Interactive effects of light, CO2 and temperature on growth and resource partitioning by the mixotrophic dinoflagellate, Karlodinium veneficum

There is little information on the impacts of climate change on resource partitioning for mixotrophic phytoplankton. Here, we investigated the hypothesis that light interacts with temperature and CO2 to affect changes in growth and cellular carbon and nitrogen content of the mixotrophic dinoflagellate, Karlodinium veneficum, with increasing cellular carbon and nitrogen content under low light conditions and increased growth under high light conditions. Using a multifactorial design, the interactive effects of light, temperature and CO2 were investigated on K. veneficum at ambient temperature and CO2 levels (25°C, 375 ppm), high temperature (30°C, 375 ppm CO2), high CO2 (30°C, 750 ppm CO2), or a combination of both high temperature and CO2 (30°C, 750 ppm CO2) at low light intensities (LL: 70 μmol photons m-2 s-2) and light-saturated conditions (HL: 140 μmol photons m-2 s-2). Results revealed significant interactions between light and temperature for all parameters. Growth rates were not significantly different among LL treatments, but increased significantly with temperature or a combination of elevated temperature and CO2 under HL compared to ambient conditions. Particulate carbon and nitrogen content increased in response to temperature or a combination of elevated temperature and CO2 under LL conditions, but significantly decreased in HL cultures exposed to elevated temperature and/or CO2 compared to ambient conditions at HL. Significant increases in C:N ratios were observed only in the combined treatment under LL, suggesting a synergistic effect of temperature and CO2 on carbon assimilation, while increases in C:N under HL were driven only by an increase in CO2. Results indicate light-driven variations in growth and nutrient acquisition strategies for K. veneficum that may benefit this species under anticipated climate change conditions (elevated light, temperature and pCO2) while also affecting trophic transfer efficiency during blooms of this species.

Introduction toxicity may alter trophic interaction by this species. However, these interactions may also be affected by cellular nutrient status under anticipated changes in CO 2 , temperature and light. Carbon and nitrogen acquisition strategies by K. veneficum in response to the environmental stressors (CO 2 , temperature, light) have yet to be evaluated.
The study presented here investigated the interactive effects of light, CO 2 and temperature on autotrophic growth and resource partitioning in Karlodinium veneficum. Given prior research results discussed above, one would expect that increases in pCO 2 and temperature alone would increase cellular carbon and nitrogen due to enhanced metabolic activities and reduced energy required for carbon assimilation. When combined with higher light intensity, however, any increases in cellular carbon and nitrogen content may be offset or attenuated by an increase in growth rates, effectively dampening changes in cell-specific nutrient status. Using a multifactorial design, we tested this hypothesis-that an increase in pCO 2 and temperature under low light conditions would increase cellular carbon and nitrogen assimilation while limiting growth, while an increase in pCO 2 and temperature under high light conditions would enhance growth, while limiting cellular carbon and nitrogen assimilation. Results of this study indicated significant interactions between light and temperature on growth and resource partitioning for K. veneficum. These interactions may benefit this species while also affecting trophic transfer efficiency and phytoplankton community dynamics during blooms of K. veneficum under future climate change conditions.

Culture maintenance
Karlodinium veneficum was originally isolated from Delaware Inland Bays, DE, and is available through the Provasoli-Guillard National Center for Marine Algae and Microbiota (https:// ncma.bigelow.org/home; CCMP2936). Stock cultures were grown on a f/2-Si culture medium [47] made up in low nutrient 0.2 μm-filtered offshore seawater diluted to the salinity of 20 psu and autoclaved before addition of nutrients, trace metals, and vitamins. Batch cultures were acclimated at 25˚C and 30˚C for over 12 months, with a 12 h light: 12 h dark cycle, and transferred to fresh medium every 10-11 days. While cultures were not axenic, kanamycin (50 mg L -1 ) was added to cultures periodically to control bacterial growth.

Experimental design
Cultures (N = 4) were grown in 1-L polycarbonate bottles, fitted with silicone tubing attached to a glass-frit gas diffuser, and acclimated to gentle bubbling with air prior to the start of the experiment. Cultures at each experimental temperature were then gently bubbled with ambient (375 ppm) or elevated (750 ppm) CO 2 in commercially prepared and certified air/CO 2 gas mixture (Scott TM Gas Mixture, Scott Company, Plumsteadville, PA) for 15 days (at least four generations) before the start of the experiment. The gas was delivered through 0.2 μm air filters. For the low light (LL) and high light (HL) experiments, cultures were grown under cool white fluorescent lights at an irradiance of 70 (±2) μmol quanta m -2 s -1 and 140 (±3) μmol quanta m -2 s -1 , respectively. At each light level, cultures were subjected to four treatment conditions: 375 ppm CO 2 at 25˚C, 375 ppm CO 2 at 30˚C, 750 ppm CO 2 at 25˚C, or 750 ppm CO 2 at 30˚C (Fig 1). Cultures (N = 4) were maintained in semi-continuous growth for an additional nine days by diluting cultures to initial cellular concentrations (50,000 cells mL -1 ) every other day with sterile f/2 medium, pre-equilibrated with air/CO 2 mixture at the appropriate CO 2 level.

Growth rates and cell volume
Cells were fixed with gluteraldehyde at a final concentration of 0.1%, and counted using a hemocytometer (Hausser Scientific, Horsham, Pennsylvania). Growth (μ, d -1 ) was calculated as in Eq (1): where N 1 and N 2 were the cell abundance at times t 1 and t 2 .
Cell size was measured on Day 9 for high light (HL) experiments only, using a Multisizer 3 Coulter Counter (Beckman Coulter, Inc., Brea, CA). Samples were diluted in filtered seawater and cell radius was calculated from the average of four measurements for each culture. Cell biovolume was calculated from the cell radius (r) as in Eq (2), assuming a spherical shape:

Particulate carbon and nitrogen
Samples were collected on Day 9 and filtered onto pre-combusted (450˚C for 4 hours) glass fiber filters for particulate carbon (PC) and nitrogen (PN) analysis. Filters were dried at 55˚C and analyzed using a Costech Elemental Combustion System 4010 (Costech Analytical Technologies, Inc., Valencia, CA). EDTA and phenylalanine were used as standards.

Carbon and nitrogen production rates
Particulate carbon (PC) and nitrogen (PN) production rates were calculated as in Eqs (3) and (4) [31]: where μ is the growth rate and PC and PN are particulate carbon and nitrogen content per cell.

RuBisCO expression
Primer design. Primers targeting ribulose-1,5-bisphosphate carboxylase oxygenase (RuBisCO) large subunit (rbcL) and β-actin were designed for K. veneficum based on sequences available in GenBank (GenBank accession number AF463410 and AY345907, respectively). Primers were designed to amplify 250 bp of the rbcL gene and 207 bp of the endogenous control β-actin for K. veneficum (Table 1). To confirm the sequence of each gene amplified with these primers, DNA was extracted from K. veneficum as described in Coyne et al. [50], and amplified by PCR in 20 μL reactions. Each reaction contained 0.2 mM dNTPs, 0.5 μM primers (KvRUBISCO 508F and KvRUBISCO 758R or KvACTIN 703F and KvACTIN 910R; Table 1 The cycling parameters for all reactions were 2 min at 50˚C, 10 min at 95˚C, followed by 40 cycles at 95˚C for 30 seconds, an annealing step at 56˚C for 30 seconds, and extension at 60˚C for 1 minute. Products were evaluated by plotting the dissociation of duplex DNA with the stepwise increase in temperature from 60˚C to 95˚C. Transcript abundance was calculated by linear regression analysis, and rbcL transcript abundance was normalized to βactin to determine the relative rbcL expression for each treatment.

Statistics
Results for each analysis were compared using a one-way ANOVA, followed by Tukey HSD post hoc testing using PAST (PAlaeontological STatistics, ver.3; [51]. The data were evaluated for normality and homoscedasticity prior to statistical analysis. Differences were considered significantly different when p<0.05. Results for growth rates, particulate nitrogen and carbon analysis, productivity, and rbcL expression were also compared using a two-way ANOVA to assess significant interactions among combinations of light intensity, temperature and CO 2 for all treatments and between temperature and CO 2 within each light level.

Dissolved inorganic carbon and CO 2 concentrations
pH, pCO 2 and DIC are shown in Table 2. Overall, pH was significantly lower at 750 ppm CO 2 than at 350 ppm CO 2 but there were no significant differences in pH between different light levels or temperatures. Dissolved inorganic carbon (DIC) concentrations were not significantly different between different light levels for each treatment. Under LL, DIC in the 30˚C, 750 ppm CO 2 treatment was significantly higher than the 350 ppm CO 2 treatments. Under HL, the DIC concentration in the 25˚C, 350 ppm CO 2 treatment was significantly lower than other treatments (P<0.05). No significant differences were observed between CO 2 levels within different CO 2 treatments.

Growth rates
For LL treatments, there were no significant differences in growth rates among any of the treatments with an average of 0.25 ± 0.01 day -1 (Fig 2A). Under HL conditions, growth rates were significantly higher than under LL, and ranged from 0.36 ± 0.03 day -1 to 0.47 ± 0.01 day -1 (Fig 2A). Growth rates were also significantly higher in 30˚C treatments compared to ambient conditions at HL (p<0.05). Analysis by two-way ANOVA revealed no significant interaction by CO 2 and temperature with respect to growth either within light treatments or across both light treatments. However, there was a significant interaction between light and CO 2 as well as light and temperature across all treatments (p<0.05).
Under HL conditions, elevated temperature and CO 2 , as well as the interaction of temperature with CO 2 , all had significant effects on the cell volume (p<0.005), shifting cell dimensions in favor of the smaller cells compared to those at ambient conditions of temperature and CO 2 in HL ( Fig 2B).

Particulate carbon and nitrogen analysis
Particulate carbon (PC) and nitrogen (PN) significantly increased in response to elevated temperature compared to ambient conditions under LL (p<0.00005, Fig 3A and 3B). Particulate carbon and nitrogen in HL were significantly higher than all LL treatments (p<0.05). In contrast to LL cultures, PC and PN were significantly higher in HL cultures maintained under ambient temperature and CO 2 than in elevated temperature and/or CO 2 under HL (p<0.05; Fig 3A and 3B).
Two-way ANOVA revealed a significant positive interaction between light and temperature for both PC and PN, but no significant interactions between light and CO 2 , or CO 2 and temperature when all results were included. When each light level was evaluated separately, there was a significant interaction between temperature and CO 2 on both PN and PC under HL but not LL.
Significant differences in carbon: nitrogen (C:N) under LL occurred only in the combined high temperature with high CO 2 treatment (P<0.005) (Fig 3C). Two-way ANOVA analysis revealed a significant positive interaction between CO 2 and temperature on cellular C:N under LL (p<0.001). C:N in HL cultures were significantly higher than LL cultures for all treatments (p<0.05; Fig 3C). CO 2 also had a slight but significant positive effect on cellular C:N ratios under HL compared to other cultures at HL (Fig 3C). There were no significant interactions between temperature and CO 2 for C:N within the HL treatment, and no significant interactions between light and CO 2 or light and temperature across all treatments.

Carbon and nitrogen production rates
Particulate carbon (PC) and nitrogen (PN) production rates mirrored PC and PN in LL treatments, so that elevated temperature had a significant positive effect on carbon and nitrogen assimilation rates compared to cultures in ambient conditions under LL (p<0.01; Fig 4A and  4B). Under HL, carbon production rates were significantly higher in response to a combination of high temperature and CO 2 (0>0.05; Fig 4A). Nitrogen production rates under HL significantly decreased in response to elevated temperature under ambient CO 2 and under elevated CO 2 under ambient temperature compared to cultures maintained at ambient temperature and CO 2 or cultures maintained at elevated temperature and CO 2 ( Fig 4B). Two-way ANOVA analysis revealed a significant interaction between temperature and light for both PC and PN production rates, but no significant interaction between light and CO 2 . Within HL treatments, there was also a significant interaction between CO 2 and temperature for both PC and PN production rates (p<0.05).

RuBisCO (rbcL) gene expression
Transcript abundance of rbcL (relative to actin transcript abundance) was significantly higher in LL treatments under high temperature and a combination of high temperature and CO 2 (p<0.05; Fig 5). Under HL conditions, however, no significant difference was detected among the treatments. Two factor ANOVA revealed a significant interaction between light and temperature (p<0.0005), but no significant interaction between light and CO 2 or between CO 2 and temperature within light levels or across all light levels.

Discussion
This study investigated the interactive effects of light, temperature and CO 2 on growth and cellular carbon and nitrogen assimilation for the mixotrophic dinoflagellate, Karlodinium veneficum. Several lines of evidence suggest that K. veneficum used different strategies for growth and resource partitioning under climate change conditions depending on light intensity. Comparison of cultures maintained under ambient temperature and CO 2 showed that light alone had a significant impact, with higher growth rates, cellular carbon and nitrogen content and production rates and C:N at high light compared to low light conditions. Within each irradiance level, however, there were distinct differences in patterns of response to elevated CO 2 , elevated temperature and combined increases in CO 2 and temperature treatments (Fig 6), with statistically significant interactions between light level and temperature for all parameters tested. This may be reflective of the need for cells to balance light-responsive activities such as light-harvesting and electron transport, with the effects of temperature and/or pCO 2 on metabolic activities that serve as energy sinks, such as carbon and nitrogen assimilation [52,53]. These changes in nutrient partitioning strategies of K. veneficum may have broader impacts on trophic transfer efficiency and phytoplankton community dynamics with anticipated changes in light regime under climate change conditions.

Growth and resource partitioning under low light
Under low light conditions, there were no significant differences in growth rates among treatments (Fig 2A), suggesting that light was a limiting factor to population growth under all conditions tested. The addition of CO 2 alone also had no significant effect on any of the parameters measured under low light conditions. Coastal environments such as those inhabited by K. veneficum experience wide fluctuations in CO 2 , with increases from terrestrial input and heterotrophic activities and decreases in CO 2 from photosynthetic activity [21]. Dinoflagellates inhabiting these environments have efficient carbon concentrating mechanisms (CCMs; [54]) that increase CO 2 concentration at the active  Increased particulate carbon (PC), particulate nitrogen (PN), and both PC and PN production rates by K. veneficum in response to elevated temperature suggest increased metabolic activity in cultures acclimated to 30˚C. This is also supported by an increase in rbcL gene expression (Fig 5), providing evidence for enhanced RuBisCO activity and/or turnover in response to temperature [57]. The synergistic effects of elevated CO 2 and temperature on PC indicate a substantial increase in carbon fixation as a result of combined temperature-induced increase in RuBisCO activity and an increase in CO 2 -saturated rate of carbon assimilation with the increased availability of CO 2 substrate [58]. In contrast, there were no significant differences in PN among cultures exposed to high temperature, high CO 2 , or a combination of high temperature and CO 2 under low light conditions. As nitrate was the only nitrogen source provided in these cultures, it would have been assimilated through nitrate reductase activity, which competes with RuBisCO for reductant generated by photosynthesis [59,60]. Under low light, high temperature conditions, enhanced RuBisCO activity with elevated CO 2 may have shifted the limited amount of reductant toward carbon fixation activities, at the expense of nitrogen assimilation. While cell size was not measured in low light cultures, it is reasonable to suggest that increased cellular PC and PN production rates in treatments at 30˚C compared to ambient conditions would result in an increase in biovolume for K. veneficum, even without higher growth rates.

Growth and resource partitioning under high light
While effects of elevated temperature and CO 2 under low light conditions indicate an increase in productivity for K. veneficum without an increase in cell abundance, comparison to these same treatments at high light suggest that this species uses distinctly different strategies for resource partitioning when irradiance increases along with temperature and CO 2 . In contrast to low light, for example, resources were allocated to cell division at the expense of biomass when cultures were maintained at high light. Significant decreases in PC and PN on a per cell basis for cultures maintained in high temperature and/or CO 2 suggested a potential decrease in cell size, which was confirmed by measurements of cell biovolume (Fig 2B). When calculated on a per biovolume basis, there were no significant differences in PC, PN or C:N across all treatments at high light, indicating a size dependency on carbon and nitrogen assimilation when light was not limiting.
Of particular interest here is the interaction between temperature and CO 2 under high light conditions with respect to cellular carbon and nitrogen production rates (Fig 4). Results presented here suggest that (1) when compared to cultures under ambient conditions of temperature and CO 2 , the decrease in PC and PN content for high temperature and CO 2 treatments under HL can be attributed to decreases in cell biovolume, whereas (2) the significant increase in PC and PN production rates in cultures maintained with a combination of high temperature and CO 2 implies that productivity was limited by low temperature at high CO 2 and by low CO 2 at high temperature. Studies on coccolithophores have demonstrated a link between pCO 2 optima for carbon fixation and temperature [61]. Results of this study suggest a similar shift in pCO 2 optima for K. veneficum, so that when light was not limiting, a combination of increased CO 2 with increased temperature was necessary to achieve significant gains in growth and productivity compared to each individual driver.

Relevance to natural populations of Karlodinium veneficum
Expanded seasonal warming [62] along with increased toxicity in K. veneficum with elevated CO 2 [46] may have devastating consequences for local fisheries, as blooms become more toxic and occur over a wider temporal and geographic range. The temperatures used in this study spanned a range that K. veneficum would encounter in mid-Atlantic environments, where blooms of this species occur annually [34,35]. Previous analysis of field data for Delaware's inland bays, however, indicated that the abundance of natural K. veneficum populations peaked at temperatures between 25˚C and 29˚C, with a sharp drop in cell density at temperatures used here for high temperature treatments (30˚C) [35]. The field data is consistent with reports of laboratory culture experiments using the same strain of K. veneficum, in which Vidyarathna et al.
[35] identified a temperature optimum for this species of 28.6˚C with a drop in growth rate at 30˚C. The ability of K. veneficum to maintain growth rates after long-term exposure to elevated temperature (> 1 year) in the present study suggests a capacity for either phenotypic acclimation or genetic adaptation to conditions [53,63,64] that may benefit K. veneficum under anticipated future temperature conditions. The reduction in cell biovolume for K. veneficum maintained in high light under elevated temperature and/or CO 2 conditions may also improve its competitive ability under climate change conditions. Smaller cells can more efficiently harvest light and nutrients, and are better able to maintain their position in the water column [65,66]. It has been predicted that species with smaller cell volumes will become dominant under climate-induced stresses such as lightsaturated conditions along with high temperature, and elevated CO 2 [67,68]. The decrease in cell biovolume and increase in growth rates for K. veneficum measured here for high light cultures in response to high temperature and CO 2 may provide this species with a competitive advantage under climate change conditions, and suggests an increase in intensity for K. veneficum blooms.

Trophic implications
In addition to abiotic factors such as light, pCO 2 and temperature regimes encountered in the natural environment, trophic interactions also have an effect on K. veneficum population dynamics and may be altered by anticipated changes in climate.
Zooplankton play a key role in regulating phytoplankton abundance and community composition through top-down control [69,70]. Grazing effects are modulated by physical parameters such as temperature and light [71], as well as nutrient concentrations [72][73][74], taxonomic composition [75,76] and the quantity and nutritional quality of prey [77,78]. In addition, a shift in phytoplankton toward HAB species can negatively affect both the growth and grazing activity of microzooplankton [38].
Results of this study point toward potential impacts of climate change on trophic transfer efficiency by K. veneficum as both predator and as prey [42,79]. Increased toxicity of K. veneficum as would be expected under elevated CO 2 [42], for example, may reduce top down control on this species by meso-and microzooplankton predators [41, 80] while increasing predation by K. veneficum on other protists [38,79]. For example, mixed prey experiments described in Adolf et al. [41] showed that the presence of toxic K. veneficum inhibited grazing by the heterotrophic dinoflagellate, Oxyrrhis marina, on co-occurring non-toxic strains. Effects of K. veneficum toxicity on copepod grazers has also been shown to be species-dependent [81], suggesting that an increase in toxicity along with increased intensity of K. veneficum blooms under elevated pCO 2 and high light conditions as demonstrated here might skew copepod populations, favoring more tolerant species. Nutrient cycling dynamics may also be affected. For example, Saba et al. [82] reported an increase in DOM released by the copepod grazer, Acartia tonsa, when feeding on toxic K. veneficum vs. a non-toxic strain, suggesting a feedback loop that may contribute to an increase in toxic blooms of this species [82].
In addition to changes in toxicity, altered nutritional quality of K. veneficum under climate change conditions may have a negative effect on trophic transfer. Carbon: nitrogen ratios (C: N) are considered a proxy for cellular protein content, with a higher ratio indicating lower nutritional value for zooplankton [8,83]. Even when phytoplankton biomass was reported to increase in response to climate change conditions, the low nutritional quality of the phytoplankton (high C:N) negatively affected higher trophic levels, reducing mass transfer efficiency [84] and enhancing the stoichiometric mismatch between phytoplankton and their consumers [83]. Zooplankton abundance and community composition may also be negatively impacted by high C:N, with a decrease in production [85] and hatchability [86]. Relative to results reported here, Smith [87] also observed significantly reduced egg production and hatching success when Acartia tonsa grazed on K. veneficum under elevated temperature and pCO 2 treatments. Hence, the trend toward increased C:N in K. veneficum suggests that blooms of this species under anticipated climate change conditions will have a negative effect on zooplankton grazers, in spite of the potential increase in the availability of algal prey due to higher growth rates.

Conclusions
Results of this study highlight differential responses to anticipated changes in temperature, CO 2 , and light intensity for the mixotrophic dinoflagellate, K. veneficum. Growth rate was enhanced under high light in response to both elevated temperature and a combination of high temperature and CO 2 , suggesting an increase in the intensity of blooms of this species with climate change. The increase in C:N for the elevated temperature and CO 2 treatments may also impact trophic interactions, with a reduction in nutritional value for zooplankton grazing on K. veneficum. When coupled with the increase in toxicity [46], results presented here indicate that top-down control of K. veneficum blooms by predators may decrease, providing a competitive advantage to this species under anticipated climate change conditions.