Impacts of elevated CO2 on phytoplankton community composition and organic carbon dynamics in nutrient-depleted Okhotsk Sea surface waters

Title Impacts of elevated CO2 on phytoplankton community composition and organic carbon dynamics in nutrient-depleted Okhotsk Sea surface waters Author(s) Yoshimura, T.; Nishioka, J.; Suzuki, K.; Hattori, H.; Kiyosawa, H.; Watanabe, Y. W. Citation Biogeosciences Discussions, 6(2), 4143-4163 https://doi.org/10.5194/bgd-6-4143-2009 Issue Date 2009-04-15 Doc URL http://hdl.handle.net/2115/48181 Rights(URL) http://creativecommons.org/licenses/by/3.0/ Type article File Information BD6-2_4143-4163.pdf

Interactive Discussion significant direct impacts on a wide range of marine organisms (Raven et al., 2005;Kleypas et al., 2006), the responses of individual organisms, populations, communities, and ecosystems are largely unknown (Vézina and Hoegh-Guldberg, 2008;Doney et al., 2009). Considerable research efforts have mainly focused on predicting the impact of ocean acidification on calcifying marine organisms (Shirayama and Thornton,5 2005; Gazeau et al., 2007;Hoegh-Guldberg et al., 2007). However, studies on planktonic primary producers in relation to ocean acidification are still limited despite the established importance of these organisms in marine ecosystems and biogeochemical cycles (Riebesell, 2004;Rost et al., 2008). Therefore, manipulative experiments are required for assessing the impacts of acidification on phytoplankton, and for integrating 10 the impacts into biogeochemical modeling approach (Orr et al., 2009). Several past studies using natural phytoplankton assemblages have predicted that change in CO 2 concentration could alter the phytoplankton species composition. For example, Tortell et al. (2002) showed that an increment of CO 2 changed the taxonomic composition of the phytoplankton assemblage in the Equatorial Pacific, in which 15 150 ppm favor the prymnesiophyte Phaeocystis sp. and 750 ppm favor the diatoms. Kim et al. (2006) showed that increasing CO 2 stimulated the growth of the diatom Skeletonema costatum in coastal waters of southern Korea. Tortell et al. (2008) also reported that elevated CO 2 enhanced primary production and relative abundance of chain-forming Chaetoceros in the Southern Ocean. On the other hand, in the Bering 20 Sea, diatoms were replaced by autotrophic nanoflagellates under a high CO 2 condition (Hare et al., 2007). Recently, Riebesell et al. (2007) demonstrated that increased CO 2 concentration enhanced the efficiency of the biological carbon pump during a diatom bloom in a coastal mesocosm experiment. These studies indicate that CO 2 increase can alter not only the phytoplankton species composition but also the organic carbon 25 dynamics in the sea. However, it should be noted that all these results were obtained from artificial algal blooms induced under macro-and/or micro-nutrient replete conditions. At present, macro-and/or micronutrients are depleted in most surface areas of the oceans. Physiological responses of phytoplankton can be expected to differ BGD 6,2009  between under nutrient-depleted and replete conditions. Additionally, in such nutrientdepleted waters, regenerated production via the microbial loop may become dominant (Azam et al., 1983). Further studies are clearly essential to address the effects of CO 2 on phytoplankton community composition and carbon dynamics in nutrient-depleted waters. 5 We thus conducted a CO 2 manipulated bottle incubation experiment using the natural plankton community collected from nutrient-depleted surface water in the Sea of Okhotsk in summer 2006, as a part of the framework of the Plankton Ecosystem Response to CO 2 Manipulation Study (PERCOM). Care was taken to ensure that no nutrient enrichment of the incubation bottles occurred and we measured changes over time 10 in phytoplankton community structure and organic carbon accumulation, and discuss the effects of CO 2 on these parameters.

Sampling and experimental setup
The study was carried out aboard the R/V Professor Khromov in summer of 2006. Sur-15 face seawater sampling for a CO 2 manipulation experiment was conducted at a station (49.5 • N, 148.25 • E; 1050 m depth) in the Sea of Okhotsk on 25 August 2006 (Fig. 1).
The water was collected from 10 m depth with acid-cleaned Niskin-X bottles on a Kevlar wire. A total of 160 L of seawater was poured into four 50 L polypropylene carboys (40 L seawater for each of the given CO 2 concentrations) through silicon tubing with a 243 µm 20 mesh Teflon net to remove any large plankton. Subsamples were taken from the each carboy and poured into three acid-cleaned 9 L polycarbonate bottles (12 bottles in total) for incubation. Similarly, initial samples also were collected from the each carboy. All of the sampling was carried out using trace metal clean techniques to avoid any effects of trace elements other than CO 2 on phytoplankton growth. Introduction Tables  Figures   Back  Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion for 14 days under in situ temperature of 13.5 • C and 50% surface irradiance. To control CO 2 concentrations in the incubation bottles, air mixtures containing 180, 380, 750, and 1000 ppm CO 2 , purchased from a commercial gas supply company (Nissan-Tanaka Co., Japan), were bubbled into triplicate incubation bottles. The flow rate of the gas was set at 100 mL min −1 for the first 24 h, and thereafter maintained at 50 mL min −1 .

5
Subsamples were collected from each incubation bottle on days 1, 3, 5, 9, and 14 at 10:00 a.m. during the course of the experiment.

Sample treatments and analyses
Subsamples for total dissolved inorganic carbon (DIC) and total alkalinity (TA) were collected in 120 mL glass vials, poisoned with HgCl 2 . Concentrations in DIC and TA 10 were measured by coulometric titration (Johnson et al., 1985) and single point titration (Culberson et al., 1970), respectively. CO 2 and pH were calculated by using a set of CO 2 constants as "K1, K2 from Mehrbach et al., (1973) and refitted by Dickson and Millero (1987)", and on pH scale as "total hydrogen scale" under in situ temperature of 13.5 • C and salinity of 32.52. In addition, certified reference material distributed by Interactive Discussion tles were pooled together and filtered onto GF/F filters, which were stored at −85 • C until analysis on land. The pigments extracted in DMF were measured with a highperformance liquid chromatography (CLASS-VP system, Shimadzu) according to the method detailed in Suzuki et al. (2005). Particulate organic carbon (POC) were collected on pre-combusted (450 • C, 4 h) 5 25 mm GF/F filter under a gentle vacuum at <0.01 MPa and the filter samples were stored at −20 • C until analysis on land. Before POC analysis, the filter samples were freeze-dried, and then exposed to HCl fumes overnight in a desiccator to remove inorganic carbon from the samples. Carbon contents on the filter were measured with a CN elemental analyzer (NA-1500, Fisons). Subsamples for dissolved organic carbon BGD 6,2009 Impacts of elevated CO 2 on phytoplankton T. Yoshimura et al.

Results and discussion
We could create significant gradients in pCO 2 and pH among the four treatments by bubbling the four different air-CO 2 mixtures into the incubation seawater medium (Fig. 2). Mean seawater pCO 2 and pH in 180, 380, 750, and 1000 ppm CO 2 treatments during the experiment were 145-148, 275-276, 479-489, and 583-597 µatm 5 and 8.40, 8.18, 7.97, and 7.89, respectively. The CO 2 systems were manipulated significantly by comparison with the initial value of pCO 2 of 195-200 µatm and pH of 8.30. The seawater pCO 2 did not fully equilibrate with the bubbling air CO 2 concentrations. This may be attributable to the insufficient flow rates of the air bubbling. Our results clearly shows that direct measurements of at least two parameters in seawater 10 CO 2 system are necessary to confirm whether the CO 2 condition in the experimental system equilibrates with the bubbling gas. Hereafter we will refer 150, 280, 480, and 590 µatm for the 180, 380, 750, and 1000 ppm CO 2 treatments, respectively. Macronutrients were depleted within the surface mixed layer and the phytoplankton biomass in terms of chl-a concentration was low (Table 1). Nitrogenous nutrients were 15 completely depleted, and PO 4 and SiO 2 concentrations were 0.25 and 1.06 µmol L −1 , respectively, before incubation. Initial chl-a biomass was comprised of 0.30 µg L −1 of small-sized (<10 µm) algal cells and 0.01 µg L −1 of large-sized (>10 µm) ones. Phytoplankton cells larger than 5 µm in size were rarely observed using light microscopy. Diatoms were scarcely observed in that size range. Some dinoflagellates such as, 20 Prorocentrum, Gonyaulax, and Ceratium were identified, but their abundances were low. The ultraplankton community (ca. <5 µm in size), revealed by flow cytometry, was comprised of Synechococcus, eukaryotic ultraphytoplankton, and heterotrophic bacteria. Initial abundances of the cells were 9.0×10 3 , 1.8×10 4 , and 3.8×10 5 cells mL −1 , respectively.

25
During the 14-day incubation experiment, chl-a concentrations fluctuated between 0.1 and 0.4 µg L −1 , but the variation pattern was similar in all treatments (Fig. 3).
Chlorophyll-a was mainly derived from small-sized cells throughout the experiment.
BGD 6,2009 Impacts of elevated CO 2 on phytoplankton T. Yoshimura et al.  Liu et al. (2009) reported that the growth rate of the in situ phytoplankton assemblage at the site where the present study was carried out was estimated to be 0.38 d −1 , representing the mean doubling time of the cells was approximately 2 days. If we set the endpoint as the end of the experiment (day 14), net chl-a production ranged from −0.04 to −0.16 between the treatments (Fig. 4a). No significant difference in the net 5 chl-a production was observed between the treatments. Change in the concentrations of NH 4 , NO 3 +NO 2 , and PO 4 were small (Table 1). Silicate was consumed to 53% of the initial concentration; the consumption in 150 µatm was significantly larger than other treatments ( Fig. 4b; Tukey's test, p<0.05).
Although apparent changes in chl-a biomass were small under the nutrient depleted 10 conditions, fucoxanthin/chl-a ratios showed variations between the treatments (Fig. 4c).
At the end of the experiment, a decreasing trend was observed in the fucoxanthin/chl-a ratio (g g −1 ) from 0.30 in 150 µatm to 0.16 in 590 µatm. These fucoxanthin/chl-a ratios were higher than the initial value of 0.12. Since the pigment samples were pooled from the triplicate bottles in each treatment, the statistical significance for these pigments 15 values could not be evaluated. It is known that fucoxanthin is contained in diatoms, prymnesiophytes, chrysophytes, and raphidophytes (Jeffrey et al., 1997). In this study, 19 -butanoyloxyfucoxanthin and 19 -hexanoyloxyfucoxanthin, which are chrysophyte and cryptophyte markers, respectively (Jeffrey et al., 1997), were very low throughout the experiment. In addition, raphidophytes were not detected using light microscopy. 20 These results indicate that diatoms might sensitively respond to changes in CO 2 concentration and under high CO 2 environmental conditions their growth relative to other taxonomic groups in the assemblage will be reduced. They are supported by the higher efficiency of carbon fixation by ribulose-1,5-bisphosphate carboxylase-oxygenase (Rubisco) in diatoms than other algal groups (Tortell, 2000). The results of higher diatom 25 contribution in the assemblage in the lowest pCO 2 were also consistent with SiO 2 consumption in each treatment (Fig. 4b). A similar result could not be obtained from our microscopic examination because of the lack of significant data on diatoms identified in the >5 µm fraction. Alternatively, variation of cellular fucoxanthin content in diatoms BGD Introduction

Tables Figures
Back Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion may explain the difference in fucoxanthin/chl-a ratios between the treatments. Increasing fucoxanthin/chl-a ratios were reported with increasing nitrogen limitation (Hou et al., 2007). Light intensity also influences the cellular fucoxanthin content but does not alter the fucoxanthin/chl-a ratio (Goericke and Welshmeyer, 1992). Since nutrient and light conditions were assumed to be equal between the treatments, the change in cellular 5 fucoxanthin content is unlikely to be the cause of the difference in fucoxanthin/chl-a ratios at the end of the experiment. For the effect of CO 2 increase on coccolithophores, no coccolithophores were detected either at the initial or at the end of the experiment with SEM. Cell abundance of heterotrophic bacteria was stable at around 4×10 5 cells mL −1 throughout the incu-10 bation period, and no significant difference in the net change in bacterial abundance was observed between the treatments (Fig. 4d). Synechococcus and autotrophic ultraeukaryotes showed decreasing trends from 9×10 3 to around 2×10 3 cells mL −1 and from 1.8×10 4 to 0.4-1.0×10 4 cells mL −1 , respectively, during the experiment. No significant differences in the net change in the abundances of Synechococcus (Fig. 4e) 15 and ultra-eukaryotes (Fig. 4f) were also observed between the treatments. Effects of pCO 2 change on carbon dynamics appeared more in DOC than in POC. Net accumulation of POC ranged from 1.2±1.4 µmol L −1 in 280 µatm to −1.7±0.5 µmol L −1 in 590 µatm (Fig. 4g). Net POC accumulation in 280 µatm was significantly higher than that in 480 µatm and 590 µatm (Tukey's test, p<0.05). 20 On the other hand, net DOC accumulations decreased from 4.9±2.8 µmol L −1 to 0.49±0.50 µmol L −1 with increasing pCO 2 (Fig. 4h). Net DOC accumulation in 150 µatm was significantly higher than that in 480 µatm and 590 µatm (Tukey's test, p<0.05). Difference in net DOC accumulation of 4.4 µmol L −1 between 150 µatm and 590 µatm was larger than that of POC of 2.9 µmol L −1 between 280 µatm and 590 µatm.

25
The clear trend of decreasing net DOC accumulation with increasing pCO 2 can be explained by changes in production and/or decomposition processes. If bacterial DOC decomposition was not changed between the treatments, DOC production was elevated in the 150 µatm condition. If DOC production did not change between the treat-BGD 6,2009 Impacts of elevated CO 2 on phytoplankton T. Yoshimura et al. Interactive Discussion ments, DOC decomposition may have been enhanced in the 480 µatm and 590 µatm condition. In the present experiment, temporally stable and indistinguishable dynamics in cell abundance of heterotrophic bacteria between the treatments (Fig. 4d) partly support the former case, although we have no data on rate processes such as bacterial production. Furthermore, DOC production can change between the treatments, 5 because the community structure may have changed between the treatments. Similar decreasing trend of fucoxanthin/chl-a ratio and ∆DOC with increasing pCO 2 ( Fig. 4c and 4h) support this. Pigments analysis suggested that relative abundance of diatoms were enhanced in the 150 µatm condition (Fig. 4c). The amount of exudation of photosynthetic products as DOC is highly species-specific (Nagata, 2000). Thus net DOC 10 production can vary under different phytoplankton species compositions. This is likely to be the most reasonable explanation for our results. On the other hand, bacterial DOC decomposition is an enzymatically-driven reaction, so the process would be sensitive to the change in pH. Bacterial cell-associated and free dissolved enzymes to hydrolyze DOC (e.g. α-and β-glucosidase) play a major role in DOC decomposition 15 (Arnosti, 2002). However, there is no studies on pH dependency of the degradative reactions with these enzymes. We need to further clarify the cause of the differences in DOC dynamics under different pCO 2 conditions. The present results suggest that the future direction of nutrient depleted surface ocean biogeochemistry in the subpolar region is less diatom dominance and less DOC 20 production in the high CO 2 world. Diatoms are responsible for the biological carbon pump even in the nutrient depleted low productive season in the subarctic regions (Honda et al., 1997(Honda et al., , 2002Takahashi et al., 2000). Therefore decrease in diatoms production would have negative impact on the marine carbon uptake and thus play a positive feedback on global warming and ocean acidification. The trends observed 25 in the present study lead to the opposite conclusion to various recent studies carried out under nutrient-replete conditions (Riebesell et al., 2007;Tortell et al., 2008). In these previous studies, the growth of diatoms and organic carbon production accelerated under high CO 2 conditions, and it was concluded that ocean acidification may BGD 6,2009 Impacts of elevated CO 2 on phytoplankton T. Yoshimura et al. concentrating mechanisms, including the role of intra-and extracellular carbonic anhydrases, also are key processes for regulating a given species' reaction to pCO 2 increase. For DOC dynamics, on the other hand, previous studies demonstrated that DOC production may be enhanced under high CO 2 . Engel et al. (2004) andRiebesell 10 et al. (2007) revealed that the formation of transparent exopolymer particles (TEP), which are made of the aggregation of dissolved carbohydrates, was accelerated under high CO 2 condition. High CO 2 potentially enhances photosynthetic carbon assimilation through lowering energetic costs associated with carbon acquisition, as found for natural assemblages (Hein and Sand-Jensen, 1997;Tortell et al., 2008) and isolated single strains (Riebesell et al., 1993;Rost et al., 2003;Hutchins et al., 2007). Some parts of the synthesized organic carbon fail to be incorporated into proteins and lipids, and are exuded to outside of the cell as DOC. When the community structure does not change or single isolated species are used in a experiment, DOC production may increase with increasing pCO 2 . However, in the case that the phytoplankton species 20 composition changed in the experimental system, intraspecies difference in DOC production under different pCO 2 conditions may be masked by the interspecies difference in DOC production as concluded for the present experiment. Decrease in DOC production can have significant impacts on biogeochemical cycles of carbon and nutrients through the change in bacterial activity (Thingstad et al., 2008). Decrease in labile 25 DOC production may lower the competitive advantage of bacteria against phytoplankton for inorganic nitrogen and phosphorus, and thus potentially have a positive effect on phytoplankton nutrient acquisition especially in nutrient-depleted environments. Decrease in semi-labile and refractory DOC production may decrease carbon fixation as

BGD Introduction
Conclusions References Tables  Figures   Back  Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion DOC which is transported into the ocean interior. Given that decrease in dissolved carbohydrate production will lower the ambient TEP concentration, which enhances the particulate aggregation, it may down-regulate the sinking POC flux. On the other hand, direct effect of pH reduction is also reported to reduce the TEP-driven particle aggregation and sedimentation processes (Mari, 2008). Therefore the assessment of the 5 impacts of change in DOC and TEP production is not straightforward, but potentially have a significant impact on marine ecosystem and carbon cycle. Thus, the phytoplankton community structure and organic carbon dynamics are fully coupled, so we need a better understanding of the impact of ocean acidification on these critical components of marine biogeochemistry. Previous experimental results 10 have recently been incorporated into a simulation model to assess the impacts of ocean acidification on the biological carbon pump (Hofmann and Schellnhuber, 2009). The present study suggests that responses of planktonic primary producer to pCO 2 increase is different between nutrient-replete and deplete condition. Further research is required based on in situ CO 2 manipulation experiments using a nutrient limited 15 (including iron as a micro-nutrient) plankton ecosystem and results incorporated into global models.  R., Tilbrook, B., Millero, F. J., Peng, T.-H., Kozyr, A., Ono, T., and Rios, A. F.: The oceanic sink for anthropogenic CO 2 , Science, 305, 367-371, 2004. Shirayama, Y. and Thornton, H.: Effect of increased atmospheric CO 2 on shallow water marine benthos, J. Geophys. Res., 110, C09S08, doi:10.1029/2004JC002618, 2005 of phytoplankton and heterotrophic bacteria in the northwest subarctic Pacific to in situ iron fertilization as estimated by HPLC pigment analysis and flow cytometry, Prog. Oceanogr., 64, 167-187, 2005. Suzuki, R. and Ishimaru, T.: An improved method for the determination of phytoplankton chlorophyll using N,N-dimethylformamide, J. Oceanogr., 46, 190-194, 1990.  -Prog. Ser., 236, 37-43, 2002. Tortell, P. D., Payne, C. D., Li, Y., Trimborn, S., Rost, B., Smith, W. O., Riesselman, C., Dunbar, R. B., Sedwick, P., and DiTullio, G. , 39, 1985-1992, 1994. BGD 6,2009 Impacts of elevated CO 2 on phytoplankton T. Yoshimura et al.     , and (C) fucoxanthin (Fucox) to chl-a ratios at the end of the 14-day incubation experiment. Pooled samples of triplicate bottles were analyzed for Fucox/chl-a ratios. Different letters above bars denote significant difference between the treatments (Tukey's test, p<0.05).