Combined effects of elevated p CO 2 and temperature on biomass and carbon 1 fixation of phytoplankton assemblages in the northern South China Sea

Abstract. The individual influences of ocean warming and acidification on marine organisms have been investigated intensively, but studies regarding the combined effects of both global change variables on natural marine phytoplankton assemblages are still scarce. Even fewer studies have addressed possible differences in the responses of phytoplankton communities in pelagic and coastal zones to ocean warming and acidification. We conducted shipboard microcosm experiments at both off-shore (SEATS) and near-shore (D001) stations in the northern South China Sea (NSCS) under three treatments, low temperature (30.5 °C at SEATS and 28.5 °C at D001) and low pCO2 (390.0 µatm at SEATS and 420.0 µatm at D001) (LTLC), high temperature (33.5 °C at SEATS and 31.5 °C at D001) and low pCO2 (390 µatm at SEATS and 420 µatm at D001) (HTLC), and high temperature (33.5 °C at SEATS and 31.5 °C at D001) and high pCO2 (1000 µatm at SEATS and 1030 µatm at D001) (HTHC). Biomass of phytoplankton at both stations were enhanced by HT. HTHC did not affect phytoplankton biomass at station D001 but decreased it at station SEATS. At this offshore station HT alone increased daily primary productivity (DPP, µgC (µg chl a)−1 d−1) by ~ 64 %, and by ~ 117 % when higher pCO2 was added. In contrast, HT alone did not affect DPP and HTHC reduced it by ~ 15 % at station D001. HT enhanced the dark respiration rate (µg C (µg chl a)-1 d−1) by 64 % at station SEATS, but had no significant effect at station D001, and did not change the ratio of respiration to photosynthesis at either station. HTHC did not affect dark respiration rate (µg C (µg chl a)−1 d−1) at either station compared to LTLC. HTHC reduced the respiration to photosynthesis ratio by ~ 41 % at station SEATS but increased it ~ 42 % at station D001. Overall, our findings indicate that responses of coastal and offshore phytoplankton assemblages in NSCS to ocean warming and acidification are contrasting, with the pelagic phytoplankton communities being more sensitive to these two global change factors.


Introduction
From 1979 to 2012, the mean global sea surface temperature (SST) increased at a rate of ~0.12°C per decade, based on in situ data records (IPCC, 2013).
Particularly, the warming rate in the surface of South China Sea (~0.26°C per decade) from 1982 to 2004 (Fang et al., 2006) appears to be about is approximately 2 times faster than the global ocean mean rate (~0.14 per decade) from 1960 to 1990 (Casey and Cornillon, 2001;Fang et al., 2006).It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings.Ocean warming is projected to rise approximately by 0.6 °C (Representative Concentration Pathway (RCP) 2.6) to 2.0 °C (RCP 8.5), in the upper 100 m of the water column by the end of the 21 st century (IPCC, 2013).
Ocean warming is known to affect primary productivity both directly and indirectly.In situ surface chl a declined exponentially with rise of  in the northeast Atlantic Ocean from latitudes 29 to 63° N in spring and summer, which was attributed to enhanced stratification and consequent reduced upward transport of nutrients into the upper mixed layer (Poll et al., 2013).The seawater volume-specific primary produtivity also decreased with temperature rise due to lower phytoplankton biomass (Poll et al., 2013).Mesocosm experiments have also demonstrated that ocean warming (an increase of 6 °C ) can decrease phytoplankton biomass by ~80% in the Baltic Sea (Lewandowska et al., 2012;Sommer et al., 2015).On the other hand, ocean warming did not affect volume-specific primary produtivity in the Baltic Sea (Lewandowska et al., 2012).
Increased atmospheric CO 2 is responsible for global warming, and the oceans., As a main sink for CO 2 , the ocean has have absorbed approximately 30% of the emitted anthropogenic carbon dioxide (IPCC, 2013).Such a dissolution of CO 2 increases seawater CO 2 partial pressure and bicarbonate ion levels and decreases pH and carbonate ion concentrations, leading to ocean acidification (Orr et al., 2005OA).
By 2100, the projected decline in global-mean surface pH is projected to be approximately 0.065 (RCP 2.6) to 0.31 , (RCP 8.5) depending on the RCP scenario (IPCC, 2013).In terms of the South China Sea, an accelerated trend of ocean acidification has been reported and the rate of pH decline almost tripled between 1951 and 2000, compared to that between 1840 and 1950 (Liu et al., 2014).
Dissolved CO 2 may be a potentially limiting factor for marine primary productivity because of the low CO 2 level in seawater and the low affinity of the enzyme Rubisco for dissolved CO 2 (Falkowski and Raven, 2013).In addition, CO 2 in seawater diffuses approximately 10,000 times slower than in air, leading to its supply rate being much lower than the demand of photosynthetic carbon fixation (Raven, 1993;Riebesell et al., 1993).Although phytoplankton have evolved carbon-concentrating mechanisms (CCMs) to cope with these problems challenges (Giordano et al., 2005;Raven et al., 2012;Reinfelder, 2011), increased CO 2 concentration may still be beneficial since energy saved due to down-regulation of CCMs under elevated CO 2 can be utilized in other metabolic processes (Gao Wu et al., 2012a2010).Early laboratory and shipboard experiments suggested that increased CO 2 indeed could enhance phytoplankton growth rates and thus marine primary productivity (Riebesell et al., 1993;Hein and Sand-Jensen, 1997;Schippers et al., 2004).Since then, neutral effects of increased CO 2 on growth of phytoplankton assemblages have also been reported (Tortell and Morel, 2002;Tortell et al., 2000Gao et al., 2012 andreferences therein).Furthermore, ocean acidification can even reduce primary productivity of surface phytoplankton assemblages when exposed to incident solar radiation (Gao et al., 2012b).Therefore, the effects of ocean acidification on marine primary productivity remain controversial and its interactions with other environmental factors, such as warming, solar UV radiation, hypoxia, etc. are incompletely understood (Gao et al., 2012a;Hä der and Gao, 2015;Mostofa, 2016).
Ocean warming and acidification, both caused by increasing atmospheric CO 2 , are proceeding simultaneously.The interactive or combined effects of warming and OA could be completely different from that of either one stressor (Hare et al. 2007, Feng et al., 2009;Gao et al., 2012a, Tatters et al. 2013).Several oceanographic cruises and ship board experiments in the European sector of the Arctic Ocean, showed that gross primary production increased with pCO 2 (145-2099 µatm) and the greatest increase was observed in lower temperature regions, indicating CO 2 -enhanced primary production in the European Arctic Ocean is temperature-dependent (Holding et al., 2015).
The South China Sea (SCS) is located between the equator and 23.8° N, from 99.1 to 121.1° E, and is one of the largest marginal seas in the world, with a total area of about 3.5 × 10 6 km 2 .Therefore, understanding the effects of ocean warming and acidification on primary production in SCS would help us to define the role of marginal seas in the global carbon cycle.However, only a very few studies on the effects of ocean acidification or warming on primary productivity in the SCS have been reported.Wu and Gao (2010) reported that CO 2 enrichment (700 µatm) did not affect the photosynthetic carbon fixation rate assimilation number of phytoplankton at a near-shore site in SCS, compared to the ambient CO 2 level (380 µatm).Gao et al. (2012b) demonstrated that increased pCO 2 (800 or 1000 µatm) reduced primary productivity in off-shore stations of the SCS.Therefore, we hypothesized that the effects of ocean acidification or/and warming on primary productivity in SCS would be site-dependent.None of the previous studies have examined co-effects of warming and increased CO 2 on primary production in the SCS.In this study, to test this hypothesis we conducted shipboard microcosm experiments at both near-shore and off-shore stations to determine the combined effects of ocean warming and acidification on biomass, photosynthetic carbon fixation, and dark respiration of phytoplankton assemblages in the SCS.

Experimental setup
The experiments were conducted at one off-shore station SEATS (17.9963 o N, 115.9621 o E) and one near-shore station D001 (18.9740 o N, 110.7166 o E) in the NSCS (Fig. 1).Surface seawater (0-2 m) was collected before sunrise with a 10 L acid-cleaned plastic bucket, filtered (180 µm) to remove large grazers and dispensed into nine microcosms.Microcosms consisted of cylindrical polymethyl methacrylate tanks (32 L, 0.34 m water depth) with water-jacketed space for circulating cooled water, .In the microcosms, phytoplankton assemblages could be exposed to 81-91% and 64-91% of full sunlight at the off-shore and near-shore stations respectively, due to the shielding of the cover, water-jacket and the depth of the water., allowing 91% photosynthesis active radiation (PAR, 400-700 nm), 63% ultraviolet-A (UVA, 315-400 nm) and 6% ultraviolet-B (UVB, 280-315 nm) transmission under incident solar radiation.Since we aimed to investigate the impacts on surface phytoplankton assemblages, and provide comparable data with the previous study in the SCS (Gao et al., 2012b), we decided not to shield the microcosms.Two levels of temperature (in situ, in situ + 3 o C) and pCO 2 (ambient, ambient + 610 µatm) were used.There were three triplicated treatments: low temperature and low pCO 2 , LTLC; high temperature and low pCO 2 , HTLC; high temperature and high pCO 2 , HTHC.The treatment of low temperature and high pCO 2 was missing due to the lack of microcosms.Microcosm temperature was controlled and monitored with via circulating coolers with a variation of ± 1.0 o C (CTP-3000, EYELA, Japan) and stable CO 2 equilibrium with the sea water (variation of pCO 2 < 5%) was achieved within 24 hours using a CO 2 enricher (CE-100, Wuhan Ruihua Instrument & Equipment Ltd, China).The incubations were conducted for seven days for the off-shore station SEATS (Aug 3 rd -9 th 2012) and six days for the near-shore station D001 (Aug 14 th -19 th 2012).

Solar radiance
The incident solar radiation was continuously monitored using an Eldonet broadband filter radiometer (Eldonet XP, Real Time Computer, Germany) that was fixed at the top of the ship.It measured every second and recorded the means over each minute.

Carbonate chemistry parameters
The seawater pH in the microcosm was recorded with a pH meter (FE20, Mettler Toledo, Greifensee, Switzerland) every hour during the first day of incubation and daily afterwards.The pCO 2 in seawater was maintained with the CO 2 enricher and measured by an automated flowing pCO 2 measuring system (Model 8050, GO, USA).
Other carbonate system parameters were derived via CO2SYS (Pierrot et al., 2006), using the equilibrium constants of K 1 and K 2 for carbonic acid dissociation (Roy et al., 1993).

Chlorophyll a analysis
For the measurement of chlorophyll a (chl a), 500 mL of seawater were filtered onto a Whatman GF/F glass fiber filter (25 mm).Then the filter was placed in 5 ml 93% acetone at -20 °C for 24 h.Chl a concentration was determined with a fluorometer (Trilogy, Turner Designs, USA), following the protocol of Welschmeyer (1994).The concentrations of chl a in situ and in microcosms were measured at the beginning and end of the experiment, respectively.

Primary productivity and dark respiration
Seawater samples taken from each microcosm at the end of the experiment were dispensed into 50 mL quartz tubes, inoculated with 5 µCi (0.185 MBq) NaH 14 CO 3 (ICN Radiochemicals, USA) and then incubated for 12 h (from 6:00 a.m. to 6:00 p.m.) and 24 h (from 6:00 a.m. to 6:00 a.m. next day) under natural light and day-night conditions.The incubation temperature of every treatment was the same as the corresponding microcosm treatment.After the incubation, the cells were filtered onto a Whatman GF⁄ F glass fiber filter (25 mm), which was immediately frozen and stored at -20 o C for later analysis.In the laboratory, each frozen filter was placed into a 20 mL scintillation vial, exposed to HCl fumes overnight, and dried (55 o C, 6 h) to expel non-fixed 14 C (Gao et al., 2007).Then 3 mL scintillation cocktail (Perkin Elmer®) was added to each vial and incorporated radioactivity was counted by a liquid scintillation counting (LS 6500, Beckman Coulter, USA).The daytime primary productivity (DPP) was defined as the amount of carbon fixation during 12 h incubation.The dark respiration was defined as the difference in amount of carbon fixation between 12 h and 24 h.Carbon fixation over 24 h was taken as daily net primary productivity (NPP).Ratio of respiration to photosynthesis (R/P) was expressed as that of respiratory carbon loss to daytime carbon fixation.The primary productivity and dark respiration in situ were measured at the beginning of the experiment.

Statistical analyses
Results were expressed as means of replicates ± standard deviation.Data were analyzed using the software SPSS v.21.The data from each treatment conformed to a normal distribution (Shapiro-Wilk, P > 0.05) except the DPP (P = 0.034) and dark respiration (P = 0.034) under HTLC at the near-shore station, and the variances could be considered equal (Levene's test, P > 0.05).One-way ANOVAs were conducted to assess the significant differences in carbonate chemistry parameters, chl a, DPP, NPP, dark respiration, ratio of dark respiration to photosynthesis between three treatments.
Tukey HSD was conducted for post hoc investigation.Independent samples t-tests were conducted to compare in situ chl a, DPP, NPP, dark respiration, and ratio of dark respiration to photosynthesis between both stations.The threshold value for determining statistical significance was P < 0.05.

Results
The incident solar radiation during the experiment was recorded (Table 1).The daytime (12 h) mean solar radiation ranged from 927 to 1592 µmol photons m -2 s -1 at the off-shore station SEATS while the lowest solar radiance was only 111 µmol photons m -2 s -1 , with the highest of 1583 µmol photons m -2 s -1 , at the near-shore station D001.The average of daytime mean solar radiation over the microcosm incubation at the off-shore station SEATS was 15.52% higher than that at the near-shore station D001.
The changes of the seawater carbonate system under different conditions are shown in Table 2.At the off-shore station SEATS, an increase of 3 o C in temperature (HTLC) did not alter carbonate parameters except leading to enhanced CO 3 2-(Tukey HSD, P = 0.009).HTHC resulted in a significant decrease in CO 3 2-(Tukey HSD, P < 0.001) and TA (Tukey HSD, P = 0.016) but an increase in CO 2 (Tukey HSD, P < 0.001) compared with LTLC.The effect of temperature on carbonate parameters at the near-shore station D001 were similar to the off-shore station SEATS, while HTHC increased HCO 3 -(Tukey HSD, P = 0.046) and did not affect TA (Tukey HSD, P = 0.203).
The in situ ratio of respiration to photosynthesis was 31.9 ± 3.5% at the off-shore station SEATS, significantly higher than that (24.2 ± 1.3%) at the near-shore station D001 (Independent samples t-test, t = 3.537, df = 4, P = 0.0024; Fig. 6a).After seven days incubation in microcosms, Tukey HSD comparison (P = 0.05) showed that higher temperature did not affect the ratio of respiration to photosynthesis but HTHC reduced it from 27.8 ± 1.6% to 16.5 ± 1.3% at the off-shore station SEATS (Fig. 6b).

Effects of increased temperature and CO 2 on biomass
The in situ chl a concentration at station D001 was higher than that at SEATS, indicating more phytoplankton biomass at station D001.The chl a concentration decreases with distance from the coast in the NSCS, mainly due to the change of nutrients (Li et al., 2011).The higher temperature increased chl a concentration at both stations, which might be attributed to increased active uptake of nutrients at the elevated temperatures through enhanced enzymatic activities.Algal and cyanobacterial growth commonly increases with temperature within a suitable range and then decreases after the optimal temperature point/range (Goldman and Carpenter, 1974;Montagnes and Franklin, 2001;Savage et al., 2004;Boyd et al. 2013) and optimum temperatures for growth of marine phytoplankton are usually several degrees higher than the environmental temperatures (Thomas et al., 2012), which could explained the increase chl a level of phytoplankton grown at the higher temperature . in the present study.
On the other hand, the elevated CO 2 offset the positive effect of the higher temperature on chl a at the near-shore station D001, and even reduced chl a at the off-shore station SEATS.High CO 2 can sometimes enhance algal photosynthesis and growth, since CO 2 in seawater is suboptimal for full operation of Rubisco enzymes (Wu Giordano et al., 20082005 and references therein).On the other hand, positive effects of elevated CO 2 can be affected by other environmental factors.Gao et al.
(2012b) demonstrated that rising CO 2 could enhance growth of diatoms at low light intensity, but decrease it at high light intensity.It was found that rising CO 2 concentration lowered the threshold for diatom growth above which photosynthetic active radiation becomes excessive or stressful, owing to reduced energy requirements for inorganic carbon acquisition at elevated CO 2 (Gao et al., 2012b).In the present study, the mean daily solar radiation levels during incubation were 1312 and 1136 µmol photons m -2 s -1 (Table 1), corresponding to phytoplankton in the microcosms exposed to 1068-1194 µmol photons m -2 s -1 at the off-shore station and 729-1034 µmol photons m -2 s -1 at the near-shore station, which were far above the threshold light intensity reported for diatoms (Gao et al., 2012b).Consequently, the higher CO 2 combined with the high solar radiation in summer of NSCS may have imposed negative effects on phytoplankton biomass at both the off-shore and the near-shore stations SEATS and D001.In addition, the inhibitory effect of higher CO 2 on biomass was more significant at the off-shore station SEATS than the near-shore stationD001.
This can be attributed to the higher sensitivity of picoplankton to high solar radiation (Li et al., 2011;Wu et al., 2015), which could be delivered to the interaction of high solar radiation and high CO 2 .As shown in Li et al.'s (2011) study, the proportion of picoplankton in phytoplankton assemblages increased with distance off the coasts.
Therefore, the dominant species at the off-shore station SEATS are pico-and nano-phytoplankton, but micro-phytoplankton at the near-shore station D001 (Table 3).

Effects of increased levels of temperature and CO 2 on primary productivity
The seawater volume-specific DPP at the near-shore station D001 was higher than the off-shore station SEATS.This should result from both higher chl a concentration and chl a-specific DPP at the near-shore stationD001.It has been shown that more smaller cells exist at the off-shore stationSEATS than at the near-shore stationD001 (Table 3).Smaller cells have been considered to have larger DPP, according to Laws's model (1975).The discrepancy between our finding and Laws's model may be due to the availability of nutrients.Laws's model was based on growth rates obtained from the same nutrient level.Nevertheless, the nutrient level at the near-shore station D001 is higher than at the off-shore station SEATS (Table 3), leading to higher DPP.The higher temperature increased seawater volume-specific DPP at both stations.This could be attributed to more biomass produced at the warmer conditions, as indicated by chl a. High temperature enhanced the chl a-specific DPP at the off-shore station SEATS.However, no positive effects of temperature were found on chl a-specific DPP at the near-shore station D001.The differential effects of temperature on chl a-specific DPP between the stations may be due to the phytoplankton community composition, since cyanobacteria and/or pico-phytoplankton have the strongest temperature response in terms of photosynthetic carbon fixation compared to micro-and nano-phytoplankton (Andersson et al., 1994).This finding contributes to the explanation of the dominance of pico-phytoplankton in a warmer ocean (Chen et al., 2014;Montagnes and Franklin, 2001;Hare et al. 2007;Morá n et al., 2010).HTHC reduced chl a-specific DPP at the near-shore station D001, but increased it at the off-shore station SEATS.High CO 2 also reduced chl a-specific DPP in a previous study, which could result from the interaction of high CO 2 and high solar radiation during the summer in the NSCS (Gao et al., 2012b).The reason that HTHC stimulated chl a-specific DPP at the off-shore station SEATS may be due to a dramatic decline in chl a concentration under the HTHC treatment.

Effects of increased temperature and CO 2 on respiration
The dark respiration rate of phytoplankton at the near-shore station D001 was higher than that at the off-shore stationSEATS, regardless of normalizing to seawater or chl a.The respiration rate of algae or cyanobacteria usually increases with cell size (López-Sandoval et al., 2014).The off-shore sStation SEATS has more pico-phytoplankton, which would lead to a lower chl a-specific dark respiration rate and then lower seawater volume-specific dark respiration, particularly when combined with lower chl a level.The higher temperature increased seawater volume-specific dark respiration at both stations, which could be related to increased chl a concentration and/or enhanced respiratory carbon loss at the higher temperature.
The higher temperature also increased chl a-specific dark respiration rate at the off-shore station SEATS.This is consistent with Butrón et al.'s (2009) study, in which respiration rates of phytoplankton along Nervión-Ibaizabal estuary showed a positive correlation with temperature.Robarts and Zohary (1987) also found that respiration rate of bloom-forming cyanobacteria was temperature-dependent, with optima over 25 °C.
HTHC reduced seawater volume-specific dark respiration at both stations, which should be the consequence of the decreased chl a in this treatment.The higher temperature increased chl a-specific dark respiration rate at the off-shore station SEATS, but there was no significant difference between HTHC and LTLC, indicating the higher CO 2 inhibited the chl a-specific dark respiration rate.Similarly, reduced respiration was found in mesocosm studies (Spilling et al., 2016).In theory, higher CO 2 would inhibit respiratory release of CO 2 .Nevertheless, enhanced respiration rate at higher CO 2 conditions have been found in laboratory-grown diatoms (Wu et al., 2010), coccolithophores (Jin et al., 2015), mixed phytoplankton assemblages (Jin et al., 2015), and macroalgae as well (Zou et al., 2011).Such increased respiration has been attributed to extra energy demand to cope with increased seawater acidity caused by higher CO 2 (Gao and Campbell, 2014;Raven et al., 2014).Therefore, the effect of increased CO 2 on phytoplankton respiration could be due to the combined effects of CO 2 diffusive resistance and seawater acidity stress.Meanwhile, neither higher temperature alone nor HTHC significantly affected chl a-specific dark respiration rate at the near-shore station D001.One possible reason could be that larger cells are less sensitive to CO 2 diffusive resistance and acidic stress due to thicker diffusion boundary layers around the cells.This hypothetical explanation is worthy of future testing.

Effects of increased temperature and CO 2 on R/P
Laws's model (1975) has suggested that large phytoplankton cells are likely to consume a smaller fraction of their biomass to compete with small phytoplankton cells in terms of the growth rate, considering small cells have higher gross production rates.Our finding that phytoplankton in at the off-shore station SEATS had a higher R/P than the near-shore station D001 confirms Laws's model.It was theorized that autotrophic respiration is more sensitive to temperature than photosynthesis and the ratio of R/P was predicted to increase with temperature (Ryan, 1991;Woodwell, 1990;Woodwell et al., 1983).However, the assumption that plant respiration is highly temperature dependent was primarily based on short-term (a few minutes or hours) responses of plants to changes of temperature (Gifford, 1994).In long-term experiments (days or months), the increase in respiration with temperature tended to disappear due to acclimation (Gifford, 1995;Jones, 1977;Reich et al., 2016;Slot and Kitajima, 2015;Ziska and Bunce, 1998).Photosynthetic acclimation to warming is variable (Chalifour and Juneau, 2011;Hancke and Glud, 2004;Schlüter et al., 2014;Staehr and Birkeland, 2006).However, a general acclimation response to long-term increased temperature is a rise in optimal temperature of photosynthesis (Gunderson et al., 2010;Kattge and Knorr, 2007;Staehr and Birkeland, 2006).Such shifts in the temperature response of photosynthesis and respiration via physiological acclimation can dampen the increase in R/P at high temperatures, or climate warming would not increase R/P (Reich et al., 2016).In other words, phytoplankton would down-regulate the high sensitivity of respiration to temperature, and maintain a relatively consistent net primary production and hence food web structure in a warming ocean.Although the ratio of R/P did not vary with increased temperature at both either stations in our work either, both photosynthesis and respiration were enhanced by the higher temperature.The constant ratio was due to the similar amplitude of increase in photosynthesis and respiration.The incubation period of 7-8 days might not be enough for phytoplankton to acclimate to the increased temperature completely, and therefore the stimulated effects of high temperature on photosynthesis and respiration were still notable.Opposite effects of HTHC on R/P were detected at the two stations SEATS and D001, negative at the off-shore stationSEATS and positive at the near-shore stationD001.This can be attributed to differential responses of photosynthesis at both stations to HTHC, considering the responses of respiration were similar.

Conclusions
This study demonstrates that ocean warming expected to occur by the end of the century would simulate the DPP and also dark respiration of phytoplankton assemblages in NSCS, but this positive effect can be damped or offset when ocean acidification is combined.The responses of phytoplankton assemblages locating different areas to ocean warming and acidification could be contrasting due to various phytoplankton compositions and physical and chemical environments.It seems that phytoplankton assemblages in pelagic areas are more sensitive to ocean warming and acidification.More exhaustive investigations are needed to obtain an accurate view of primary production under future ocean environment.
This study demonstrates that a short-term rise of SST appeared to simulate the DPP and dark respiration of phytoplankton assemblages in the NSCS.However, this positive effect could bewas dampened or offset when warming and ocean acidification treatments were combined.The regional responses of phytoplankton assemblages at the two stations to ocean warming and acidification may differ due to differences in physical and chemical environment as well as phytoplankton community structure.
The combined treatment of warming and acidification reduced biomass and dark respiration rate at the off-shore, but did not affect them at the near-shore station.
Ecologically and geographically, our data implies differential responses of primary production to ocean climate change.This short-term experiment suggests the need to determine whether similar effects may occur over the longer timescales of future anthropogenic change.Table 2. Parameters of the seawater carbonate system at different incubation conditions.Measurements and estimation of the parameters were described in Methods.Data are the means ± SD (n = 3).LTLC, low temperature and low pCO 2 ; HTLC, high temperature and low pCO 2 ; HTHC, high temperature and high pCO 2 .DIC = dissolved inorganic carbon, TA = total alkalinity.Different superscript letters indicate significant differences between treatments within one station.Figure 1 27 1312 a The dates for measurements of photosynthetic carbon fixation in situ.b The dates for measurements of photosynthetic carbon fixation experiencing temperature and pCO 2 treatments.c Mean represents the average of daytime mean solar radiation over seven or six days microcosm incubation.

Figure captions Figure 1 .
Figure captions Figure 1.Experimental stations in the northern South China Sea.

Figure 2 .
Figure 2. Chl a concentration in situ (a) and after temperature and pCO 2 treatments in microcosms (b).The microcosm incubations lasted seven days at off-shore station SEATS and six days at near-shore station D001.The error bars indicate the standard deviations (n = 3).The different letters above the error bars represent significant (P < 0.05) differences between stations in panel (a) and between treatments in panel (b).

Figure 3 .
Figure 3. Daytime primary productivity (DPP) in situ (a, b) and after temperature and pCO 2 treatments in microcosms (c, d).The microcosm incubations lasted seven days at off-shore station SEATS and six days at near-shore station D001.The error bars indicate the standard deviations (n = 3).The different letters above the error bars represent significant (P < 0.05) differences between stations in panels (a, b) and between treatments in panels (c, d).

Figure 4 .
Figure 4. Net primary productivity (NPP) in situ (a, b) and after temperature and pCO 2 treatments in microcosms (c, d).The microcosm incubations lasted seven days at off-shore station SEATS and six days at near-shore station D001.The error bars indicate the standard deviations (n = 3).The different letters above the error bars represent significant (P < 0.05) differences between stations in panels (a, b) and between treatments in panels (c, d).

Figure 5 .
Figure 5. Dark respiration in situ (a, b) and after temperature and pCO 2 treatments in microsoms (c, d).The microcosm incubations lasted seven days at off-shore station SEATS and six days at near-shore station D001.The error bars indicate the standard deviations (n = 3).The different letters above the error bars represent significant (P < 0.05) differences between stations in panels (a, b) and between treatments in panels (c, d).

Figure 6 .
Figure 6.The ratio of respiration to photosynthesis in situ (a, b) and after temperature and pCO 2 treatments (c, d) in microcosms.The microcosm incubations lasted seven days at off-shore station SEATS and six days at near-shore station D001.The error bars indicate the standard deviations (n = 3).The different letters above the error bars represent significant (P < 0.05) differences between stations in panels (a, b) and Figure 2

Table 3 .
Physical, chemical, and biological parameters at off-shore stations SEATS and near-shore station D001.SST: seawater surface temperature; N: NO