Interactive Effects of Ocean Acidification and Nitrogen-Limitation on the Diatom Phaeodactylum tricornutum

Climate change is expected to bring about alterations in the marine physical and chemical environment that will induce changes in the concentration of dissolved CO2 and in nutrient availability. These in turn are expected to affect the physiological performance of phytoplankton. In order to learn how phytoplankton respond to the predicted scenario of increased CO2 and decreased nitrogen in the surface mixed layer, we investigated the diatom Phaeodactylum tricornutum as a model organism. The cells were cultured in both low CO2 (390 μatm) and high CO2 (1000 μatm) conditions at limiting (10 μmol L−1) or enriched (110 μmol L−1) nitrate concentrations. Our study shows that nitrogen limitation resulted in significant decreases in cell size, pigmentation, growth rate and effective quantum yield of Phaeodactylum tricornutum, but these parameters were not affected by enhanced dissolved CO2 and lowered pH. However, increased CO2 concentration induced higher rETRmax and higher dark respiration rates and decreased the CO2 or dissolved inorganic carbon (DIC) affinity for electron transfer (shown by higher values for K1/2 DIC or K1/2 CO2). Furthermore, the elemental stoichiometry (carbon to nitrogen ratio) was raised under high CO2 conditions in both nitrogen limited and nitrogen replete conditions, with the ratio in the high CO2 and low nitrate grown cells being higher by 45% compared to that in the low CO2 and nitrate replete grown ones. Our results suggest that while nitrogen limitation had a greater effect than ocean acidification, the combined effects of both factors could act synergistically to affect marine diatoms and related biogeochemical cycles in future oceans.


Introduction
Rising atmospheric CO 2 concentrations enhance its absorption into the world's oceans, which currently accounts for removal of nearly one third of anthropogenic CO 2 emissions from the atmosphere [1]. Atmospheric CO 2 concentrations are expected to reach 800-1000 ppmv by the end of this century according to the ''business as usual'' CO 2 emission scenario [2]. Dissolution of CO 2 into seawater has already induced a global drop in pH of 0.1 units since the end of the Industrial Revolution, and values are expected to drop another 0.3-0.4 units by the end of this century. This decline in pH driven by increased CO 2 is termed ocean acidification (OA) [3]. The decrease in seawater pH is a consequence of changes in marine chemistry, where increased dissolved CO 2 leads to increases in H 2 CO 3 and hence to increases in H + and HCO 3 2 concentrations and decreased CO 3 22 concentration and CaCO 3 saturation state. Changes in pH also affect biogeochemical processes such as alterations to trace metal speciation, which can have significant biological effects [4], [5].
If the photosynthesis of marine eukaryotic phytoplankton were supported solely by the diffusional supply of CO 2 to the active site of the CO 2 fixing enzyme Rubisco, then this process would be severely limited at the concentrations of CO 2 currently found in seawater (,10-30 mmol) [6]. However, most algae have been shown to be able to make extremely efficient use of low levels of dissolved inorganic carbon (DIC) by virtue of inducible carbon concentrating mechanisms (CCMs) [7]. The CCMs act to maintain internal CO 2 concentrations higher than can be accounted for by diffusion-mediated entry of inorganic carbon.
CCM activity is down-regulated under high CO 2 ; enhanced CO 2 availability could thus reduce the energy cost for CO 2 transport [8], and the re-allocation of energy may play a critical role in modulating primary production as well as elemental stoichiometry and species composition [6]. This, however, also may depend on other environmental factors. The effects of ocean acidification can have positive, neutral or negative aspects depending on the physiological processes involved, and may be species-specific [9]. While increased primary production under high CO 2 has been found in many studies [10], energy loss due to enhanced respiration has also been reported under high CO 2 /low pH conditions [11], possibly due to enhanced energy demand associated with the need to maintain intracellular acid-base stability [12]. The effects of ocean acidification are controversial with contradictory trends reported in the literature. While some of this could be due to species-specific responses it could also result from interactive effects with other environmental factors [13], [14].
Nutrient availability is well known to affect algal growth and production. This is especially so for nitrogen availability, which is seen in many cases as a major limiting factor for algal growth in the oceans [15]. Marine phytoplankton may experience increased nutrient limitation in the euphotic layer in the future due to intensified stratification in a warming ocean [16], [17]. Ocean acidification, at the same time, may affect ion and nutrient assimilation of algae either directly by altering proton or ion channels or indirectly by changes in chemical speciation and nutrient availability [18], [19]. Thus, ongoing ocean acidification together with intensified stratification could further decrease marine nutrient availability and uptake rates. Decreased nitrogen availability is expected to lead to decreased synthesis of chlorophyll and proteins in algae, which would have a strong influence on photosynthesis and physiological performance. Nitrogen limitation is known to affect carbon fixation because of potential impacts on levels of Rubisco and other proteins and also because nitrate assimilation is energy dependent and will compete with carbon fixation for ATP and reductant [20]. Thus changes in C and N acquisition may be reflected in altered cell carbon and nitrogen contents [21].
While the impacts of nutrient limitation under present day CO 2 are well understood and there is an increasing literature on the effects of elevated CO 2 on phytoplankton physiology and ecology (see reviews by Beardall et al., Riebesell and Tortell and references therein) [10], [22] and elemental ratios [23], most studies on the effects of ocean acidification have been carried out under nutrient replete conditions [11], [24] and there is very little information on interactive effects between nutrient limitation and elevated CO 2 [10]. Since elevated CO 2 and ocean acidification in a future world is likely to go hand-in-hand with a more restricted nutrient supply in the low-to mid-latitude open ocean, this is an important issue that needs addressing.
Thus, this paper considers how phytoplankton responses to ocean acidification may be affected by nitrogen limitation. Specifically, we have measured the cell size, growth, pigmentation, quantum yield, respiration, and CCM activity as well as cell carbon and nitrogen contents of Phaeodactylum tricornutum grown under high CO 2 and low nitrogen conditions, to determine the possible interactive effects of ocean acidification and nitrogen limitation on this model diatom species.

Statement of ethics
The strain of the diatom Phaeodactylum tricornutum Bohlin (strain CCMA 106), originally isolated from the oligotrophic waters of the South China Sea in 2004, was obtained from the Center for Collections of Marine Bacteria and Phytoplankton (CCMBP) of the State Key Laboratory of Marine Environmental Sciences (Xiamen University). No specific permits were required for using this species.

Algal culture conditions
The diatom Phaeodactylum tricornutum Bohlin (strain CCMA 106) were grown in artificial seawater with Aquil medium enrichment [25] except that the NO 3 2 concentration was adjusted to 110 mmol L 21 NO 3 2 (HN) or 10 mmol L 21 NO 3 2 (LN). The nitrogen-limiting level of 10 mM was based on the surface inorganic nitrogen concentrations (unpublished) obtained from the oligotrophic South China Sea, ranging from 0 (undetectable) to 20 mM. Cultures were continuously aerated with ambient air of 390 matm of CO 2 (LC) or with high CO 2 of 1000 matm (HC) within plant CO 2 chambers (HP1000G-D, Ruihua Instrument & Equipment Co. Ltd, China) and bubbled at a constant flow rate of 300 ml min 21 . This allowed the following treatments, combining different N and CO 2 levels, to be performed: LC-HN, LC-LN, HC-HN, HC-LN. The cells were grown semi-continuously at 20 u C under 70 mmol photons m 22 s 21 illumination with a 12L: 12D photoperiod. Dilutions were carried out every 24 h to ensure cell concentrations did not exceed 3610 5 cells ml 21 at their exponential growth phase so that pH change during growth at each CO 2 level was less than 0.02 (Table 1). Cells were acclimated to each NO 3 2 and CO 2 combination for more than 10 generations before being used in the experiments described below.
The pH of cultures was measured daily, prior to dilution, with a pH probe (Mettler Toledo DL15 Titrator, Sweden), which was calibrated with standard NBS (National Bureau of Standards) buffer solutions (Hanna) at three pH points (pH 10.01, pH 7.01 and pH 4.01). Measurement of dissolved inorganic carbon (DIC) was carried out using an automated system (AS-C3, Apollo Scitech), which was connected to an infrared gas detector (Li-Cor 7000, Li-Cor). Calculation of the carbonate system components (HCO 3 2 , CO 3 22 , CO 2 and TA) was carried out using known values of DIC, pH, nutrient concentration, salinity and temperature with a CO 2 system analysis software (CO2SYS) [26] ( Table 1). Carbonic acid dissociation constants (K 1 and K 2 ) were according to Roy et al. [27], and that for boric acid (K B ) was taken from Dickson [28].

Growth rate and cell size measurements
Cell numbers, mean cell volumes and size distributions were acquired with a Z2 TM Coulter Counter (Beckman, Buckinghamshire, UK). Determinations of growth rates were based on the cell number changes every 24 h and were calculated according to the equation: m = (lnN 1 -lnN 0 )/(t 1 -t 0 ), where N 1 and N 0 are the cell concentrations before dilution (t 1 ) and after the previous dilution (t 0 ) respectively. Growth rates were calculated based on measurements of 11-12 replicates for triplicate cultures under each CO 2 level.

Carotenoid and Chlorophyll measurements
To determine the carotenoid and chlorophyll a and c contents of cells cultured in the different CO 2 and NO 3 2 conditions, cells were collected by filtration on to Whatman GF/F filters (pore size, 0.22 mm) and extracted overnight with absolute methanol at 4uC. The extracts were then centrifuged for 10 min at 50006g and the absorbance of the supernatant was scanned with a spectrophotometer (DU800, Beckman, Fullerton, California, USA). Calculation of chlorophyll a from the absorbance spectra followed the equation of Porra [29], chlorophyll c was after Ritchie [30] and carotenoid was calculated according to Strickland and Parsons [31].

Quantum yield measurements
The quantum yield of cells grown in the different CO 2 and NO 3 2 conditions was measured with a XE-PAM (Walz, Germany) at both mid-light phase (F v '/F m ') and at the end of the dark phase (F v /F m ). The saturation light was set at 5000 mmol photons m 22 s 21 for 0.8 s.

Determination of CCM activity from rETR vs DIC curves
To estimate the affinity of cells for DIC (used as a proxy for CCM activity), cells cultured in different CO 2 and NO 3 2 conditions were collected, washed with, and re-suspended into, DIC-free seawater with a pH of 8.20. Cell densities after resuspension were between 3 and 4610 4 cells ml 21 . The DIC-free seawater was prepared by adding 1 mol L 21 HCl to drop the pH below 3 and then bubbling with pure N 2 for 1 h. Tris-buffer was added to 20 mmol L 21 to adjust the pH back to 8.20. Cells suspended in the DIC-free seawater were incubated at 150 mmol m 22 s 21 for 15 min to exhaust any intracellular DIC, and NaHCO 3 solution was then added into each vial of algal suspension to obtain different DIC concentrations. After further incubation under a photon flux of 70 mmol m 22 s 21 for 10 min (less than 0.1% DIC was consumed), a rapid light curve was determined with the XE-PAM, and the resulting data fitted with the equation of Eilers and Peeters [32]: y = x/(ax 2 +bx+c), where a, b, c are estimated parameters, x photon flux density, and y the rETR value. Light saturated rates of electron transport (rETR max ) at the different DIC concentrations can be calculated from the fitted rapid light curve: rETR max = 1/[b+2(ac) 1/2 ], and the light harvesting efficiency (a) was calculated with the equation: a = 1/c. To quantify the relationship between rETR max and DIC concentrations, we fitted the two parameters using the Michaelis-Menten equation to determine light-and DIC-saturated rates of photosynthesis and the half-saturation constant K 1/2 DIC and K 1/2 CO2 for DIC-dependent electron transport.

Dark respiration measurements
Cells were gently filtered on to polycarbonate membrane filters (0.22 mm, Q/YY8-1-88, Xinya, China) with a vacuum pump at a pressure of less than 0.02 Pa. to ensure cells were intact, based on a microscopic checkup, and were then re-suspended into 20 mmol L 21 Tris-buffered media of the respective composition (LC-HN, LC-LN, HC-HN and HC-LN). Each treatment had a known cell concentration of around 1610 6 cells ml 21 . Dark respiration rates were determined with a Clark type oxygen electrode (5300A, YSI) from changes in oxygen concentration over time at 20uC. A twopoint calibration (seawater bubbled with air until equilibrium saturation and O 2 deprivation with excess sodium sulfite as zero oxygen) was carried out before respiration measurements. The possible contribution of bacterial respiration was tested on the culture filtrate that passed through a filter pore size of 1mm (mixed cellulose lipid membranes), which would not exclude the few bacteria present, and there was no detectable bacterial O 2 consumption.

Measurement of carbon and nitrogen contents
To determine particulate organic carbon (POC) and nitrogen (PON) in Phaeodactylum tricornutum grown under different CO 2 and NO 3 2 concentrations, cells were collected in the mid-light period by filtration onto pre-combusted (450uC, 6 h) GF/F filters (Whatman). Filters were acidified with 0.1N HCl fumes for 12 h and then dried overnight in an oven at 60uC. Carbon and nitrogen contents were determined with a PerkinElmer Series II CHNS/O Analyzer 2400.

Statistical analysis
One-way analysis of variance (ANOVA) and Tukey's test were used to establish differences among treatments at a confidence level of 95%. Interactive effects between CO 2 and NO 3 2 were analyzed using a Tukey post hoc test.

Growth rate
Growth rates were inhibited under nitrogen limited conditions under both LC (30%, P,0.001) and HC (38%, P,0.001) conditions ( Figure 1). No direct effects on growth rate were found between the CO 2 treatments (nitrogen replete, P = 0.24; nitrogen limited, P = 0.72). No interactive effect was found between CO 2 and NO 3 2 levels (P = 0.3).

Chlorophyll and carotenoid contents
The effects of nitrogen and CO 2 treatments on cellular chlorophyll and carotenoid concentrations showed the same trend as cell size and growth rate. Nitrogen limitation decreased the carotenoid, chlorophyll a and c contents by 50% and 62% (Figure 2A), 48% and 60% ( Figure 2B), 45% and 63% ( Figure 2C) in the LC and HC groups, respectively (P,0.05). No direct effects  on pigmentation were found between the CO 2 treatments (P.0.05). The ratio of chlorophyll a to carotenoids showed no significant difference among the treatments (P.0.05) ( Figure 2D). No interactive effects on chlorophyll a (P = 0.086), c (P = 0.133), carotenoid (P = 0.475) and the ratio of chlorophyll a to carotenoids (P = 0.657) were found between CO 2 and NO 3 2 levels.

Quantum yield
Maximum quantum yields (F v /F m ), measured at the end of the dark period, showed no significant differences among treatments (P.0.05) ( Figure 3A). However, in the mid-light period, cells cultured under nitrogen limited conditions showed decreases in effective quantum yield (F v '/F m ') by 12% and 15% in the LC (P = 0.01) and HC (P = 0.002) treatments respectively, compared to those under nitrogen replete conditions ( Figure 3B). No significant (nitrogen replete, P = 0.43; nitrogen limited, P = 0.91) change in the yield was found between the low and high CO 2 levels ( Figure 3A, B). No interactive effect on yield was found between CO 2 and NO 3 2 levels (F v /F m , P = 0.24; F v '/F m ', P = 0.58).

Cell size
Increased dissolved CO 2 concentration did not affect the mean cell size (as Effective Spherical Diameter) and cell volume of Phaeodactylum tricornutum under nitrogen replete conditions. However, nitrogen limitation did cause significant (P,0.001) decreases in cell size and cell volume (Figure 4, 5). Mean cell volumes were 65.3 and 67.6 mm 3 in LC and HC treatments, and nitrogen limitation significantly decreased these values to 43.0 and 41.3 mm 3 (by 34% and 39%), respectively (P,0.001) ( Figure 5A). Mean cell sizes (as Effective Spherical Diameter) in LC and HC were 4.9160.08 and 4.9560.00 mm respectively, under nitrogen replete conditions ( Figure 5B). Under nitrogen limitation, cell size decreased by 13% and 15% (P,0.001), to 4.2560.02 mm and 4.1960.01 mm in LC and HC cells, respectively ( Figure 5B). No significant differences in both cell size (P = 0.56) and volume (P = 0.39) were found between low and high CO 2 levels under the nitrogen replete conditions. However, high CO 2 cells showed a small but significant decrease of both cell size (P = 0.02) and volume (P = 0.04) when nitrogen was limited. No interactive effects were found between CO 2 and NO 3 2 levels in both cell size (P = 0.16) and cell volume (P = 0.14).

P vs DIC characteristics
With increasing DIC concentration in the medium, high CO 2 , N-replete cultures had a significantly (p = 0.04) elevated rETR max (light and DIC-saturated rate of electron transfer) compared to low  Figure 6C). The K 1/2 CO2 shows the same trend as K 1/2 DIC and values were 0.34, 0.66, 0.62 and 0.81 mmol L 21 in the LC-HN, LC-LN, HC-HN and HC-LN treatments respectively. The light harvesting efficiency (a) of the cells, derived from RLC at different DIC levels, was not affected (P = 0.12) by CO 2 levels when nitrogen was limited, but was significantly (P = 0.01) elevated by the CO 2 enrichment in Nreplete conditions at a DIC level of 138 mmol L 21 . Such a trend was still observed at 275 mmol L 21 , although with the differences being statistically insignificant (nitrogen replete, P = 0.35; nitrogen limited, P = 0.32) (Figure 7). At the ambient DIC level of 2200 mmol L 21 , no significant difference (nitrogen replete, P = 0.22; nitrogen limited, P = 0.23) in light harvesting efficiency was found among the treatments (Figure 7).

Dark respiration rates
Cells grown under nitrogen limitation, irrespective of CO 2 level, showed approximately a doubling in respiration rate, compared to ambient CO 2 , nitrogen replete cells. With N-replete cells, high CO 2 also resulted in an increase in cellular respiration rates ( Figure 7A). When dark respiration was expressed on a per chlorophyll a basis, rates were enhanced by 298%, 110% and 305% in LC-LN, HC-HN, HC-LN treatments respectively, compared to the LC-HN conditions ( Figure 7B). No interactive effects on dark respiration were found between CO 2 and NO 3 2 levels (P = 0.16).

Carbon and nitrogen contents
Nitrogen limitation significantly decreased the nitrogen content of cells cultured in both LC (by ca. 32%) and HC (by ca. 28%) conditions compared with the nitrogen-replete treatments (P,0.001) ( Table 2). Nitrogen limitation led to an increase in the C:N ratio by 21% in the LC condition, and this enhancement increased to 45% in the HC condition (P,0.001). However, under the nitrogen replete treatment the CO 2 concentration did not affect the C:N ratio (P = 0.24), even though the cells at the high CO 2 level significantly increased their nitrogen content by 13% (P = 0.04). Turkey's post hoc test showed that there were significant interactive effects on C:N found between CO 2 and NO 3 2 levels (P,0.001).

Discussion
While the elevated CO 2 concentration of 1000 matm did not cause significant differences in growth, pigment contents, effective quantum yield and cell size, nitrogen limitation decreased all these parameters in the diatom Phaeodactylum tricornutum (Figures 1, 2, 3,  4). Both the elevation of CO 2 and N-limitation led to a down-regulation of CO 2 concentrating mechanism (CCM) activity, as reflected in the increased/decreased K 1/2 DIC or K 1/2 CO2 . Nlimitation and increased pCO 2 /reduced pH led to the lowest light use efficiency under Ci-limited conditions (Figure 7), with this trend effect being minimized under elevated levels of DIC. Ocean acidification increased dark respiration under N-limited conditions   ( Figure 8). The high CO 2 and low nitrate combination altered the cells' elemental stoichiometry, with the C:N ratio increased by 45% compared to the low CO 2 and nitrate replete grown cells ( Table 2).

Basic cell parameters
Confirming previous studies on growth of diatoms, including P. tricornutum, under elevated CO 2 [33], [34], growth at the elevated CO 2 levels that are expected by the end of the century did not cause a significant increase in growth rate. This was the case regardless whether the organism was grown under N depletion or N repleted conditions. Earlier studies in our laboratory on the same species showed an enhanced growth rate (ca. 5%) under elevated CO 2 and a PAR of 120 mmol m 22 s 21 [11]. In the current study we did not find improved growth under a PAR level of 70 mmol m 22 s 21 . Recently, changes in light levels have recently been shown to mediate diatoms' responses to ocean acidification [14]. Even although both light levels are subsaturating for photosynthesis, less photosynthetic carbon fixation under 70 mmol m 22 s 21 should have resulted in less or no growth stimulation due to the enhanced respiratory carbon loss (Figure 8).
Although growth rates were clearly limited by a decrease in Nsupply and, as has commonly been reported [35], chlorophyll levels were significantly decreased in N-limited cells, there was no change in the chl:carotenoid ratio, sometimes used as an indicator for N-limitation [35], under our experimental conditions (Figure 2). A lack of effect of CO 2 levels on cellular pigment content is consistent with previous data on diatoms [11], [36] and other microalgae [37]. Furthermore, dark-adapted maximal quantum yield showed no decrease in N-limited cells, although the effective quantum yield show a small, but significant drop; neither parameter was affected by growth at elevated CO 2 (Figure 3). Previous studies on a green alga Ulva rigida showed that both F v / F m and F v '/F m ' were significantly down-regulated by high CO 2 , and that nitrogen limitation further decreased both parameters [38]. A decline in F v /F m is a general response to nitrogen limitation [39]. There is a possibility that the urea cycle found in Phaeodactylum functions to support the photosynthetic machinery by recycling N under the N-limited conditions [40]. In the light period, in contrast, carbon fixation and nitrogen acquisition are both competing for energy [20], thereby leading to lower yield under the N-limited conditions.
Cell size was significantly smaller by ca. 15% in N-limited cells, a phenomenon that has been reported for dinoflagellates and for the coccolithophore, Emiliania huxleyi [41], [42], [43]. Considering the lower specific growth rate and increased C:N ratio under future high CO 2 and low nutrient conditions (Table 2), the diatoms' carbon fixation efficiency per unit nitrogen will increase in the surface seawater.

Photosynthesis vs DIC characteristics
Elevated CO 2 caused a slight increase in photosynthetic capacity (measured as rETR max under light and DIC-saturated conditions) under N-replete, but not under N-limited conditions ( Figure 6). Small increases in the DIC-saturated photosynthetic capacity of diatoms grown under elevated CO 2 have been reported previously [11], [36], [44]. Growth under N-limited conditions caused an increase in K 1/2 DIC in low-CO 2 grown cells.   This indicates a down-regulation of CCM activity by nitrogen limitation. This is in contrast to the reports of Young and Beardall for the green alga Dunaliaella tertiolecta showing cells grown under low CO 2 exhibited increasing affinity (lower K 1/2 DIC ) for DIC under N-limited growth and work on Chlorella emersonii in which cells under 5% CO 2 showed partial induction of CCM activity when N-limited [45], [46]. Such responses have been viewed as a way of improving N-use efficiency and maintaining Rubisco activity with less Rubisco protein when resources such as nitrogen are in short supply [45]. However, induction of CCM activity under N-limitation does not always occur, especially when cells are grown under low CO 2 or are not severely N-limited (see Table 1) [39], [47]. In P. tricornutum, nitrogen storage strategy using the urea cycle could mediate the diatom's CCM to decrease the influence of elevated CO 2 levels [40].
The changes in rETR max under elevated CO 2 and N-replete conditions may be partially negated by the rise in dark respiration and contribute to the lack of effect of high CO 2 on growth rate. However, low CO 2 grown cells that were N-limited also exhibited higher respiration rates, contributing to the lower growth rates found in N-limited cells. Growth at elevated CO 2 exposes cells to a lower pH, which might impose additional energetic costs for acidbase regulation to sustain metabolic integrity [48]. Metabolic processes also influence the pH in the immediate proximity to the cell surface [49], and under low N levels (here supplied as nitrate), the near cell pH would become less alkaline (NO 3 2 uptake leads to OH 2 extrusion), again leading to increased energy demand to maintain pH gradients across the cell membrane.

Elemental composition
The elemental composition and macromolecular composition of phytoplankton is critically important for secondary producers such as copepods, fish and shrimp, and food nutritional quality influences energy flow through marine food chains [50]. Recently Rossoll et al. found that the fatty acid composition of the diatom Thalassiosira pseudonana cultured at elevated CO 2 was altered and that this significantly affected the growth and egg production of a copepod, Acartia tonsa [51]. Riebesell et al. reported an increased C:N ratio in a mesocosm study dominated by diatoms [52], whereas Burkhardt et al. reported both increases and decreases in C:N ratio with increasing CO 2 , dependent on the species [23]. In our study, nitrogen limitation decreased both the carbon and nitrogen contents per cell but these changes were not parallel and led to an enhanced C:N ratio in both LC and HC conditions. Cells cultured under high CO 2 showed increased cell quotas for both carbon and nitrogen, irrespective of nitrogen supply ( Table 2). The highest C:N was found in the high CO 2 /low nitrogen conditions that are expected to dominate the open ocean in the near future, indicating that these synergistic effects of ocean acidification and nitrogen limitation could decrease the food quality of marine phytoplankton.
Although increased marine dissolved CO 2 may bring some benefits in terms of improved carbon supply to some phytoplankton [8], these organisms also face an extra cost associated with changed marine chemistry, especially pH stress, which could also lead to more energetic constraints on growth. Thus, the net benefit of higher CO 2 will be a balance between gains and losses determined by the various environmental factors associated with climate change.
In conclusion, ocean acidification together with ocean change can act in the marine environment synergistically or antagonistically to affect diatom performance, depending on the levels of sunlight [14]. Intensified stratification may push the marine phytoplankton into nitrogen-limited status, and will thereby influence the physiological or biochemical characteristics of the phytoplankton cells. Increases in respiratory metabolism may counteract any increase in the rate of C gain through photosynthesis (and hence affect net growth) [11], [14], [36]. Enhanced C:N ratios, induced by high CO 2 and low nitrogen, can influence secondary producers as well as predators at higher levels. Effects of ocean acidification on marine primary producers can be speciesspecific due to their physiological diversities and vary between different oceanic regions correlated with differed physical, chemical or biological conditions.