Photochemical responses of the diatom Skeletonema costatum grown under elevated CO 2 concentrations to short-term changes in pH

Variability in pH is a common occurrence in many aquatic environments, due to physical, chemical and biological processes. In coastal waters, lagoons, estuaries and inland waters, pH can change very rapidly (within seconds or hours) in addition to daily and seasonal changes. At the same time, progressive ocean acidification caused by anthropogenic CO2 emissions is superimposed on these spatial and temporal pH changes. Photosynthetic organisms are therefore unavoidably subject to significant pH variations at the cell surface. Whether this will affect their response to long-term ocean acidification is still unknown, nor is it known whether the short-term sensitivity to pH change is affected by the pCO2 to which the cells are acclimated. We posed the latter open question as our experimental hypothesis: Does acclimation to seawater acidification affect the response of phytoplankton to acute pH variations? The diatom Skeletonema costatum, commonly found in coastal and estuarine waters where short-term acute changes in pH frequently occur, was selected to test the hypo thesis. Diatoms were grown at both 390 (pH 8.2, low CO2; LC) and 1000 (pH 7.9, high CO2; HC) μatm CO2 for at least 20 generations, and photosynthetic responses to short-term and acute changes in pH (between 8.2 and 7.6) were investigated. The effective quantum yield of LC-grown cells decreased by ca. 70% only when exposed to pH 7.6; this was not observed when exposed to pH 7.9 or 8.2. HC-grown cells did not show significant responses in any pH treatment. Non-photochemical quenching showed opposite trends. In general, our results indicate that while LCgrown cells are rather sensitive to acidification, HCgrown cells are relatively unresponsive in terms of photochemical performance.


INTRODUCTION
The concentration of H + in water (and its negative log, the pH) are subjected to dynamic changes that occur over time scales ranging from seconds to seasons.This phenomenon is especially obvious in coastal waters (Hinga 2002, Cornwall et al. 2013).In intertidal pools, pH variations of up to 3 units over 24 h have been observed, mostly due to biological activities (Poole & Raven 1997).In highly productive coastal waters, estuarine waters or coral reef systems, conspicuous diel pH variations of between 0.3 and 0.5 units over 24 h have been documented (Gao et al. 2005, Gray et al. 2012), although the range of these variations is strongly influenced by hydrodynamics and by the rates of exchange with the bulk seawater (Middelboe & Hansen 2007, Unsworth et al. 2012).In ABSTRACT: Variability in pH is a common occurrence in many aquatic environments, due to physical, chemical and biological processes.In coastal waters, lagoons, estuaries and inland waters, pH can change very rapidly (within seconds or hours) in addition to daily and seasonal changes.At the same time, progressive ocean acidification caused by anthropogenic CO 2 emissions is superimposed on these spatial and temporal pH changes.Photosynthetic organisms are therefore unavoidably subject to significant pH variations at the cell surface.Whether this will affect their response to long-term ocean acidification is still unknown, nor is it known whether the short-term sensitivity to pH change is affected by the pCO 2 to which the cells are acclimated.We posed the latter open question as our experimental hypothesis: Does acclimation to seawater acidification affect the response of phytoplankton to acute pH variations?The diatom Skeletonema costatum, commonly found in coastal and estuarine waters where short-term acute changes in pH frequently occur, was selected to test the hypo thesis.Diatoms were grown at both 390 (pH 8.2, low CO 2 ; LC) and 1000 (pH 7.9, high CO 2 ; HC) µatm CO 2 for at least 20 generations, and photosynthetic responses to short-term and acute changes in pH (between 8.2 and 7.6) were investigated.The effective quantum yield of LC-grown cells decreased by ca.70% only when exposed to pH 7.6; this was not observed when exposed to pH 7.9 or 8.2.HC-grown cells did not show significant responses in any pH treatment.Non-photochemical quenching showed opposite trends.In general, our results indicate that while LCgrown cells are rather sensitive to acidification, HCgrown cells are relatively unresponsive in terms of photochemical performance.).Therefore, typical changes linked to ocean acidification (OA) are increased concentrations of pCO 2 , H + and HCO 3 − , decreased concentrations of CO 3 2− and de creases in the CaCO 3 saturation state (Gattuso et al. 2010).These changes will lead to a decrease in the alkalinity of seawater, and the consequent decrease in buffering capacity will increase the magnitude of short-term pH changes in the future.In combination with eutrophication and hypoxia events, anthropogenic acidification may occur earlier and at a more accelerated rate in coastal and estuarine waters than in pelagic environments (Cai et al. 2011).

KEY WORDS: Diatoms
It is known that changes in external pH may have repercussions on cytosolic pH (Raven 2013 and references therein) and cell homeostasis (Montechiaro & Giordano 2010, Giordano 2013, Fanesi et al. 2014). Flynn et al. (2012) showed that the pH at the cell surface of phytoplankton can deviate appreciably from that in bulk seawater, and increasingly so with an increase in metabolic activities and with the thickness of the diffusion boundary layer, which, in the size range of phytoplankton, is a direct function of cell dimension.Thus, in the proximity of the cell, pH can vary across a broad range in a relatively short term depending on the extent of biological (photosynthesis or respiration) and physical (e.g.boundary layer thickness and diffusional gradients) processes.Little is known about how longer-term changes in average bulk water pH may affect organisms already subject to large diel pH variations in their microenvironment.
In coastal waters and estuaries (e.g.Changjiang Estuary, Wang 2002 andJiulong River estuary, Li et al. 2011) where Skeletonema costatum is usually the most abundant algal species, pH can vary by up to 0.5 units d −1 (Hinga 2002, Gao et al. 2005).S. costatum responds rapidly to changes in pH or pCO 2 by down-regulating the activity of periplasmic carbonic anhydrase within 60 to 120 min (Chen & Gao 2003, Wu & Gao 2009).Elevated pCO 2 was shown not to stimulate its growth in the lab (Chen & Gao 2003) or under high solar irradiance (Gao et al. 2012); however, enhanced growth of S. costatum was observed under elevated CO 2 levels in a mesocosm study (Kim et al. 2006).Here, we hypothesize that acclimation to increased CO 2 may decrease the sensitivity of S. costatum to acute pH variations.For this reason, we acclimated S. costatum to growth in media equilibrated with gas phases containing either 390 (present) or 1000 µatm pCO 2 (the level predicted at the end of this century; IPCC 2007); we then ex posed these cultures to acute acidification and measured their photosynthetic responses in terms of Photosystem II (PSII) variable fluorescence.

Culture conditions
The diatom Skeletonema costatum (Grev.)Cl. (CCMBP 112) was pre-cultured at 20°C under a photon flux density (PFD) of 100 µmol photons m −2 s −1 , with a 12 h light:12 h dark photoperiod.The cells were grown in filtered and autoclaved seawater collected in the South China Sea, supplemented with nutrients using f/2 medium (Guillard & Ryther 1962).Culture CO 2 was maintained in equilibrium with gas phases containing either 390 (pH 8.2; LC) or 1000 (pH 7.9; HC) µatm CO 2 in a CO 2 plant growth chamber (HP1000G-D, Wuhan Ruihua In strument and Equipment), with variations < 4%.For all treatments, 3 independent cultures were used as replicates.Prior to the experiments, the cells were grown for at least 20 generations (about 15 d), under either LC or HC; the cultures were diluted every 24 h to maintain cell concentrations between 2 × 10 4 and 1.5 × 10 5 ml −1 .This daily dilution and cell density maintained the cells in mid-exponential growth phase and minimized changes in nutrient availability and seawater carbonate chemistry (see Table S1 in the Supplement at www.int-res.com/articles/ suppl/ b023 p109_ supp.pdf).A pH meter (Benchtop pH 510, Oakton), calibrated daily with standard buffers (NBS, Hanna) was used to measure the pH in the cultures before and after dilution.Other parameters of the seawater carbonate system were calculated using CO2SYS software (Table S1; Lewis & Wallace 1998), taking into account salinity (35 psu), total dissolved inorganic carbon (DIC), pH, nutrient concentrations and temperature (20°C); DIC was determined with a DIC analyzer (AS-C3, Apollo SciTech).The equilibrium constants K 1 and K 2 for carbonic acid dissociation and K B for boric acid were determined according to Roy et al. (1993) and Dickson (1990), respectively.
Cells were harvested by filtration on cellulose membranes (Xinya), washed with 10 ml of growth medium and then resuspended in 100 ml of medium in which the pH had been adjusted by the addition of appropriate amounts of HCl to obtain the 3 experimental pH levels of 7.6, 7.9 and 8.2.These cell suspensions were transferred to sterile 100 ml lighttransparent polyethylene syringes.The syringes were maintained under the same conditions used for the pre-cultures.We opted to use syringes because they allowed for the rapid collection of samples and the maintenance of a negligible air space (only the needle tip of the syringes), and thus minimal CO 2 leakage.

Variable fluorescence measurements
At time zero (t 0 ), a 2 ml aliquot was collected from the syringe and incubated in the dark for 15 min in a quartz tube and then directly placed in a Xe-PAM (Walz) holder for variable fluorescence measurements.Subsequently, 2 ml aliquots were transferred to quartz tubes after 2 (t 2 ), 4 (t 4 ), 6 (t 6 ), 15 (t 15 ), 30 (t 30 ), 45 (t 45 ) and 60 (t 60 ) min since the transfer into the syringes, and subjected to the measurements without prior dark adaptations.Samples were also collected after 15, 30, 45 and 60 min, and dark-adapted for 15 min to obtain maximum quantum yield of fluorescence (F v /F m ).This parameter was derived from the fluorescence induction curves (Dorigo & Leboulanger 2001) and obtained according to the following equation (Genty et al. 1989): where F m is the maximum fluorescence measured after a saturating flash of 5000 µmol photons m −2 s −1 , and F o is the basal fluorescence of PSII with measuring light of 0.1 µmol photons m −2 s −1 .The light-adapted cells were used to determine the effective quantum yield of fluorescence (ΔF/F m ').This parameter was also derived from the induction curves, but in this case the cells were exposed to actinic light with an irradiance similar to that used for growth (100 µmol photons m −2 s −1 ).The induction curve was conducted as follows: the first saturating pulse (5000 µmol photons m −2 s −1 ) was provided after irradiating the cells with measuring light for 10 s; the actinic light was turned on 10 s after the first saturating pulse, then after a further 10 s and at 10 s intervals, for a sequence of 10 saturating pulses.The ΔF/F m ' values were obtained from the last 1 or 2 sat-urating pulses when the F m ' was steady by application of the following equation (Genty et al. 1989): where F m ' is the instant maximum fluorescence and F t is the steady-state fluorescence under the actinic light.
Non-photochemical quenching (NPQ), which provides an indication of regulated thermal energy dissipation, was estimated using the following equation The relative electron transport rate (rETR) was obtained from the rapid light curves (White & Critchley 1999), which were measured under the same experimental conditions used for the induction curve and calculated as (Genty et al. 1989): where E PAR is the irradiance of the actinic light and 0.75 is a factor used to take into account that, in diatoms, approximately 75% of the photons reach PSII (Johnsen & Sakshaug 2007).For the rapid light curves, samples were subject to a series of saturating pluses (5000 µmol photons m −2 s −1 ) in the presence of incremental actinic light PFD (0, 125, 185, 285, 410, 600, 840, 1200 and 1650 µmol photons m −2 s −1 ) at 10 s intervals.
Other photosynthetic parameters were obtained from the rapid light curves, which were fitted by the mathematical model of Jassby & Platt (1976): rETR = rETR max × tanh(α × E / rETR max ) (5) where E is the photon flux density (µmol m -2 s -1 ), α the initial slope of the rapid light curve and E k is the photon flux density at which the onset of photosynthesis light saturation occurs.

Statistics
The results are presented as means ± SD of measurements, conducted on samples collected from 3 independent cultures (n = 3).For the rapid light curves, the Jassby & Platt (1976) model was fitted to individual curves, and averages and standard deviations of the photosynthetic parameters were then calculated from 3 independent cultures (n = 3).The statistical significance of differences was assessed by a 3-way repeated measured ANOVA, using pH (df = 2), CO 2 (df = 1) and time (df = 7 for non-dark adapted cells or df = 4 for dark-adapted cells) as variables, with a Tukey's post-hoc test for comparison.Differences were considered significant at p < 0.05.

Quantum yield and NPQ
In the LC-acclimated cells, ΔF/F m ' of PSII showed an initial increment for the first 5 min when the cells were exposed to acute acidification (Fig. 1A).When the cells were incubated at pH 7.6, after the initial increase ΔF/F m ' decreased rapidly for about 30 min (p = 0.001), then plateaued at about 34% of the maximum value.When the cells were incubated at pH 7.9 and 8.2, ΔF/F m ' did not vary significantly (p = 0.612).In the HC-acclimated cells, the ΔF/F m ' value also underwent a slight increase at the beginning, but then reached a steady value and did not change, even in the lowest pH treatment (p = 0.981; Fig. 1B).F v /F m showed a similar change pattern, except that the drop in pH to 7.6 led to less of a decline in the dark-adapted cells but more reduction in the lightadapted ones in LC-acclimated cells (p < 0.001; Fig. 1A,C).If the variations were expressed as a function of the total time at a given pH, the rate of yield decrease was smaller in dark-adapted (0.00475 min −1 ) than in light-adapted cells (0.0056 min −1 ).
NPQ values in the LC-grown cells exposed to pH 8.2 and 7.9 (Fig. 2A) remained low (always below 0.4), with no statistical difference between the values measured at these 2 pH levels (p = 1.000).At pH 7.6, however, NPQ increased from ca. 0.4 to 3.4 over 60 min (p = 0.003).Interestingly, unlike ΔF/F m ', NPQ did not plateau.In the HC-acclimated cells, NPQ did not change as obviously as in the LC-grown cultures, even when the cells were subjected to pH 7.6 ) high (1000 µatm; HC) CO 2 levels and grown for 1 h at pH treatment levels of 7.6, 7.9 or 8.2.
Error bars: SD of 3 independent measurements (n = 3 cultures) (Fig. 2B).Similar trends were observed for the NPQ of the cells that were dark-adapted for 15 min, after having been in the light for 15, 30, 45 or 60 min (Fig. 2C).In the HC-grown cells, however, the drop of pH to 7.6 did enhance the NPQ by 53.8 to 93.8% (p = 0.001) in the light-acclimated cells, and by 51.2 to 135.5% (p = 0.001) in the dark-adapted cells compared to the pH 8.2 treatment.A comparison of ΔF/F m ' of LC-and HC-grown cells indicated that the yield of the former was always lower than that of the latter (p < 0.001).At pH 7.9 and 8.2, ΔF/F m ' was about 20% lower in LC-grown cells and this difference did not vary over time; at pH 7.6, the difference between the ΔF/F m ' of LC-and HCgrown cells was much higher, and after 60 min the value for HC-grown cells was 285% of that of their LC counterparts (see Fig. S1 in the Supplement at www.int-res.com/articles/ suppl/ b023 p109_ supp.pdf).At pH 7.9 and 8.2, NPQ was higher in HC-than in LC-grown cells by about 130% (p < 0.001).Only at pH 7.6 was the NPQ of LC-grown cells much higher (almost by one order of magnitude) than that of cells acclimated to HC (Fig. S2 in the Supplement).

Light response curves
The rETR did not show major differences as a function of time of sampling and pH treatment, but showed that dark-adapted cells (included those at t 0 in Fig. 3) tended to have lower rETR at sat urating light than light-adapted cells (p < 0.001; Figs. 3 & 4).HC-grown cells always had higher rETR max than LCgrown cells (p < 0.001; Fig. S3 in the Supplement).α, which represents the apparent light-use efficiency for the rETR, was 60 to 79% higher in the HC-than in the LC-grown cells (Fig. 5, Fig. S3C).In the LCgrown cells, the α value was fairly constant within a ) high (1000 µatm; HC) CO 2 levels, grown for 1 h at pH treatment levels of 7.6, 7.9 or 8.2.Error bars: SD of 3 independent measurements (n = 3 cultures); NPQ(d) is the NPQ value of dark-adapted cells 60 min period (only a modest decline was observed at pH 7.9).In the HC-grown cells, however, α increased over the first 15 min and then remained constant (Fig. 5).At the beginning of the pH treatments (t 0 ), higher light levels were required to saturate rETR in HC-than in LC-grown cells; this difference was not observed at subsequent times (Fig. S3A,B).

DISCUSSION
The main finding of this work is that Skeletonema costatum cells are well equipped to maintain photochemical performance in response to acute pH changes when acclimated to growth at CO 2 concentrations and pH values on the order of those expected for the next 100 yr (i.e. in the HC treatment).Acute exposures to acidification at pH 7.6 had different im-pacts on the cells that were acclimated to present conditions (LC) than from those acclimated to elevated CO 2 (HC) concentrations, with the former being much more susceptible to changes in their quantum yield (which decreased upon acidification) and NPQ (which increased upon acidification).The HC-acclimated cells also showed higher light-use efficiency for rETR.
S. costatum has been shown not to express periplasmic (external) carbonic anhydrase when the medium CO 2 concentration was raised to the HC levels used for this study (Chen & Gao 2003).This is usually considered as a line of evidence that energydependent CO 2 acquisition is turned off or strongly down-regulated (Giordano et al. 2005).Under these conditions, diffusional influx of CO 2 into the cells may become relevant, provided that cellular consumption takes the internal CO 2 concentration below the ambient concentration (Raven et al. 2014)

E F
Fig. 3. Relative electron transport rate (rETR) from rapid light curves of Skeletonema costatum cells acclimated to (A−C) low (390 µatm; LC) or (D−F) high (1000 µatm; HC) CO 2 levels, grown for 1 h at pH treatment levels of 7.6, 7.9 or 8.2 and measured at time points 0 (t 0 ), 15 (t 15 ), 30 (t 30 ) and 60 (t 60 ) min.Error bars: SD of 3 independent measurements (n = 3 cultures) diffusion is a direct function of the concentration gradient across the membrane, according to Fick's first law (Fick 1855).Also, the absolute CO 2 concentration in the medium under the conditions in which photosynthesis prevails over respiration cannot be higher than that allowed by the equilibrium between the gaseous and liquid phase, according to Henry's law (Henry 1803).Therefore, assuming that the intracellular pH is relatively constant and that the CO 2 concentration within the cell is not at equilibrium with the external environment due to positive net photosynthetic CO 2 consumption, the acidification of the external medium, in concomitance with a higher pCO 2 in the gas phase, would favor a diffusional flux of CO 2 into the cell.The same would not occur in the LC-grown cells, where the presence of an active CO 2 concentrating mechanism (CCM; Giordano et al. 2005) makes diffusional CO 2 influx un likely (since the internal CO 2 concentration is higher than that in the medium).
Therefore, HC-grown cells may benefit more from acidification than LC-grown cells.The down-regulation of the CCM due to elevated (but not saturating) CO 2 is also likely to enhance photorespiration.The enlargement of the sink for electrons constituted by photorespiration (Raven et al. 2014) may contribute to the explanation of the higher effective quantum yield and light-use efficiency in HC-grown cells (Figs. 1, 5 & S1), and the higher NPQ of LC-grown cells (Fig. 2) at the lower pH (7.6) treatment.This result implies that HC-acclimated cells can reduce their sensitivity to acute pH stress.Increased NPQ with time in the LC-grown cells when exposed to pH 7.6 (Fig. 2A,C) reflects enhancement of q E (energy-dependent quenching, a key component of NPQ) with down-regulation of CCMs in this species.A drop of pH by 0.  We are well aware of the limits of our relatively simple and short-term experiment; nevertheless, our results indicate that diel pH variations in coastal waters, as reflected in the acute pH drop treatment, can alter the photosynthetic performance of S. costatum, and that acclimation to elevated pCO 2 may decrease the impact of ocean acidification on this diatom.The fact that S. costatum is a common inhabitant of coastal and estuarine waters (Wang 2002, Li et al. 2011) can at least partially be attributed to its capability to cope with rapid pH changes in these environments.While coastal and estuary waters are predicted to be acidified faster through anthropogenic activities (Sunda & Cai 2012), we might predict that S. costatum would still cope well with these changes in coastal and estuary habits based on our experimental data.Certainly, multifactorial experiments, comprising concomitant and realistic variations of pCO 2 , pH and warming are required to further clarify the complex interplay among the dif ferent facets and consequences of an increase in pCO 2 .
With progressive changes such as OA and warming, the abundance of diatoms may increase or decrease in different latitudes or different waters of constant or fluctuating seawater carbonate chemistry.While diatom abundance in the northeast Atlantic and North Sea increased during a half-century time period (Hinder et al. 2012), global phytoplankton biomass in the world's oceans has been shown to decline (Boyce et al. 2010).While there is still a controversy over whether diatoms will benefit from or be harmed by OA (as reviewed by Gao & Campbell 2014), beneficial or inhibitory effects of OA on some diatoms appear to depend on levels of light exposures (Gao et al. 2012).In coastal waters, phytoplankton cells usually experience dramatic changes in pH/pCO 2 due to biological production and consumption, and exposure to a low pH/high CO 2 milieu can affect their physiological performance when the cells move (e.g.some dinoflagellates) or are mixed in with waters of different acidities.Al though the present study showed the tolerance of S. costatum to acidic stress (especially in those acclimated to high CO 2 conditions), a mechanistic understanding of different diatoms' responses to acidic stress is essential in order to predict the general responsive patterns of this group to ocean changes.

Fig. 4 .Fig. 5 .
Fig. 4. Relative electron transport rate (rETR) in the Skeletonema costatum cells that had been dark-adapted for 15 min('d').Other information regarding symbols and panels is the same as in Fig.3