INTRODUCTION
Diatoms dominate the phytoplankton community in coastal ecosystems, contributing to ~20% of global primary production (Nelson et al., Reference Passow1995). Because of their large size and silica ballast, they contribute a major fraction of the downward flux of particulate organic matter into deep ocean (Alldredge & Jackson, Reference Alldredge and Jackson1995) and therefore constitute major players in carbon sequestration from the atmosphere to the deep ocean (Boyd et al., Reference Boyd, Strzepek, Fu and Hutchins2010). Chain-forming centric diatoms are the most successful group of eukaryotic primary producers in productive upwelling coastal ecosystems, supporting higher trophic levels. A prerequisite for their high growth rates is an efficient and regulated acquisition of inorganic carbon (Ci) that compensates for the catalytic inefficiency of the enzyme Ribulose-1,5-biphosphate carboxylase/oxygenase (RUBISCO) (Burkhardt et al., Reference Burkhardt, Amoroso, Riebesell and Sültemeyer2001).
Ocean acidification (OA) is a consequence of increased inorganic carbon content of the ocean surface water due to rising atmospheric CO2. The associated drop in the average surface water pH from ~8.2 to 7.8 represents one of the most rapid OA events on earth over the past 300 Myr (Caldeira & Wickett, Reference Caldeira and Wickett2003). It is a controversial issue whether high CO2/low pH in seawater would significantly promote growth and primary productivity. Responses of diatoms to high pCO2 and decreased pH are likely to be species-specific, with potential winners, neutral and losers (Gao & Campbell, Reference Gao and Campbell2014). Diatoms can downregulate the activity of extracellular carbonic anhydrase (Burkhardt et al., Reference Burkhardt, Amoroso, Riebesell and Sültemeyer2001), but they differ in their CO2 concentrating mechanisms (CCMs) (Hopkinson et al., Reference Hopkinson, Meile and Shen2013). Photosynthetic responses to enhanced CO2 under OA are remarkably diverse, and there is a large variability both between and within taxonomic groups. Nevertheless, despite the growing body of literature on the topic, clear trends in the photosynthetic responses of phytoplankton to elevated CO2 have not emerged, and positive effects, if any, are small (Mackey et al., Reference Müller-Navarra, Brett, Park, Chandra, Ballantyne, Zorita and Goldman2015).
In addition to growth and primary production, the elemental composition of phytoplankton might vary under OA. Increasing CO2 can lead to increased particulate organic carbon (POC) content relative to N and P quotas, i.e., higher C:N and C:P under elevated CO2 concentrations (Riebesell et al., Reference Riebesell, Schulz, Bellerby, Botros, Fritsche, Meyerhöfer, Neill, Nondal, Oschlies, Wohlers and Zöllner2007; Feng et al., Reference Feng, Warner, Zhang, Sun, Fu, Rose and Hutchins2008). Nevertheless, systematic increases of particulate organic matter (POM) and C:N ratios have not been observed in response to rising pCO2 and temperature (Burkhardt et al., Reference Burkhardt, Zondervan and Riebesell1999; Wohlers-Zöllner et al., Reference Wohlers-Zöllner, Breithaupt, Walther, Jürgens and Riebesell2011). Moreover, the current evidence demostrated that large differences in the elemental composition of marine phytoplankton can arise from nutrient limitation (Geider & LaRoche, Reference Geider and LaRoche2002), physical factors (Laws & Bannister, Reference Laws and Bannister1980; Burkhardt et al., Reference Burkhardt, Zondervan and Riebesell1999) and interspecific variability among algal species with different C:N:P requirements (Geider & LaRoche, Reference Geider and LaRoche2002). Elevated pCO2 and temperature may lead to a greater partitioning of organic carbon into the dissolved organic carbon (DOC) pool (Kim et al., Reference Kim, Lee, Shin, Yang, Engel, Karl and Kim2011).
Other important proxies for food quality in marine food webs are fatty acid (FA) content and composition. Polyunsaturated fatty acids (PUFA) are considered especially important as they represent essential FAs that cannot be synthesized de novo by heterotrophic consumers (Müller-Navarra et al., Reference Nimer, Warren and Merrett2004). Phytoplankton production of PUFA is highly dependent on the algal physiology and nutrients, and therefore on the environmental conditions (Klein Breteler et al., Reference Klein Breteler, Schogt and Rampen2005; Leu et al., Reference Leu, Daase, Schulz, Stuhr and Riebesell2013). There is a paucity of information on how OA might impact FA contents. Here, we examine the effects of high CO2/low pH and excess nutrients on growth and physiological rates, elemental composition, carbon partitioning and nutritional quality of Skeletonema pseudocostatum in order to evaluate: (1) how the growth rate, photosynthesis, respiration and carbon partitioning of Skeletonema pseudocostatum is affected by a drop of ~0.5 units of pH caused by elevated pCO2 levels and (2) how physiological rates may affect the fatty acid content under condition of increased CO2/lowered pH.
MATERIALS AND METHODS
Semi-automatic system for seawater carbonate manipulation
Experiments were conducted using a semi-automatic mesocosm system for seawater carbonate chemistry manipulation at Calfuco Marine Laboratory in south-central Chile (39°78′S, 73°39′W). CO2-enriched seawater was produced by bubbling seawater with air-CO2 mixtures, following the method described by Torres et al. (Reference Torres, Manriquez, Duarte, Navarro, Lagos, Vargas and Lardies2013). The system uses mass flow controllers (MFC) to blend atmospheric air with ultra-pure CO2 (i.e. research grade) to produce different pCO2 levels (see Table 1). The seawater was continuously bubbled with either ambient or enriched pCO2-air. The high level of pCO2, 1123 µatm, corresponded approximately with projected atmospheric levels between years 2100 and 2150 under the RCP 8.5 scenario (Meinshausen et al., Reference Meinshausen, Smith, Calvin, Daniel, Kainuma, Lamarque, Matsumoto, Montzka, Raper, Riahi, Thomson, Velders and van Vuuren2011). Air/CO2 mixtures were produced using a bulk technique, where dry air with pure CO2 were supplied to seawater using an air mass flow controller (MFC) (Aalborg, model GFC; http://www.aalborg.com) and a CO2 MFC (Aalborg, Model GFC). Dry and filtered air (through a 1 μm particulate filter) was generated by compressing atmospheric air (117 psi) using an oil-free, 4 piston air compressor (Schulz, model MSV12). Pressure in the air and pure CO2 were maintained at ~10 psi. Air flow in MFC was set manually to 5 l min−1 for treatment and CO2 flow was set manually to 4.25 ml min−1 to produce the high CO2 treatment. The CO2 of blended gas was monitored to allow fine regulation of CO2 through MFC to reach target pCO2 in seawater. The pCO2 monitoring system was based on a CO2 analyser (Qubit System, model S151), primarily for measuring the CO2 content in the air-CO2 mixture.
Cell concentration of the marine diatom S. pseudocostatum and average (±SE, N = 4) of carbonate system parameters during (A) acclimation and (B) experimental period under two different pCO2/pH levels; Low pCO2 (L); High pCO2 (H).
DIC, Dissolved inorganic carbon; TA, total alkalinity; HCO3 −, bicarbonate; CO2, carbon dioxide.
First round of acclimation was carried out between 21 and 25 November and second round between 26 and 29 November. Experiment started on 29 November and ended on 3 December.
Culture conditions
Skeletonema pseudocostatum (CSA 48, non axenic) was isolated from Yaldad Bay, southern Chile (43.1°S–73.7°W) in March 2009 and obtained from the COPAS Sur-Austral strain collection (http://www.ficolab.cl/), at the Department of Botany, Concepción University. Cells were grown and acclimated in autoclaved (1 l) and filtered (0.1 µm) natural seawater (salinity: 29.8 PSU) within autoclaved glass bottles (1 L) at the same temperature (14.7 ± 0.6°C), light intensity (195 µmol m−2 s−1) and at a 16/8 light/dark cycle. Light was measured with a sensor Li-192SA (Li-Cor) and was provided by cold white LED tubes (22 Watts, 6000 K). The time of sampling was kept throughout the acclimatization and experiment. For the acclimatization, seawater was enriched to f/2 medium (Guillard & Ryther, Reference Guillard and Ryther1962). Cells were maintained in exponential growth phase using a semicontinuous culture. To maintain balanced exponential growth, cultures were diluted with fresh medium every 3–4 days, keeping cell concentrations <43 × 104 cell ml−1 during the acclimation. During the first (21–25 November) and second round (26–29 November) of the algal acclimation under low pCO2 levels, mean pH values ranged between 8.106 and 8.230, respectively (Table 1). In cultures under high pCO2 conditions, mean pH values ranged between 7.676 and 7.654, respectively. Cultures were acclimated to the respective pH/pCO2 values for 10 generations.
After acclimatization, cells from respective pH/pCO2 treatments were inoculated in autoclaved polycarbonate carboys filled with 20 l of autoclaved seawater (29 November) at the same temperature and light intensity, and carboys were positioned randomly in the experimental system. Carboys were closed with rubber stoppers pierced with glass capillaries for inlet and outlet of air/CO2 mixture. Four carboys were used for low pCO2 and four for high pCO2 treatments, while two control carboys without cells were followed for monitoring abiotic changes in carbonate system parameters. Under these culture conditions, cells were grown for ~6 generations. Samples for carbonate system parameters were taken on 1 and 3 December (Table 1). The harvesting of samples was carried out on 3 December. Cell concentrations at the time of sampling were ~22 × 104 cell ml−1 at low pCO2 and ~14.5 × 104 cell ml−1 at high pCO2. During the experimental period, the photosynthetic activity and cell density increased, leading to an increase in pH (0.11 unit under low CO2 and 0.018 unit under high CO2). The concentration of DIC and total alkalinity (TA) decreased by 35 and 28% under low CO2 conditions and 20 and 19% under high CO2 conditions, respectively.
Chemical analysis
Samples for nutrient analysis (NO3 −, NO2 −, PO4 3− and Si (OH)4) were taken every day during the acclimation and experimental periods (Table 2). Samples were filtered (GF/F) and frozen (−20°C) until analysis following Strickland & Parsons (Reference Strickland and Parsons1968). Daily pH samples were collected in 50 ml syringes and immediately transferred to a 25 ml thermostatted cell at 25.0 ± 0.18°C for standardization, and measurements were conducted with a pH electrode with a glass combined double Ag/AgCl junction (Metrohm model 6.0258.600) calibrated with standard National Bureau of Standards (NBS) calibration buffer Metrohm® 4 (Code 6.2307.200), 7 (Code 6.2307.210) and 9 (Code 6.2307.220). The estimated analysis error for this analysis was estimated as <0.01 pH. For dissolved inorganic carbon (DIC) and DOC determination, separate 30 ml subsamples were collected with a sterile syringe and filtered through a Swinex containing a GF/F filter that had been precombusted for 4–5 h at 450°C directly into 40 ml glass 200 Series I-CHEM® vials. For DIC analyses, the septa of vials were exchanged for butyl rubber septa to prevent diffusion of CO2 (DOE, Reference Dickson, Afghan and Anderson1994). Samples for DIC analysis were preserved with 50 µl of a saturated solution of mercuric chloride (DOE, Reference Dickson, Afghan and Anderson1994). Immediately after opening the sample bottle, a digital syringe withdrew a small amount of sample (0.5 ml), acidified it with 10% phosphoric acid and subsequently measured the evolved CO2 with a LICOR 6262 non-dispersive infrared gas analyser. Certified seawater reference materials from A. Dickson were used to ensure the quality of DIC determination by preparing a calibration curve covering the range of DIC from 200–2000 µeq l−1 (Dickson et al., Reference Dickson and Millero2003), with a resulting precision averaging ≈ 0.1% (range 0.05–0.5%). Temperature and salinity data were used to calculate the other carbonate system parameters (e.g. pCO2, HCO3). Analyses were performed using CO2SYS software for MS Excel (Pierrot et al., Reference Mehrbach, Culberson, Hawley and Pytkovicz2006) set with Mehrbach solubility constants (Mehrbach et al., Reference Nelson, Treguer, Brzezinski, Leynaert and Queguiner1973) refitted by Dickson & Millero (Reference Di Martino, Delne, Alvino and Loreto1987). The KHSO4 equilibrium constant determined by Dickson (Reference Dickson1990) was used for all calculations.
Inorganic nutrient concentrations (±SE, N = 4) and nutrient ratios (±SE, N = 4) during (A) acclimation and (B) experimental period and under two different CO2 concentrations; Low pCO2 (L); High pCO2 (H).
First round of acclimation was carried out between 21 and 25 November and second round between 26 and 29 November. Experiment started in 29 November and ended on 3 December.
For POC and particulate organic nitrogen (PON) analysis, a subsample (1 l) was filtered through combusted (4–5 h at 450°C) GF/F filters to concentrate particles. Filters were dried at ~60°C for 24 h and held in a desiccator until analysed. DOC samples were bubbled with CO2-free nitrogen for 7 min to ensure complete removal of DIC. DOC and POC measurements were conducted by the G.G. Stable Isotope Hatch, Laboratories at the University of Ottawa, Canada, with an analytical precision of 2%. All DOC and POC samples were run on an total inorganic carbon-total organic carbon (TIC-TOC) analyser (OI Analytical Analyzer, Model 1030). Data were normalized using internal standards.
Determination of fatty acids (FAs) was conducted from water samples onto filtered through MFS GF/F filters to concentrate particles. Saturated (SAFA) and unsaturated fatty acids (MUFA: monounsaturated and PUFA: polyunsaturated) were measured on separate filters dried at 50°C for 24 h and held in a desiccator until analysed. The fatty acid concentrations were measured after extraction and methylation (Kattner & Fricke, Reference Kattner and Fricke1986) with a gas chromatograph Perkin Elmer Sigma 300 equipped with a programmable temperature vaporizer-injector, a fused Omegawax 53 capillary column, and a flame ionization detector.
Biological measurements
Concentrations of S. pseudocostatum cells were determined from samples preserved with acid Lugol's. Cell counting was performed using a Neubauer hemocytometer and optical microscope (OLYMPUS CX31). Specific growth rates were calculated using an exponential curve fitted for each replicate of the treatments. The slope of the exponential curve was considered as the growth rate for each pCO2 treatment. Cell size was measured using an epifluorescence microscope (OLYMPUS IX51), choosing at random 40 individual cells for each replicate of each pCO2 treatment. For biovolume calculation we used a cylinder geometric model according to Sun & Liu (Reference Sun and Liu2003). For determination of total chl-a, samples were filtered onto GF/F filters and stored at −20°C. Chl-a was extracted in acetone 95% and measured with a fluorometer (Trilogy Model 7200–040, Turner Designs, Sunnyvale, CA, USA) before and after acidification (Lorenzen, Reference Lorenzen1966). Chl-a, POC, PON and DOC concentrations were normalized per cell (pg cell−1), assuming that changes in Chl-a, N and C from other sources (e.g. lysis) were not significant in our cultures. Gross photosynthesis (GP) and dark respiration (DR) rates were estimated from changes observed in dissolved oxygen concentrations after incubating in vitro in light and dark bottles (Strickland, Reference Strickland1960). Water from three 20 l carboys was transferred to 125 ml borosilicate (i.e. gravimetrically calibrated) using a silicone tube; three time-zero bottles, three light bottles and three dark bottles per replicate were used. The light and dark bottles were incubated at the same temperature and light regime as the 20 l polycarbonate carboy cultures for 6 h; the dissolved oxygen from time-zero bottles was measured at the beginning of the experiment. Dissolved oxygen was measured using a fibre optical oxygen transmitter (Optical Oxygen meter FIBOX, PreSens®). The average of coefficient of variation for replicates was 0.8%. Net photosynthesis (NP) was calculated as the difference in the dissolved oxygen concentration between ‘light’ incubated samples and ‘time zero’ samples. Dark respiration (DR) was calculated as the difference between ‘dark’ incubated samples and ‘time zero’ samples. Dark respiration rates are expressed as a negative O2 flux. Gross photosynthesis (GP) was calculated as the difference between NP and DR (Gaarder & Gran, Reference Gaarder and Gran1927). GP and DR per cell were expressed in fmol cell−1 h−1.
Statistical analysis
In order to evaluate algal responses to experimental conditions, Student's t-test was used for each chemical and biological parameter. The Shapiro–Wilk statistic (Shapiro & Wilk, Reference Shapiro and Wilk1965) was used to check the data for normality distribution and a Levene test checked the homoscedasticity.
RESULTS
Carbonate system
During the experimental period (day 3.12.2013), the carbonate system parameters under simulated CO2-driven ocean acidification showed significant differences in the pCO2 concentration (t = −11.52, df = 6, P < 0.0001) and pH values (t = 26.05, df = 6, P < 0.0001) between both CO2 treatments (Table 1). Significant differences were also found in HCO3 − (t = −2.88, df = 6, P = 0.028) and CO2 (t = −11.44, df = 6, P < 0.0001). There were no significant differences in DIC concentration (t = −2.41, df = 6, P = 0.05) between CO2 treatments, although it was close to the minimal acepted probability. As expected, no significant differences were found in the total alkalinity (TA) (t = −1.22, df = 6, P = 0.26).
Biological parameters
The impact of high pCO2 on S. pseudocostatum physiology was assessed by comparing 12 parameters between low and high pCO2 treatments (Figure 1). Although cell volume was higher at low pCO2 level (442 ± 103 µm3) compared with high pCO2 level (361 ± 42 µm3), there were no significant differences between pCO2 treatments (t = 1.38, df = 6, P > 0.05). There were also no significant differences in growth rates (t = 0.53, df = 6, P > 0.05), cell-normalized Chl-a (t = 2.28, df = 6, P > 0.05), cellular abundance (t = 2.16, df = 4, P > 0.05), gross photosynthesis and respiration rates (t = 2.08, df = 4, P > 0.05; t = −0.74, df = 1.1, P > 0.05, respectively), POC (t = −0.08, df = 5, P > 0.05), and PON (t = −0.75, df = 5, P > 0.05) C:N ratio (t = 1.57, df = 5, P > 0.05), R:P ratio (t = −1.72, df = 1.1, P > 0.05). In contrast, DOC/POC ratio and DOC per cell significantly increased at high pCO2 levels (t = −3.91, df = 2.4, P < 0.05; t = −2.51, df = 6, P < 0.05, respectively). In percentage terms, DOC/POC and DOC per cell increased by 40.4 and 48.4% at high pCO2, respectively.
Physiological parameters: growth rate, gross photosynthesis (GP), dark respiration (DR), cellular concentration of Chl-a, particulate organic carbon (POC) and particulate organic nitrogen (PON), respiration losses in % (R:P ratio), C/N ratio, carbon partitioning (DOC/POC ratio) and DOC concentration measured during semi-continuous cultures of S. pseudocostatum grown at exponential growth phase under two different CO2 concentrations; Low pCO2: 212 ± 30 µatm; High pCO2: 1050 ± 40 µatm. Error bars indicate standard errors.
The high pCO2 treatment exhibited significant differences in FA concentration and composition. Total FA concentration was significantly decreased (t = 5.69, df = 6, P < 0.05) under high pCO2 (0.208 ± 0.04 µg l−1) compared with low pCO2 (0.07 ± 0.01 µg l−1) (Figure 2A). The relative amount of SAFAs was significantly higher (t = −4.53, df = 6, P < 0.05) and the amount of MUFAs lower (t = 16.1, df = 6, P < 0.05) at high pCO2 compared with low pCO2 treatment. In contrast, polyunsaturated fatty acids (PUFA) did not show significant differences between low pCO2 (t = −2.2, df = 6, P > 0.05) (Figure 2B). This is exemplified by some essential fatty acids such as docosahexaenoic acid (DHA, 22:6) and arachidonic acid (ARA, 20:4), which showed similar concentrations under low (0.0008 µg l−1 and 0.001 µg l−1, respectively) and high (0.0002 µg l−1 and 0.002 µg l−1, respectively) pCO2 treatments.
Fatty acid concentration and composition of Skeletonema pseudocostatum cultured at different pCO2 treatments. (A) Total fatty acid and (B) percentage of saturated (SAFA), monounsaturated (MUFA) and polyunsaturated (PUFA) fatty acids relative to total fatty acids during the exponential growth phase cultured at low pCO2: 212 µatm (N = 3) and high pCO2: 1050 µatm (N = 3) treatments. Error bars indicate standard errors.
DISCUSSION
In the coastal domain, surface waters are commonly exposed to levels in partial pCO2 higher than expected at equilibrium with the atmosphere (Hofmann et al., Reference Hofmann, Butler and Tans2011; Yu et al., Reference Yu, Matson, Martz and Hofmann2011), which is mostly associated with biological processes such as daily time cycles of photosynthesis and respiration (Shamberger et al., Reference Shamberger, Feely, Sabine, Atkinson, DeCarlo, MacKenzie, Drupp and Butterfield2011) and oceanographic processes such as riverine discharges and coastal upwelling events (Cao et al., Reference Cao, Dai, Zheng, Wang, Li, Zhai, Meng and Gan2011). In consequence, diatoms inhabiting coastal areas may be capable of tolerating larger ranges of pH and pCO2. However, this high variability also may mean that planktonic organisms inhabiting coastal regions are already operating at the limits of their physiological tolerances. Thus, future OA may drive the physiology of these marine organisms up to the edge in their tolerance range.
The present study showed that high pCO2 had no significant impact on cell volume, growth rate, abundance, chl-a, C/N ratio, and photosynthesis rates in the diatom S. pseudocostatum (Figure 1). Although some studies have indicated that elevated pCO2 is expected to have a stimulative effect on growth rates (Kim et al., Reference Kim, Lee, Shin, Kang, Lee, Kim, Jang and Jang2006; King et al., Reference King, Sañudo-Wilhelmy, Leblanc, Hutchins and Fu2011; Low-Décarie et al., Reference Low-Décarie, Fussmann and Bell2011) and primary productivities (Riebesell et al., Reference Riebesell, Schulz, Bellerby, Botros, Fritsche, Meyerhöfer, Neill, Nondal, Oschlies, Wohlers and Zöllner2007), many other studies have shown that elevated pCO2 concentration did not affect the growth rates in diatom monocultures, including S. costatum (Chen & Gao, Reference Chen and Gao2003, Reference Chen and Gao2004), Thalassiosira pseudonana (Crawfurd et al., Reference Crawfurd, Raven, Wheeler, Baxter and Joint2011; King et al., Reference King, Jenkins, Wallace, Liu, Wikfors, Milke and Shannon2015), and Chaetoceros brevis (Boelen et al., Reference Boelen, Van de Poll, Van der Strate, Neven, Beardall and Buma2011), or in diatom-dominated natural phytoplankton assemblages (Tortell et al., Reference Tortell, Rau and Morel2000). Furthermore, Berge et al. (Reference Berge, Daugbjerg, Andersen and Hansen2010) have also shown that low pH/high pCO2 conditions do not affect the growth rate and production rates of eight species of phytoplankton representing diatoms, dinoflagellates, cryptophytes and haptophytes. Berge et al. (Reference Berge, Daugbjerg, Andersen and Hansen2010) also showed that 49 strains of a total 33 species of phytoplankton exhibited similar growth rates at pH ~7.8 compared with more alkaline levels of pH (8.1–8.2), suggesting that marine phytoplankton are adapted to tolerate the modelled average pH drop due to ocean acidification by the year 2100. Our findings are consistent with these findings, suggesting that S. pseudocostatum tolerates well changes of 0.5 units of pH due to manipulated pCO2 levels.
Diatoms have the capacity for simultaneous transport of CO2 and HCO3 − during photosynthesis and increase their affinities of both transport systems in response to diminishing supply of carbon substrate (Burkhardt et al., Reference Burkhardt, Amoroso, Riebesell and Sültemeyer2001). However, the proportion at which CO2 and HCO3 − are taken up and the extent to which Ci uptake is affected by changes in CO2 supply vary among phytoplankton species (Nimer et al., Reference Pierrot, Lewis and Wallace1998; Elzenga et al., Reference Elzenga, Prins and Stefels2000; Burkhardt et al., Reference Burkhardt, Amoroso, Riebesell and Sültemeyer2001; Rost et al., Reference Rost, Riebesell, Burkhardt and Sültemeyer2003). For example, Phaeodactylum tricornutum takes up CO2 preferentially over HCO3 − from seawater whereas Thalassiosira weissflogii takes up HCO3 − preferentially to CO2 under depletion of CO2; thus for Phaeodactylum tricornutum one would expect a pronounced response in photosynthetic C fixation under enhanced CO2. However, P. tricornutum showed increased photosynthetic electron transport rates, but no change or very modest increases in growth (5–13%; Wu et al., Reference Wu, Gao and Riebesell2010; Li et al., Reference Li, Xu and Gao2014) or carbon fixation (Burkhardt et al., Reference Burkhardt, Amoroso, Riebesell and Sültemeyer2001) under high pCO2. In our experiment (3 December), significant differences in HCO3 − concentrations were observed between pCO2 treatments (Table 1), suggesting that both free CO2 and HCO3 − were probably an important inorganic carbon source for Skeletonema pseudocostatum cells.
Rising pCO2 might affect primary producers in terms of saving energy required for active inorganic carbon acquisition, whereas low pH could potentially increase metabolic demand to maintain celular homeostasis relative to the increased acidity (Gao et al., Reference Gao, Helbling, Häder and Hutchins2012). Nevertheless, our results did not show significant differences in the respiration rates between low and high pCO2 levels (Figure 1), which suggest no significant changes in the balance between production and consumption (see R:P ratio) between low and high pCO2 levels.
Furthermore, our results also showed that there were no significant differences in the elemental composition (C:N ratio) between the pCO2 treatments (Figure 1), indicating that increasing CO2 would not increase the POC:PON ratio. In contrast, DOC per cell and the DOC/POC ratio were significantly higher at high pCO2, suggesting that extracellular carbon release relative to particulate carbon production increased under elevated pCO2. The extracellular release of photosynthesis products is especially common during nutrient-depleted growth conditions (Kim et al., Reference Kim, Lee, Shin, Yang, Engel, Karl and Kim2011; Borchard & Engel, Reference Borchard and Engel2012), since phytoplankton exude DOC to the environment to reduce the energy costs associated with storing surplus compounds (Wood & Van Valen, Reference Wood and Van Valen1990). However, it has also been reported to occur independent of nutrient availability (Hessen et al., Reference Hessen, Ågren and Anderson2004; Hessen & Anderson, Reference Hessen and Anderson2008) and under continuous CO2 enrichment (Song et al., Reference Song, Ballantyne and Smith2013). In our experiment cells were under nutrient-replete conditions (Table 2) and continuous CO2 enrichment, so the DOC increase under high pCO2 conditions seems not to be nutrient dependent. The higher DOC release triggered by elevated CO2 is consistent with other studies carried out in natural phytoplankton assemblages (Riebesell et al., Reference Riebesell, Schulz, Bellerby, Botros, Fritsche, Meyerhöfer, Neill, Nondal, Oschlies, Wohlers and Zöllner2007; Kim et al., Reference Kim, Lee, Shin, Yang, Engel, Karl and Kim2011; Engel et al., Reference Engel, Borchard, Piontek, Schulz, Riebesell and Bellerby2013) and monocultures (Engel et al., Reference Engel, Delill, Jacquet, Riebesell, Rochelle-Newall, Terbrüggen and Zondervan2004; Borchard & Engel, Reference Borchard and Engel2012). Most experimental studies have suggested that greater assimilation of carbon into organic matter at high CO2 levels may increase the extracellular organic matter release from phytoplankton. However, our study showed that gross photosynthesis, POC concentration and C:N ratio were not significantly different between pCO2 treatments, which rules out this process. Enrichment of CO2 and increased acidity have also been found to stimulate photorespiration in diatoms T. pseudonana and P. tricornutum (Gao et al., Reference Gao, Helbling, Häder and Hutchins2012), a process by which oxygen is consumed and CO2 released under light conditions, as well as a process by which glycolate is lost to the outside medium as an excreted product. It has been suggested that photorespiration is important for maintaining electron flow to prevent photoinhibition under stress conditions (i.e. high CO2 levels) (Heber et al., Reference Heber, Bligny, Streb and Douce1996), as well as under drought stress (Wingler et al., Reference Wingler, Quick, Bungard, Bailey, Lea and Leegood1999). In addition, the formation of photorespiratory metabolites, such as glycine, serine and glycolate has also been measured under salt stress in C3 plants (Downton, Reference Downton1977; Di Martino et al., Reference Dickson and Goyet1999). Therefore, photorespiration may play a protective role under stress conditions and consequently contribute partially to the release of DOC triggered under high pCO2 levels. We also recognize that our experiment was not axenic and consequently the increase of DOC in the bottles could also be produced by lysis and transformation of POC to DOC by bacteria or chemical hydrolysis (Carlson, Reference Carlson, Hansell and Carlson2002). The greater partitioning of organic carbon into the DOC pool under high CO2/low pH conditions may have implications for long-term C storage in aquatic ecosystems, as DOC components can be both important precursors in the creation of large particle aggregates and also in the formation of recalcitrant DOC (Engel Reference Engel2002) via heterotrophic metabolism in the upper layers of aquatic ecosystems (Jiao et al., Reference Jiao, Herndl, Hansell, Benner, Kattner, Wilhelm, Kirchman, Weinhauer, Tingwei, Chen and Azam2010). Important questions arise regarding the increase of DOC production under acidification condition scenarios, e.g. (i) How the increase of DOC will affect the C cycling through bacteria; (ii) the formation of transparent exopolymer particles and consequently, the export of organic matter to the deep ocean (Passow, Reference Mackey, Morris, Morel and Kranz2002); and (iii) nutrient competition between bacteria and phytoplankton.
The effects of lowered pH and increased pCO2 was also evaluated on nutritional quality of S. pseudocostatum. Total FAs were significatively different between pCO2-treatments, being 63.26% higher under low pCO2 compared with high pCO2 treatment (Figure 2A). These results agree with other studies that showed a significant decline in total FAs of the centric diatom T. pseudonana under elevated CO2 (750 µatm) compared with present-day CO2 (380 µatm) (Rossoll et al., Reference Rossoll, Bermúdez, Hauss, Schulz, Riebesell, Sommer and Winder2012). These authors found that the relative amount of SAFAs was significantly higher at high CO2 and a ~20% decline in the relative amount of PUFAs. Our findings showed that the relative amount of SAFAs was significantly higher (44 to 63%) and MUFAs significantly lower (44 to 10%) at high pCO2 compared with the low pCO2 treatment. In contrast, lowered pH and elevated CO2 did not affect the contribution of PUFAs to total fatty acids significantly (Figure 2B). The important increase of saturated and decrease of monounsaturated FA contents and total FAs under acidification may affect the transfer of lipids to higher trophic levels. However, the nutritional quality in terms of essential FAs remains unchanged. Most lipids consist mainly of hydrocarbon chains with varying numbers of double bonds. SAFA have hydrocarbon chains with single bonds while polyunsaturated FAs contain more than one double bond and include many compounds essential for higher trophic levels, such as for copepod egg production, hatching and maturity (Jonasdottir et al., Reference Jonasdottir, Trung, Hansen and Gartner2005; Klein Breteler et al., Reference Klein Breteler, Schogt and Rampen2005). Our findings are consistent with other studies showing no detrimental effects of high pCO2 on the nutritional quality in terms of essential fatty acids (Leu et al., Reference Leu, Daase, Schulz, Stuhr and Riebesell2013). Many other responses can be expected in the total FAs and components depending on phytoplankton functional group and species. For example, declining PUFA content at elevated pCO2 was reported for the Antarctic prasinophyte Pyramimonas gelidicola (Wynn-Edwards et al., Reference Wynn-Edwards, King, Davidson, Wright, Nichols, Simon, Kawagushi and Vitue2014), the sea-ice diatom N. lecointei (Torstensson et al., Reference Torstensson, Hedblom, Andersson, Andersson and Wulff2013) and the diatom Cylindrotheca fusiformis (Bermúdez et al., Reference Bermúdez, Feng, Roleda, Tatters, Hutchins, Larsen, Boyd, Hurd, Riebesell and Winder2015). No detectable differences attributable to pCO2 treatment in the fatty acids component has been observed for the centric diatoms T. pseudonana and T. weissflogii, the green algae Dunaliella salina, the euryhaline microalgae Chlorella autotrophica (King et al., Reference King, Jenkins, Wallace, Liu, Wikfors, Milke and Shannon2015) and the dinoflagellate Gymnodinium sp. (Wynn-Edwards et al., Reference Wynn-Edwards, King, Davidson, Wright, Nichols, Simon, Kawagushi and Vitue2014). In contrast, high CO2 increased the accumulation of total lipids and polyunsaturated fatty acids in the chlorophytes Scenedesmus obliquus and Chlorella pyrenoidosa (Tang et al., Reference Tang, Han, Li, Miao and Zhong2011). The cellular processes involved in FA synthesis under changing pH and pCO2 levels are not fully understood. Because pH might act as a regulation signal for the formation of cell membranes by controlling the production of its synthesizing enzymes (Young et al., Reference Young, Shin, Orij, Chao, Li, Guan, Khong, Jan, Wenk, Prinz, Smits and Loewen2010), it has been proposed that a higher saturation degree at high CO2 levels (increase of SAFA) may be a mechanism to control the internal cell-pH because a membrane built of short chain FA is less fluid and permeable to CO2 (Rossoll et al., Reference Rossoll, Bermúdez, Hauss, Schulz, Riebesell, Sommer and Winder2012).
Our findings suggest that growth, gross photosynthesis and C:N ratio were not necessarily connected to CO2-driven changes in composition and content of FAs in S. pseudocostatum. In agreement, other studies showed that CO2-driven changes in the growth rate of the centric diatom Thalassiosira weissflogii were not reflected by significant changes in the elemental composition and fatty acid composition, which indicate bidirectional responses to changes in CO2 (King et al., Reference King, Jenkins, Wallace, Liu, Wikfors, Milke and Shannon2015). Since ocean acidification has the potential to alter phytoplankton biochemistry, our results highlight the importance for understanding the cellular processes involved in FA synthesis under rising CO2/decreasing pH, which will finally determine the carbon transfer efficiency to higher trophic levels in a changing ocean.
ACKNOWLEDGEMENTS
We acknowledge the skilful help of all assistants involved in this research, especially Mr Rui Feng and Ms Paulina Contreras for their logistic support. The authors would like to thank Rodrigo Torres for his advice in carbonate chemistry and CO2-mesocosm management at Calfuco. We would like to thank the FICOLAB group of Concepcion University.
FINANCIAL SUPPORT
This work was supported by projects Fondecyt 1130254, 3120089 and 3130722 funded by CONICYT-Chile. Additional support was provided by the Millennium Nucleus ‘Center for the Study of Multiple-drivers on Marine Socio-Ecological Systems (MUSELS)’ funded by MINECON NC 120086 and the Millennium Institute of Oceanography (IMO) funded by MINECON IC120019. C.A.V. was also supported by Red Doctoral REDOC. CTA, MINEDUC project UCO1202 at Universidad de Concepción.
