition of nutrients, support our conclusion that cell growth is not possible at nutrient concentrations common in OSG surface waters. These experiments, however, did demonstrate that cells can potentially survive in OSG surface waters, and remain viable, for a long period of time (>3 weeks; Fig. 3). This survival time is consistent with the measurements of Fv/Fm on N. candelabrum, which showed that after 21 days in LNSW the value for this parameter had fallen to almost zero, indicating severe nutrient stress. Few other studies have focused on the long-term survival of phytoplankton under low nutrient concentrations. However, one study on the dinoflagellate Gymnodinium splendens demonstrated that this species could survive for 65 days under oligotrophic conditions (Dodson and Thomas, 1977), nearly three times longer than N. hexacanthum in the present study. In addition to inter-specific nutrient tolerances between Gymnodinium and Neoceratium, this significantly longer survival time is probably the result of the higher nutrient concentrations that were used in that study (NO3−, 0.08 μM; PO43−, 0.46 μM) compared with our own (NO3−, <0.02 μM; PO43−, <0.02 μM). It is likely that this ability to survive long periods of time in the absence of sufficient nutrients may partly explain the success of Neoceratium in oligotrophic regions, and also, their success in stratified coastal regions during summer.
Due to the sampling resolution (2 days), time-lags in growth response, upon nutrient addition, could only be resolved within 2-day intervals. Even so, a clear pattern emerged. The time-lag in growth response increased beyond 10 days of exposure to NAGSW. This increase in time-lag is consistent with studies that have found the delay in growth response to be positively related to duration of nutrient ‘starvation’, especially in species that are capable of storing nutrients internally. This is a strategy that is proposed to be ecologically advantageous in environments where nutrients are only encountered periodically (Collos, 1986). Once cells began to grow, however, there did not appear to be a detrimental impact on growth rates. This suggests that the adverse effects of prolonged exposure to surface oligotrophic conditions, of greater than 10 days, do not persist once cells overcome the initial delay in growth.
There are several important implications of these findings. Neoceratium cells must periodically access an external source of nutrients in order to survive and grow in surface waters of OSGs. To maintain a healthy population in a steady state of growth, this nutrient source would ideally need to be accessed, at most, every 10 days, but could be accessed over a longer timescale (10–20 days) while still allowing for a relatively high chance (∼40%) of survival. Under extreme circumstances, a small number of cells (∼3%) could survive nutrient starvation for longer than 20 days, enabling a ‘seed population’ to survive. In the long term, however, this strategy would not be ecologically successful as cell losses would be too high to maintain a population.
In the final part of this article, we discuss possible mechanisms by which Neoceratium may encounter nutrients periodically in OSGs. Karl (Karl, 1999) identified at least four physical mechanisms that can increase phytoplankton productivity by introducing nutrients into surface waters of subtropical gyres: internal waves and tides, cyclonic mesoscale eddies, wind-driven Ekman pumping and atmospheric storms. Of these four physical mechanisms, mesoscale eddies occur in the most predictable manner, and the increase in nutrient concentrations associated with these events would appear to be great enough to support the growth of Neoceratium (McGillicuddy et al., 1998; Baek et al., 2007). These events alone, occurring approximately once per month in the Sargasso Sea (Siegel et al., 1999), would be almost sufficient to enable Neoceratium to survive via periodic nutrient exposure. It is worth considering, however, that the nutrient concentrations we used (nitrate: 185 µM; phosphate: 1.4 µM) to ‘recover’ nutrient deprived cells had much higher concentrations of nitrate and phosphate than would be found at the nutricline in OSGs (nitrate: no more than 5–10 µM) (Moore et al., 2013), and therefore may not provide a completely accurate proxy for how Neoceratium react when exposed to these waters. For example, a study by Baek et al. (Baek et al., 2008) on N. fusus and N. furca showed that although growth rates peak at 5 µM of nitrate, with additional increases making no difference to how fast cells divide, nutrient uptake rates continue to increase even when growth rates have plateaued; a finding that was interpreted by these authors as Neoceratium storing excess nutrients in vacuoles.
Another strategy may be to obtain nutrients via phagotrophy: a tactic that is thought to be widespread among small eukaryotic algae (<5 μm) in the OSGs of the Atlantic Ocean (Zubkov and Tarran 2008; Hartmann et al. 2011, 2012), and one thought to be used by a number of Neoceratium species, which commonly feed on ciliate prey (e.g. Bockstahler and Coats, 1993; Li et al., 1996; Smalley et al., 1999, 2003; Smalley and Coats, 2002). Some important questions here would be: how much prey would Neoceratium need to ingest in order to obtain sufficient nutrients for survival and growth, and whether encounter rates with prey species are high enough to support this method of nutrient acquisition. It is possible, by making a few assumptions, to estimate the answer to the first of these questions by calculating how much prey (in the form of ciliates) would be required for survival and growth, based on our protein measurements. These measurements of intracellular protein showed that, in N. hexacanthum exposed to LNSW, protein concentrations decrease by 5 ng (±2 ng) over 21 days, from 9.1 to 4.3 ng cell−1. Therefore, to maintain their cells, Neoceratium need approximately 0.14–0.33 ng day−1 of protein to supplement their nitrogen demands. Assuming a nitrogen assimilation efficiency of 75% for type II mixotrophs feeding on microzooplankton (Stickney et al., 2000), Neoceratium would need to ingest 0.19–0.44 ng of protein per day. Using previously published protein contents of a number of ciliate species (1–15 ng) (Zubkov and Sleigh, 1996, 2000), it would appear that an ingestion rate of 0.01–0.44 ciliates day−1 would be required to survive in OSGs. For Neoceratium to divide at the rate observed by Weiler (Weiler, 1980) (0.09–0.16 day−1), adding the further assumption that the complete asexual division process requires the ultimate doubling of cellular protein concentrations, we calculate that Neoceratium would need to consume 0.9–1.9 ng protein day−1, or 0.06–1.9 ciliates day−1. This rate of feeding would appear to be feasible given that the maximum feeding rate observed for N. furca in coastal waters is 2.6 ciliates day−1 (Smalley and Coats, 2002).
Finally, vertical advection and vertical migration are likely to play a role in obtaining nutrients. Vertical advection, aided by turbulent mesoscale eddies, internal waves and Langmuir cells, is capable of transporting cells tens of meters over a period of a few hours (Denman and Gargett, 1983), and may transport Neoceratium cells to the nutricline, or at least close to the nutricline, where vertical migration may become important. Villareal and Lipschultz (Villareal and Lipschultz, 1995) have previously suggested that all large phytoplankton (>100 μm) in the Sargasso Sea are capable of vertical migration, and that they use this mechanism to acquire nutrients from the nutricline. A number of studies indicate that Neoceratium undergo cell division at depth around dawn (Weiler and Chisholm, 1976; Weiler and Eppley, 1979; Baek et al., 2009), followed by upward migration before the transition from dark to light, with downward migration occurring before dusk.
CONCLUSION
Our study demonstrates that Neoeceratium cells are unable to grow phototrophically at low nutrient concentrations representative of surface waters in OSGs. Nevertheless, cells can survive and remain viable for over 3 weeks, suggesting that the presence of Neoceratium in OSGs may be explained by them enduring long periods in surface waters above the nutricline, with periodic exploitation of alternative nutrient sources. Ephemeral upwelling of nutrients, mixotrophy, vertical migration and /or passive advection to the nutricline appear to be the most feasible mechanisms by which nutrients may be obtained. Knowing the relative contribution of these strategies to Neoceratium nutrient acquisition is necessary not just from the point of view of obtaining a more complete understanding of Neoceratium ecology, but also for gaining a better appreciation of how microbial food webs function within OSGs.
FUNDING
This study was supported by the UK Natural Environment Research Council through National Oceanography Centre core funding.
ACKNOWLEDGEMENTS
First, we would like to thank the two anonymous reviewers whose comments and suggestions strengthened this article. We thank Manuela Hartmann and Ross Holland for their support with the laboratory work; Martha Valiadi for providing the Neoceratium cultures; Tom Bibby, Mark Moore and Nicola Pratt for their assistance with the fast repetition rate fluorometer (FRRf); and Adrian Martin for his comments and suggestions on various aspects of the research and the article.
REFERENCES
Author notes
Corresponding editor: Zoe Finkel