Sensitivities of Marine Carbon Fluxes to Ocean Change

Sensitivities of marine carbon fluxes to ocean change

Ulf Riebesell1, Arne Ko¨ rtzinger, and Andreas Oschlies Marine Biogeochemistry, Leibniz Institute of Marine Sciences, IFM-GEOMAR, Du¨sternbrooker Weg 20, 24105 Kiel, Germany

Edited by Hans Joachim Schellnhuber, Potsdam Institute for Climate Impact Research, Potsdam, Germany, and approved August 31, 2009 (received for review December 29, 2008)

Throughout Earth’s history, the oceans have played a dominant role in the climate system through the storage and transport of heat and the exchange of water and climate-relevant gases with the atmosphere. The ocean’s heat capacity is Ϸ1,000 times larger than that of the atmosphere, its content of reactive carbon more than 60 times larger. Through a variety of physical, chemical, and biological processes, the ocean acts as a driver of climate variability on time scales ranging from seasonal to interannual to decadal to glacial–interglacial. The same processes will also be involved in future responses of the ocean to global change. Here we assess the responses of the seawater carbonate system and of the ocean’s physical and biological carbon pumps to (i) ocean warming and the associated changes in vertical mixing and overturning circulation, and (ii) ocean acidiﬁcation and carbonation. Our analysis underscores that many of these responses have the potential for signiﬁcant feedback to the climate system. Because several of the underlying processes are interlinked and nonlinear, the sign and magnitude of the ocean’s carbon cycle feedback to climate change is yet unknown. Understanding these processes and their sensitivities to global change will be crucial to our ability to project future climate change.

climate change ͉ marine carbon cycle ͉ ocean acidiﬁcation ͉ ocean warming

he ocean is presently undergoing mospheric carbon dioxide (8) and has a sidered separately. For conceptual simpli- major changes. Over the past 50 mean residence of a few hundred years. fication, however, we will here focus on years, the ocean has stored 20 Hence, small changes in this pool, caused, physical impacts on the solubility pump times more heat than the atmo- for example, by biological responses to only before discussing impacts on the bio- Tsphere (1). The Arctic Ocean surface lay- ocean change, would have a strong effect logical pump in a later section. ers have recently experienced a pro- on atmospheric CO2. Counteracting the Rising atmospheric CO2 leads to higher nounced freshening (2), and although organic carbon pump in terms of its effect sea-surface temperature (SST) and a con- there is still considerable uncertainty in on air–sea CO2 exchange is a process current reduction in CO2 solubility. It is current observational estimates (3), cli- termed the carbonate counter pump (9), estimated that this positive SST feedback mate models run under increasing atmo- also known as the alkalinity pump. The will reduce the oceanic uptake of anthro- spheric CO2 levels essentially all predict a formation of CaCO3 shell material by cal- pogenic carbon by 9–15% [Ϸ45–70 giga- slowdown of the Atlantic meridional over- cifying plankton and its sinking to depth tons carbon (Gt C)] by the end of the 21st turning circulation (MOC), which is part lowers the DIC and alkalinity in the sur- century (10–13). Global warming will also of the global thermohaline circulation (4). face ocean, causing an increase in CO2 intensify the hydrological cycle. The direct Since preindustrial times, the ocean has partial pressure. It is worth noting that the dilution effect on CO2 solubility is small also taken up Ϸ50% of fossil fuel CO2, organic and inorganic carbon pumps rein- compared with the temperature effect and which has already led to substantial force each other in terms of maintaining a is expected to largely cancel out at the changes in its chemical properties (5). vertical DIC gradient, whereas they are global scale as enhanced precipitation in Carbon fluxes from the sea surface into counteractive with respect to their impact one area tends to be balanced by en- the ocean interior are often described in on air–sea CO2 exchange. hanced evaporation in another area. Over- terms of a solubility pump and a biologi- Although the range of potential all, the expected effect of direct dilution cal pump (6). The abiotic solubility pump changes in the solubility pump and chemi- on anthropogenic CO2 uptake is of uncer- is caused by the solubility of CO2 increas- cal responses of the marine CO2 system is tain sign but estimated to be much smaller ing with decreasing temperature. In known reasonably well, our understanding than 1% (13). A potentially more signifi- present climate conditions, deep water of biological responses to ocean change is cant impact of changes in the hydrological forms at high latitudes. As a result, still in its infancy. Such responses relate cycle on the oceanic CO2 uptake can arise volume-averaged ocean temperatures are both to possible direct effects of rising at high latitudes in the North Atlantic: lower than average sea-surface tempera- atmospheric CO2 through ocean acidifica- Here, reduced surface salinities, together tures. The solubility pump then ensures tion (decreasing seawater pH) and ocean with higher SSTs, would lower the density that, associated with the mean vertical carbonation (increasing CO2 concentra- of surface waters and thereby may inhibit temperature gradient, there is a vertical tion), and indirect effects through ocean the formation of deep waters. This lower- gradient of dissolved inorganic carbon warming and changes in circulation and ing in turn would reduce meridional pres- (DIC). This solubility-driven gradient ex- mixing regimes. These changes are ex- sure gradients and tend to slow down the plains Ϸ30–40% of today’s ocean surface- pected to impact marine ecosystem struc- thermohaline-driven part of the meridi- to-depth DIC gradient (7). ture and functioning and have the poten- onal, overturning circulation. Climate A key process responsible for the re- tial to alter the cycling of carbon and model simulations indeed predict a gen- maining two thirds of the surface-to-depth nutrients in the surface ocean with likely DIC gradient is the biological carbon feedbacks on the climate system. pump. It transports photosynthetically Author contributions: U.R., A.K., and A.O. wrote the paper. fixed organic carbon from the sunlit sur- Changes in the Solubility Pump. Alterations The authors declare no conﬂict of interest. face layer to the deep ocean. Integrated in the physical state of the ocean will af- This article is a PNAS Direct Submission. over the global ocean, the biotically medi- fect both the solubility pump and the bio- 1To whom correspondence should be addressed. E-mail: ated oceanic surface-to-depth DIC gradi- logical pump. For finite-amplitude pertur- [email protected]. ent corresponds to a carbon pool 3.5 bations, changes in both pumps interact This article contains supporting information online at www. times larger than the total amount of at- nonlinearly and therefore cannot be con- pnas.org/cgi/content/full/0813291106/DCSupplemental.

20602–20609 ͉ PNAS ͉ December 8, 2009 ͉ vol. 106 ͉ no. 49 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813291106 Downloaded by guest on September 25, 2021 PCA ETR:PERSPECTIVE FEATURE: SPECIAL

eral weakening of the North Atlantic MOC for the 21st century when forced with increasing greenhouse gas concentrations (4, 14). Complementary to the simulated weakening of the Atlantic MOC, climate models also predict an intensification of the ocean circulation in the South- ern Ocean, resulting from a larger warming in mid and low latitudes compared with the waters around Antarctica. In consequence, the meridional pressure gradients are predicted to increase both across the Antarctic Circumpolar Current (15) and in the atmosphere above the Southern Ocean (16). Estimates of the total impact of the circulation changes on the solubility pump differ considerably among different models, predicting a reduction in global oceanic carbon uptake Fig. 1. Trends and projections of changes in the properties of the marine CO2 system in cold (blue) and warm (red) surface waters between 1750 and 2100 under the prescribed atmospheric CO2 increase. (A) Equilibrium ء by some 3–20% (17). A major part of the CO2 partial pressure (pCO2) and pH. (B) Equilibrium DIC and TA. (C) Concentrations of the three species CO2, ء Ϫ 2Ϫ large uncertainty can be attributed to our HCO3 , and CO3 (note that the concentration of CO2 is multiplied by a factor of 10. (D)CO2-uptake ratio and incomplete understanding of the Southern saturation states (⍀) of calcite and aragonite. (E) Temperature sensitivities of pCO2 and pH. (F) Chemical (i.e., Ocean’s role as a carbon sink. CO2) sensitivities of pCO2 and pH. See SI Text for details on the calculation and assumptions made therein.

Chemical Response of the Marine CO2 Sys- tem. A significant portion of the carbon To illustrate changes in the properties units (Fig. 1A). This increase in the hy- ϩ released by humankind in the Anthropo- of the marine CO2 system during the drogen ion (H ) concentration of up to cene era has been—and will continue to course of the 21st century, we need to 250%, also referred to as ‘‘ocean acidifica- be—redistributed to the world ocean. The prescribe the development of atmospheric tion,’’ is currently of growing concern ocean owes its huge CO2-uptake capacity CO2 concentrations. With vastly differing among marine biologists. Also note that to the presence of the carbonate ion emission scenarios and a host of open it is entirely unclear whether the corre- 2Ϫ questions, we opted for a rather simplistic sponding decrease of the hydroxide ion (CO3 ), which can react with excess CO2 taken up from the atmosphere according approach in which we extrapolated recent concentration by up to 60% could also growth rates to yield an atmospheric CO2 have consequences for chemical reactions to the following net buffering reaction Ϫ (Note that the two analytically indistin- concentration of almost 700 parts per mil- or equilibria involving OH (e.g., solubil- lion by volume (ppmv) by the end of the ity of iron hydroxides); this aspect war- guishable species, CO2 and H2CO3, are usually combined as hypothetical species 21st century (see SI Text for details on the rants a closer look. According to Eq. 1, calculations and all assumptions involved). the uptake of atmospheric CO2 in addi- Ϫ ء :(.CO*2 The oceanic response exerted by this at- tion to CO2 increases HCO3 at the ex- ϩ 2Ϫϩ 3 Ϫ 2Ϫ CO*2 CO3 H2O 2HCO3 [1] mospheric boundary condition will be re- pense of CO3 (Fig. 1C). By converting 2Ϫ gionally different depending mainly on one molecule of CO3 to two molecules The oceanic uptake capacity which Ϫ seawater temperature and the general of HCO3 this reaction is neutral, how- amounts to nearly 85% (not including the characteristics of the marine CO2 system. ever, for total alkalinity (18). buffering by sedimentary carbonates) of 2Ϫ To capture the entire sensitivity range, we The marked drop in CO3 has distinct the entire anthropogenic carbon released define a cold- and warm-water case effects on two important properties of the to date (340–420 Gt C) (5), can only be roughly representing polar and tropical marine CO2 system: First, the CO2 uptake accommodated, however, on the turnover surface waters. In both cases, changes in capacity, expressed as the equilibrium time scale of the ocean, i.e., several centu- CO2 system parameters pCO2, pH, and DIC increase per rise of atmospheric CO2 ries to millennia. The present oceanic up- ␦ ␦ atm DIC as well as concurrent changes in CO2 ( DIC/ xCO2 ), is attenuated to less than take of Ϸ27% of the current anthropo- Ϫ 2Ϫ system speciation (HCO3 ,CO3 ,CO*2) a third of its preindustrial value. Hence, at genic carbon emissions (5) therefore falls were calculated (Fig. 1 A–C). These spe- equilibrium one liter of seawater in 2100 markedly short of the thermodynamic up- ciation shifts also affect other CO2 system will sequester only one third of the atmo- take capacity. As the ocean continues to properties such as the CO2 buffering ca- spheric CO2 (per ppmv increase) it se- take up anthropogenic CO2, significant pacity expressed by the uptake factor questered in 1750 (Fig. 1D). This is a ␦ ␦ atm speciation changes in the marine CO2 sys- ( DIC/ xCO2 ), the saturation state with rather drastic change that must be (and tem take place that lead to a progressive respect to the calcium carbonate minerals usually is) taken into account in future reduction of the buffering capacity. These calcite and aragonite as well as the sensi- projections. Secondly, the ubiquitous su- changes will not only cause a decrease in tivities of pCO2 and pH to thermal and persaturation of surface waters with re- the ocean’s uptake ratio, i.e., the equilib- biologically mediated chemical forcing spect to the calcium carbonate minerals rium change in DIC per change in atmo- (Fig. 1 D–E). calcite and aragonite falls to about a half spheric CO2 concentration, but also lead As expected, surface-ocean DIC in- (50–60%, Fig. 1D) of its preindustrial to an amplification of the effects of ther- creases along with pCO2 as a consequence level. Thus, cold surface waters in high mal and biological forcing on the natural of uptake of ‘‘excess’’ atmospheric CO2 latitudes will be the first to no longer be oceanic carbon cycle. These changes in (Fig. 1B), a process known as ‘‘ocean car- supersaturated and may even turn corro- the marine CO2 system, which we will bonation.’’ The concurrent drop in sur- sive to carbonate minerals (19, 20). explore in the following, are inherently face-ocean pH is a mirror image of the Beyond these prominent and well- nonlinear and will amplify with time. pCO2 increase and amounts to 0.3–0.4 pH documented effects, there are other

Riebesell et al. PNAS ͉ December 8, 2009 ͉ vol. 106 ͉ no. 49 ͉ 20603 Downloaded by guest on September 25, 2021 changes in the properties of the marine A 200 CO2 system that do not appear to have Observed: Körtzinger et al. (2008) Observed (smoothed) received as much attention. These changes 150 Thermal forcing Chemical forcing pertain to the inherent sensitivity of the 100 CO2 system to thermal and chemical forcing, most importantly the impacts of the 50

seasonal cycle of SST and net biogenic anomaly [µatm]

2 0 production of particulate organic and in-

CO -50

organic carbon on the ocean’s source/sink p patterns for atmospheric CO . Large re- 2 -100 gions of the world ocean show marked SST seasonality, which directly affects the -150 Seasonal temperature-sensitive parameters pCO2 -200 and pH [but not the conservative proper- Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0.12 ties DIC and total alkalinity (TA), if B Observed 0.10 expressed in gravimetric units]. As the Thermal forcing 0.08 Chemical forcing anthropogenic perturbation of the marine 0.06 CO2 system enhances the temperature 0.04 sensitivity of pCO2 (␦pCO2/␦T) by almost a factor of 2.5 in our extrapolation (Fig. 0.02 1E), the effect of the natural seasonal 0.00 SST forcing will be strongly amplified, -0.02 leading to marked modulation of the sea- -0.04 -0.06

sonal air–sea CO2 flux pattern. In con- Seasonal pH anomaly trast, seawater pH will only be marginally -0.08 affected with even a slight reduction in -0.10 -0.12 the temperature sensitivity (␦pH/␦T). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec The second major driver of natural seasonal variability in the surface-ocean car- Fig. 2. Seasonal cycle of pCO2 (A) and pH (B) anomalies at present (year 2004, continuous lines) and in bon cycle are biologically driven chemical the future (year 2100, dashed lines) at a location in the central Labrador Sea (56.5° N, 52.6° W). Also shown changes, i.e., withdrawal of carbon are pCO2 observations (from 20) on which this analysis is based. See SI Text for details on the calculation through net production of organic carbon and assumptions made therein. as well as particulate carbonates (e.g., calcite and aragonite). The sensitivities of pCO2 and pH with respect to organic car- cycle features a marked summer minimum whereas the region turns into a CO2 bon production, which to a first-order ap- (winter maximum) because of the domi- source in winter. If we further take into proximation affects DIC but not TA, are nance of biological over thermal forcing account the marked seasonality in wind both amplified by the anthropogenic per- (Fig. 2A). speed, the physical driving force of air–sea turbation. The enhancement by a factor As shown by ref. 23, the observed an- exchange, it becomes obvious that strong net effects on the annual source/sink func- Ͼ3 by year 2100 for pCO2 (␦pCO2/␦DIC) nual pCO2 cycle in the central Labrador is again much larger than for pH (␦pH/ Sea can be deconvoluted into a thermal tion can occur (Fig. 3). In our case, the ␦DIC), which will increase by a factor of (i.e., isochemical) and a chemical (i.e., overall net CO2 sink in 2100 is projected to drop to about half the sink observed only Ϸ1.4 (Fig. 1F). Overall, pCO2 is thus isothermal) component, where most of affected most strongly by changes in phys- the chemical effect is mediated by carbon 2004. ical and (biologically mediated) chemical uptake via net biological production. As- The well-documented climatological forcing, which has implications for the suming no change in the thermal and CO2 source/sink patterns of the present seasonal CO sink/source pattern. The chemical forcing, hypothetical seasonal ocean (21, 22) are largely a consequence 2 of the two big, natural, counteracting seawater pH, in contrast, is much less pCO2 and pH cycles can be constructed affected. for the year 2100 (Fig. 2). The increase in forces, both of which are enhanced as surface-ocean pCO levels rise in con- The thermal and biochemical drivers of the peak-to-peak amplitudes is significant 2 cert with the atmospheric pCO2.As the natural seasonal CO sink/source pat- in both but much stronger in pCO2. Note 2 shown above, the balance between the terns of the ocean have been explored in that although in year 2100 pH is lower two responses—and hence the net detail (e.g., 21–24). Unlike for oxygen, the and shows a larger seasonal amplitude ocean-source/-sink function for the at- two have counteracting effects on pCO . than in 2004, saturation states of calcite 2 mospheric CO budget—will be af- and aragonite show a reduced seasonal 2 In major parts of the world ocean, how- fected, even without the need to invoke amplitude even though they are also ever, one factor dominates, and a clear any changes in the general physical and p seasonal pCO2 cycle results. Especially in lower. At first sight, the CO2 effects may biological forcing of the marine carbon these regions, changing thermal and appear to be a mere intensification of the cycle. If we further include the possibil- chemical sensitivities will give rise to an seasonal pCO2 and hence CO2 source/sink ity of such changes in the forcing itself, alteration of the seasonal cycle of pCO2 cycle with little net affect. In regions which may partly compensate and partly with higher peak-to-peak amplitudes and where on an annual average a disequilib- enhance the responses presented above, potentially also a change in the sign of the rium is observed, however, a significant it becomes clear that the reduction in disequilibrium. To make the superim- net effect may result. The Labrador Sea, the buffering capacity of the ocean’s posed effects of these counteracting pro- for example, is nearly neutral in winter CO2 system is a critical system response cesses more palpable, we illustrate the and a strong sink in summer (Fig. 3) with with significant feedback potential. situation with a realistic example. We an overall annual CO2 sink of 2.7 mol have chosen the Labrador Sea as a high- mϪ2 yrϪ1 in 2004 (23). In our extrapola- Biotic Responses to Ocean Warming. Ocean latitude example where the seasonal pCO2 tion, the summer sink is enhanced warming impacts the pelagic ecosystem

20604 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813291106 Riebesell et al. Downloaded by guest on September 25, 2021 100 13

12 50 ]

11 -1 0 [µatm] 2 10 -50 pCO 9

-100 8 Wind speed [m s Air-sea -150 Present (2004) Future (2100) 7 Wind speed -200 6 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Fig. 3. Monthly mean air–sea disequilibrium at a location in the central Labrador Sea (56.5° N, 52.6° W) in years 2004 (continuous line) and 2100 (broken line). Also shown is monthly mean climatological wind speed at this location.

both directly and indirectly in three ways: formed nutrients into regenerated nutri- (i) through decreased supply of plant nu- ents will be modulated by the concomi- trients because of a slowdown in vertical tant reduction in the supply of the often mixing and convective overturning; (ii) limiting micronutrient iron (26) under Fig. 4. Schematic model illustrating the effect of through increased thermal stratification, changing light conditions (27). sea-surface warming on upper-ocean processes in causing increasing light availability for Increased thermal stratification due to low (Upper) and high (Lower) latitudes. Graph de- photosynthetic organisms suspended in rising SST affects both nutrient supply picts effects on stratification and mixed-layer depths, nutrient supply (yellow arrows), plankton the upper mixed layer; and (iii) through and mixed-layer light intensities. In the biomass, and particle flux (green arrows). the effect of increasing temperatures tropics and midlatitudes, where thermal on the rates of biological processes. Which stratification restricts vertical mixing, typi- of these factors will dominate in a particu- cally low surface nutrient concentrations a complex nonlinear response to ocean lar region and at a given time depends on limit phytoplankton growth. Ocean warm- warming at the community and ecosystem the hydrographic conditions, the composi- ing will further reduce mixing, diminishing levels. Although shoaling of the upper tion of the pelagic community, and the the upward nutrient supply and lowering mixed layer primarily affects the autotro- activities of its individual components. productivity (Fig. 4 Upper). At higher lati- phic community through increased light tudes, phytoplankton is often light-limited intensities in this layer, increasing seawa- Changes in Circulation and Mixing. A re- because intense vertical mixing circulates ter temperature mainly affects the activity duction in the overturning circulation can algal cells over deep mixed layers, result- of the heterotrophic community. The net be expected to reduce the supply of nutri- ing in lower mean light intensity along a outcome depends on the balance of light ents to the surface waters. In areas in phytoplankton cell’s trajectory and hence and temperature sensitivities and dynamic which surface nutrient concentrations are lower net productivity. In these regions, interaction of the different groups of already zero, this reduction will not ocean warming and a greater influx of auto- and heterotrophic organisms and change surface nutrient concentrations fresh water, mostly from increased precip- cannot be easily deduced from physiology- and, for constant stoichiometry, surface itation and melting sea ice, will contribute based first principles. DIC levels. Despite a possible reduction to reduce vertical mixing which may in- A number of different approaches con- in biomass and its turnover, the first-order crease productivity (Fig. 4 Lower). sistently predict an overall decrease in effect on air–sea CO2 exchange will be primary and export production in the small in these areas. Regions with non- Temperature Effects on Biological Activities. tropics and midlatitudes and a pole-ward zero surface nutrient concentrations, how- Direct biotic responses to increasing tem- migration of geographic boundaries sepa- ever, can experience declining surface peratures will differ greatly among the rating biogeochemical provinces: nutrient and DIC concentrations for a various components of the pelagic ecosys- (i) coupled atmosphere ocean general slower circulation, which will allow more tem, most notably between the autotro- circulation models (AOGCMs) combined atmospheric CO2 to enter the ocean. phic and heterotrophic communities. The with mechanistic models of pelagic ecosys- Without exchange with the underlying temperature dependence of biological tems that directly predict biological re- waters, utilization of the nutrients in the processes is commonly expressed by the sponses (e.g., 31–33); upper 50 m of the global ocean would Q10 factor—the factorial increase in the (ii) climate response patterns in the correspond to a one-time additional car- rate for a 10° C increase in temperature. AOGCMs that are similar to observed bon uptake of Ϸ8 Gt C. Subduction of Although bacterial heterotrophic activities modes of interannual variability such as the newly nutrient-depleted waters and typically have a Q10 factor between 2 and the El Ninõ/Southern Oscillation or Pa- replacement by nutrient-rich subsurface 3 (28), phytoplankton growth and photo- cific Decadal Oscillation (34, 35); and waters can, however, allow for a repeated synthesis show only a moderate tempera- (iii) empirical models based on observa- Ͻ Ͻ uptake of atmospheric carbon. This pro- ture sensitivity (1 Q10 2) (29) and are tional estimates of physical controls on cess eventually leads to a reduction of pre- primarily controlled by incident light in- current distributions of chlorophyll and formed nutrients and a corresponding intensity. Bacterial growth efficiency, on the primary production (36). crease of regenerated nutrients and other hand, is an inverse function of tem- carbon below the surface layer at the ex- perature, causing an increased fraction of Biotic Responses to Ocean Acidification pense of a decline in the atmospheric car- the assimilated carbon to be respired with (OA). Oceanic uptake of anthropogenic bon pool (25). Particularly in the Southern rising temperature (30). These diverging CO2 and the resulting changes in seawater Ocean, the dynamics of converting pre- temperature sensitivities will likely induce chemistry will, in the course of this cen-

Riebesell et al. PNAS ͉ December 8, 2009 ͉ vol. 106 ͉ no. 49 ͉ 20605 Downloaded by guest on September 25, 2021 tury, expose marine organisms to condi- Table 1. Biotic responses to sea surface warming and ocean carbonation/acidification tions which they may not have experi- and their feedback potential to the climate system enced during their recent evolutionary Feedback process Sign of feedback Sensitivity Capacity Longevity history (37). The impact these changes will have on marine life is still uncertain, Responses to ocean warming but there are likely to be both winners Nutrient supply to oligotrophic ocean ϩϩϩ Ϯ0 ϩϩ Nutrient supply to HNLC areas negative ϩϩ ϩϩ ϩϩϩ and losers. In fact, with regard to the po- ϩϩ ϩϩ ϩϩϩ tential biological effects of oceanic CO Nutrient utilization efficiency negative 2 Nutrient inventory positive ϩϩ ϩϩϩ uptake, the term ‘‘ocean acidification’’ Organic matter remineralisation positive only encompasses one side of the story. Responses to ocean acidification Although the decrease in seawater pH Calcification negative ϩ† ϩϩ‡ may pose a threat to the fitness of pH- Ballast effect positive ϩϩϩ ϩϩϩ sensitive groups, in particular calcifying Stoichiometry negative ϩϩ† ϩϩ ϩϩ organisms, ‘‘ocean carbonation’’—the in- Extracellular organic matter production negative ϩϩ† ϩϩϩ ‡ Nitrogen fixation negative ϩϩ† ϩ ‡ creasing oceanic CO2 concentration—will likely be beneficial to some groups of Responses are characterized with regard to feedback sign, sensitivity, capacity, and longevity by using photosynthetic organisms, particularly best guesses; we use Ϯ0 for negligible, ϩ for low, ϩϩ for moderate, and ϩϩϩ for high. Empty boxes those that operate a relatively inefficient indicate missing information/ understanding. † CO2 acquisition pathway. Available information mainly based on short-term perturbation experiments. Adverse effects of ocean acidification ‡Potential for adaptation presently unknown. HNLC, high nutrient, low chlorophyll. are expected on various groups of calcifying organisms, including corals, bivalves, gastropods, and sea urchins, and possibly ratus and carbon-enrichment systems (58), standing of biologically driven feedback it is presently difficult to assess which of mechanisms is still rudimentary, so the on the many other groups that use CaCO3 as internal or external structural elements, these responses are specific to individual assigned values mostly represent best such as crustaceans, cnidaria, sponges, taxa or represent general phenomena guesses. In some cases, we feel that our bryozoa, annelids, brachiopods, tunicates, across phytoplankton taxa or within func- understanding is too immature to even squid, and fish (38). The loss of competi- tional groups. make a guess. tive fitness in these groups may lead to Responses to Ocean Warming and Circula- the loss of biodiversity and the restructur- Uncertainties. It is important to note that tion. The effects of sea-surface warming ing of marine ecosystems. To what extent our present knowledge of pH/CO2 sensi- such changes would affect marine produc- tivities of marine organisms is based al- on the marine biota and the resulting im- tivity, transfer of energy through the food most entirely on short-term perturbation pacts on marine carbon cycling will differ web, or biogeochemical cycling is pres- experiments, neglecting the possibility of depending on the prevailing light and nu- ently unknown. Of relevance to the car- evolutionary adaptation. Also, little is trient conditions in the upper mixed layer bonate pump are the pelagic calcifying presently known regarding the effects and the composition of the pelagic from multiple and interacting stressors, community. groups, in particular the coccolithophores, ⅐ foraminifera, and pteropods. Most studies such as sea-surface warming and eutrophi- Decreased nutrient supply to already- to date indicate a decrease in calcification cation. Moreover, there is a complete lack nutrient-limited areas of the ocean’s sur- of these groups with increasing seawater of information on the transfer of re- face layer is certain to lead to dimin- acidification, both in laboratory experi- sponses from the organism to the commu- ished primary production (sensitivity ϩϩϩ ments (e.g., 39–42) and field studies (43, nity and ecosystem levels and the replace- ) and to a decline in the strength 44). However, this point remains contro- ment of OA-sensitive by OA-tolerant of the biological pump, i.e., the amount versial (45, 46). species. of biogenic carbon transported to depth. Stimulating effects of ocean carbon- On the other hand, as nutrient supply to ation have been primarily observed for Impacts on the Biological Carbon Pumps. the surface layer slows down, so does processes related to photosynthetic activ- For a comparison of the different biologi- the supply of dissolved inorganic carbon, cal responses to ocean change and their slowing down the rate at which deep ity. These effects include CO2-enhanced rates of phytoplankton growth and carbon potential feedback to the climate system, ocean CO2 is brought back into contact fixation (47, 48), organic matter produc- we will apply four criteria to characterize with the atmosphere. Because the tion (49, 50), and extracellular organic each of the processes: change in preformed nutrients is negligi- (i) the sign of the feedback, denoting matter production (51). CO2 sensitivity of ble in nutrient-limited areas, the overall autotrophic processes is also evident from whether a feedback amplifies (positive effect on carbon sequestration can be changes in the cellular carbon:nitrogen: feedback) or dampens (negative feedback) expected to be negligible (capacity Ϯ0) phosphorus (C:N:P) ratios of marine mi- the initial forcing; unless there is a change in C:N ratios in croalgae with CO2 concentration (52). A (ii) the sensitivity, indicating the change export production. shift in the ratio of carbon to nutrient needed to trigger the response; ⅐ Decreased nutrient supply to nutri- drawdown toward higher C:N and C:P at (iii) the capacity, representing the ent-replete (e.g., high-nutrient, low- elevated pCO2 was observed during a me- feedback strength relative to the forc- chlorophyll) areas, may lead to lower socosm study in a natural plankton com- ing and relative to other feedback levels in surface nutrients and DIC and munity (53). Recent studies with the dia- mechanisms; and in consequence reduce the total zotrophic cyanobacterium Trichodesmium (iv) the longevity, referring to the amount of preformed nutrients. Associ- revealed an increase in both carbon and length of time a feedback may operate ated with these changes is an increase nitrogen fixation with increasing pCO2 (i.e., transient versus long-lasting). in regenerated nutrients and carbon in (54–56). As primary producers in the ma- In categories ii, iii, and iv,weuseϮ 0 the ocean interior (sensitivity ϩϩ), rine realm encompass very phylogeneti- for negligible, ϩ for low, ϩϩ for moder- resulting in a negative feedback of cally diverse groups of organisms (57) that ate, and ϩϩϩ for high values (Table 1). moderate capacity (ϩϩ), which can differ widely in their photosynthetic appa- We wish to stress that our present under- last for several hundred years as long

20606 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813291106 Riebesell et al. Downloaded by guest on September 25, 2021 as the supply of old preformed nutri- ocean warming. (Capacity ϩϩ, longevity sequestration (65) (positive feedback). ents to the surface ocean from below ϩϩϩ.) The capacity of this feedback depends exceeds the subduction of new pre- ⅐ Although sea-surface warming directly on the pH sensitivity of pelagic calcifiers formed nutrients (longevity ϩϩϩ). affects all biological rates, the higher tem- and the quantitative importance of ⅐ Increased nutrient-utilization efficiency perature sensitivity of heterotrophic (rela- CaCO3 as ballast for particle export, in the high latitudes, on the other hand, tive to autotrophic) processes is expected both of which are poorly understood. If has a potentially strong effect on the effi- to shift the balance between primary pro- CaCO3 turns out to be a prerequisite for ciency of the biological pump (36) (capac- duction and respiration/remineralization in deep transport of particulate organic ity ϩϩ). With the shoaling of mixed-layer favor of the latter (63). Additionally, matter, and if calcification of pelagic depths, currently nonused inorganic nutri- warming shifted the partitioning of or- calcifiers remains sensitive to high CO2, ents in high-latitude surface waters can be ganic carbon between the particulate and this feedback could have a potentially combined with CO2 to form organic mat- dissolved phase toward an enhanced accu- high capacity (ϩϩϩ) and extended du- ter and be transported to depth (negative mulation of dissolved organic carbon (63). ration (ϩϩϩ). feedback). This change in efficiency af- This shift may cause a decrease in the ⅐ A negative feedback also arises with in- fects the partitioning of the nutrient pool vertical flux of particulate organic matter creasing C:N drawdown, as observed in between preformed and regenerated nu- and a decrease in remineralization depth. response to CO2 enrichment in a meso- trients and the associated partitioning of The overall effect could be a decline in cosm experiment (53). The capacity for carbon among the atmospheric and oce- the strength and efficiency of the carbon this process depends on the relative in- anic carbon reservoirs (25). In the deeply pump (positive feedback). Because of the crease in carbon consumption in excess of mixed high-latitude areas of the northern complexity of plankton trophic interac- the Redfield ratio, which based on the North Atlantic and Southern Ocean, a tions it is difficult, however, to assess the observed 27% increase in response to a significant shoaling of the mixed-layer sensitivity and capacity of this feedback. doubling in present day CO2 can be con- depth is required to have a noticeable ef- Also unknown is the extent to which sidered moderate to high (capacity ϩϩ). fect on nutrient utilization efficiency (sen- changes in community composition may It is not presently known whether the re- sitivity ϩϩ). Enhanced surface layer strat- compensate for the shifting balance be- sponse observed for the mesocosm phyto- ification in midlatitudes may also speed up tween auto- and heterotrophic processes, plankton community, which was domi- the biomass accumulation during seasons making it impossible to make a judgement nated by diatoms (mostly Skeletonema sp. of high productivity, e.g., during spring- on the longevity of this feedback. and Nitzschia spp.) and the coccolith- bloom development, possibly affecting ophore Emiliania huxleyi, also applies to particle aggregation and sinking and the Ocean Acidification/Carbonation. Our un- other phytoplankton taxa and functional depth of remineralization. Provided that derstanding of the biological responses to groups, or whether this response may be preformed nutrients continue to be sup- CO2-induced changes in seawater carbon- modified or lost because of adaptation to plied to the surface layer, this feedback ate is still in its infancy. As research on high CO2. Hence, it appears too early to could last for several hundred years (lon- ocean acidification gains momentum, new, make a judgement on the sensitivity and gevity ϩϩϩ). unforeseen pH/CO2 sensitivities emerge, longevity of this feedback. If there is a ⅐ As sea-surface warming reduces deep- with partly opposing impacts on carbon significant enhancement in the C:N draw- ocean ventilation, this slowdown will lower sequestration in the ocean. down at the surface that translates into an the supply of oxygen to the ocean interior ⅐ Decreased pelagic calcification and the enhanced carbon export, it will lead to (59). This process, which has been termed resulting decline in the strength of the higher oxygen consumption and a possible ‘‘ocean deoxygenation’’ (60), is expected carbonate pump lower the drawdown of expansion of oxygen-minimum zones (73), to cause an overall decrease in deep- alkalinity in the surface layer, thereby in- linking to the oxygen-related feedback ocean oxygen content, including an expan- creasing the uptake capacity for atmo- described in Responses to Ocean Warming sion of oxygen-minimum zones (61, 62). spheric CO2 in the surface layer. Assum- and Acidification. Suboxic and anoxic conditions favor pro- ing an annual rate of CaCO3 export of Ϸ1 ⅐ Increased production of extracellular cesses such as denitrification and anaero- Gt C (64), the capacity of this negative organic matter under high CO2 levels bic ammonium oxidation, leading to the feedback is relatively small (65–68) (ca- (51) may enhance the formation of par- loss of bioavailable nitrogen in the ocean, pacity ϩ). The sensitivity of this response ticle aggregates (74, 75) and thereby with possible implications for marine pri- in coccolithophores appears to be highly increase the vertical flux of organic mat- mary production. Provided that the species-dependent (69), permitting ter (negative feedback) (76). This re- ocean’s nitrogen inventory ultimately de- changes in species composition to dampen sponse may in fact have been responsi- termines the amount of carbon biologi- or eliminate the response at the commu- ble for the observed increased in C:N:P cally sequestered in the ocean, reducing nity level (sensitivity ϩ). Shifting species stoichiometry observed at elevated the nitrogen inventory would provide a composition and the potential for adapta- pCO2 (53, 77). Provided that the re- positive feedback to the climate system. tion (presently unknown) could make this sponse is of general nature in bloom- Suboxic conditions at the seafloor, on the a transient feedback (longevity ϩ). In forming phytoplankton (presently un- other hand, favor the release of phosphate view of the scarcity of information on pH known sensitivity), the capacity of this from sedimentary iron phosphates, possi- sensitivities of foraminifera and ptero- process may be high because, similar to bly increasing the ocean’s content of bio- pods, however, it is premature to specu- the ballast effect, it could lead to an available phosphorus. A shift in the late on the significance of this feedback increase in remineralization depth (ca- ocean’s nitrate-to-phosphate content process. pacity ϩϩϩ). Again, not knowing could affect the composition and produc- ⅐ CaCO3 may act as ballast in particle whether evolutionary adaptation will tivity of marine primary producers. Most aggregates, accelerating the flux of par- select against this response prevents a notably, the extra phosphate may be used ticulate material to depth (70, 71; but judgement on the longevity of this by diazotrophic cyanobacteria, possibly see also ref. 72). Reduced CaCO3 pro- process. counterbalancing the nitrogen loss from duction could therefore slow down the ⅐ Enhanced nitrogen fixation at elevated denitrification by nitrogen fixation. The vertical flux of biogenic matter to depth, pCO2, as recently reported for the dia- net outcome of these complex interactions shoaling the remineralization depth of zotrophic cyanobacterium Trichodes- is difficult to assess, as is the sensitivity to organic carbon and decreasing carbon mium (54–56), has the potential to in-

Riebesell et al. PNAS ͉ December 8, 2009 ͉ vol. 106 ͉ no. 49 ͉ 20607 Downloaded by guest on September 25, 2021 crease the reservoir of bioavailable port, ocean acidification might, via a CO2 emissions, 70% of the presently nitrogen in the surface layer, which in a reduction in CaCO3 formation, lead to known reef locations will be in corrosive predominantly nitrogen-limited ocean a substantial decrease of the efficiency waters by the end of this century (82). would increase primary production and of the biological pump. This decrease Coral reefs provide the habitat for the carbon fixation (negative feedback). Be- would lead to a change in upper-ocean ocean’s most diverse ecosystems, are the cause cyanobacteria-produced biomass is nutrient and oxygen status (68) with breeding grounds for commercially impor- generally thought to not be exported to likely impacts on the pelagic commu- tant fish, protect shorelines in tropical great depth, uncertainty exists about the nity structure. The reduction in export areas from erosion and flooding, and gen- extent to which the additional bioavail- would also present a strong positive erate billions of dollars annually in able nitrogen would lead to enhanced feedback on atmospheric CO2 (68) tourism. carbon sequestration or whether it and, in turn, on ocean acidification would remain suspended in the surface (65). However, the postulated increase Summary. The global carbon cycle is both layer, eventually eliminating the ecologi- in the production of extracellular or- driven by and a driver of Earth’s climate cal niche for nitrogen-fixing cyanobacte- ganic carbon under high CO2 levels system. In this system, climate change ria. As CO2 sensitivity of nitrogen fixa- may partly compensate for the loss of therefore goes hand in hand with a tion has thus far been reported for only CaCO3 ballast by favoring aggregation change in carbon cycling and a redistribu- one species, it is too early to speculate of organic matter and non-CaCO3 min- tion of reactive carbon among the carbon about the sensitivity and longevity of erals (72). Given the current state of reservoirs in the atmosphere, terrestrial this response. our knowledge, the efficiency of the biosphere, and ocean. The distribution of biological pump could change signifi- carbon between these reservoirs is the Tipping Points. Although several of the cantly in either the positive or negative result of a multitude of interconnected cause-and-effect relationships described direction. physical, chemical, and biological pro- above have the potential to drastically al- (iii) With regard to ocean acidification, cesses, many of which are sensitive to cli- ter the ocean’s ecosystems and biogeo- a distinct tipping point may arise when mate change themselves. A major chal- chemistry, it is yet unclear if any of the seawater turns undersaturated with re- lenge in Earth system science is predicted changes may turn out irrevers- spect to calcium carbonate and becomes determining which of these processes act ible or may develop into runaway biogeo- corrosive for the shells and skeletons of as primary drivers in the natural climate chemical feedbacks. In the following, we calcifying organisms. In the Southern cycle and how they interact in controlling will give four examples wherein sensitivi- Ocean, for example, wintertime under- carbon fluxes between the reservoirs. ties of ecological and biogeochemical pro- saturation for aragonite is projected to The same challenge applies when trying cesses to ocean change may develop into occur at CO2 levels of 450 ppmv, which to forecast the system’s response to major tipping points, possibly driving the ocean under business-as-usual (Intergovernmenal perturbations of the carbon cycle, such as to a new state. Panel on Climate Change IS92a scenario) the release of CO2 from fossil-fuel burn- (i) A decrease of nutrient supply CO2 emissions will occur by 2030 and no ing and land-use changes. At first sight, to the surface layer because of sea- later than 2038 (79). Wintertime satura- the ocean’s role in this system appears to surface warming in presently production states are of particular relevance to be that of a giant buffer, sequestering tive areas of the ocean may stimulate a one of the key species of polar ecosys- enormous amounts of CO2 and thereby dominance shift in the phytoplankton tems, the aragonite-producing pteropod dampening CO2-induced climate change. community from large diatoms to Limacina helicina. As this species under- A closer look reveals that both the chang- small-celled flagellates and cyanobacte- goes its larval development primarily during climate and the extra load of CO2 se- ria. A shift to smaller-celled phyto- ing winter months, it may experience cor- questered by the ocean alter the oceanic plankton causes an extension of the rosive seawater conditions at a particularly carbon cycle to the extent of modifying its marine food web by one or more tro- sensitive phase of its life cycle within the capacity for further uptake of anthropo- . phic levels (78). With 90% of the bio- next couple of decades. As Limacina is an genic CO2 The net outcome of these mass and energy lost from one trophic important component in polar ocean food modifications is still uncertain, largely due level to the next, this shift would dras- webs, linking lower levels of the food web to our limited understanding of the under- tically reduce the efficiency of energy to the top predators, its disappearance lying mechanisms. This uncertainty holds transfer from the level of primary pro- could mark the disruption of polar pelagic true particularly for those processes in- ducers to the top predators. Accelerat- ecosystems. volving biologically mediated components, ing heterotrophic over autotrophic pro- (iv) For coral reef ecosystems, a tipping which because of the complexity and plas- cesses in a warming ocean will further point will be reached when reef erosion ticity of biotic responses and interactions reduce the transfer efficiency of prima- exceeds reef accretion. At CO2 levels of are extremely difficult to untangle. ry-produced organic matter up the 560 ppmv, calcification of tropical corals is Progress in our understanding of these food chain. Moreover, as the growth expected to decline by 30% (80, 81). At interacting processes and their sensitivities efficiency of bacteria is an inverse this stage, loss of coral structure in areas to ocean change requires the concerted function of temperature, a larger frac- of high erosion may outpace coral calcifi- effort of all relevant disciplines, from mo- tion of assimilated carbon is respired cation, in which case reefs will no longer lecular and ecosystem biology, at elevated temperatures (30, 63). As be sustainable. As the carbonate satura- marine and atmospheric chemistry, physi- these responses all contribute to accel- tion horizon—the depth below which cal- cal oceanography and palaeoceanography, erating the spinning of the wheel at cium carbonate dissolves—shallows be- to atmosphere–ocean- and Earth-system the lower trophic levels, a warmer and cause of ocean acidification, cold-water modeling, and with close interactions be- more stratified ocean may shift into a corals become exposed to corrosive wa- tween observationalists, experimentalists, new and lower state of energy-transfer ters. These slow-growing corals, which and modelers. efficiency, involving reduced fish pro- inhabit cold waters down to 3,000-m water duction. depth, have built extensive reef systems on ACKNOWLEDGMENTS. This work was supported (ii) Depending on the still incom- the shelves and along the continental mar- by the European Union Integrated Project Carbo- pletely understood role of CaCO3 as gins extending from Northern Norway to Ocean “Marine carbon sources and sinks assess- mineral ballast for organic carbon ex- the west coast of Africa. With unabated ment” contract no. 511176 (GOCE).

20608 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813291106 Riebesell et al. Downloaded by guest on September 25, 2021 1. Levitus S, Antonov J, Boyer T (2005) Warming of the 27. de Baar, HJ, et al. (2005) Synthesis of iron fertilization 56. Ramos JBe, Biswas H, Schulz KG, LaRoche J, Riebesell world ocean, 1995–2003. Geophys Res Lett 32:L020604 experiments: From the iron age in the age of Enlight- U(2007) Effect of rising atmospheric carbon dioxide on 10.1029/2004GL021592. enment, J Geophys Res 110, C09S16, 10.1029/ the marine nitrogen fixer Trichodesmium. Global Bio- 2. Curry R, Mauritzen C (2005) Dilution of the northern 2004JC002601. geochem Cycles 21:GB2028, 10.1029/2006GB002898. North Atlantic Ocean in recent decades. Science 28. Pomeroy LR, Wiebe WJ (2001) Temperature and sub- 57. Falkowski PG, et al. (2004) The evolution of modern 308:1772–1774. strates as interactive limiting factors for marine het- eukaryotic phytoplankton. Science 305:354–360. 3. Cunningham SA, et al. (2007) Temporal variability of erotrophic bacteria. Aquat Microb Ecol 23:187–204. 58. Giordano M, Beardall J, Raven JA (2005) CO2 concen- the Atlantic meridional overturning circulation at 29. Eppley RW (1972) Temperature and phytoplankton trating mechanisms in algae: Mechanisms, environ- 26.5° N. Science 317:935–938. growth in the sea. Fish Bull 70:1063–1085. mental modulation, and evolution. Ann Rev Plant Biol 4. Gregory JM, et al. (2005) A model intercomparison of 30. Rivkin RB, Legendre L (2001) Biogenic carbon cycling in 56:99–131. the Upper Ocean: Effects of microbial respiration. Sci- 59. Bopp L, Le QuereĆ, Heimann M, Manning AC, Monfray changes in the Atlantic thermohaline circulation in ence 291:2398–2400. P(2002) Climate-induced oceanic oxygen fluxes: Impli- response to increasing atmospheric CO concentration. 2 31. Cox PM, Betts RA, Jones CD, Spall SA, Totterdell IJ (2000) cations for the contemporary carbon budget. Global Geophys Res Lett 32: L12703. Acceleration of global warming due to carbon cycle Biogeochem Cycles 16:1022, 10.1029/2001GB001445. 5. Sabine CL, et al. (2004) The Oceanic sink for Anthropo- feedbacks in a 3D coupled model. Nature 408:184–187. 60. Keeling RF, Ko¨rtzinger A, Gruber N (2009) Ocean de- genic CO2. Science 305:367–371. 32. Bopp L, et al. (2001) Potential impact of climate change oxygenation in a warming world. Ann Rev Mar Sci,in 6. Volk T, Hoffert MI (1985) in The Carbon Cycle and on marine export production. Global Biogeochem Cy- press. Atmospheric CO2: Natural Variations Archean to cles 15:81–100. 61. Matear RJ, Hirst AC (2003) Long term changes in dis- Present, Geophysical Monogr Ser, eds Sunquist ET, 33. Christian JR, Verschell MA, Murtugudde R, Busalacchi solved oxygen concentrations in the ocean caused by Broecker WS (Washington, DC), Vol 32, pp 99–111. AJ, McClain CR (2002) Biogeochemical modelling of the protracted global warming. Global Biogeochem Cycles 7. Toggweiler JR, Gnanadesikan A, Carson S, Murnane R, tropical Pacific Ocean: I. Seasonal and interannual vari- 17:1125, 10.1029/2002GB001997. Sarmiento JL (2003) Representation of the carbon cycle ability. Deep-Sea Res II 49:509–543. 62. Stramma L, Johnson GC, Sprintall J, Mohrholz V (2008) in box models and GCMs: 1. Solubility pump. Global 34. Boyd PW, Doney SC(2002) Modelling regional responses Expanding oxygen-minimum zones in the tropical Biogeochem Cycles 17:1026, 10.1029/2001GB001401. by marine pelagic ecosystems to global climate change. oceans. Science 320:655–658. 8. Gruber N, Sarmiento JL (2002) in The Sea: Ideas and Geophys Res Lett 29:1806, 10.1029/2001GL014130. 63. Wohlers J, et al. (2009) Changes in biogenic carbon flow Observations on Progress in the Study of the Seas, eds 35. Behrenfeld MJ, et al. (2006) Biospheric primary produc- in response to sea surface warming. Proc Nat Acad Sci Robinson AR, McCarthy JJ, Rothschild BJ (Wiley, New tion during an ENSO Transition. Nature 444:752–755. 106:7067–7072. York), vol. 12, pp 337–399. 36. Sarmiento JL, et al. (2004) Response of ocean ecosys- 64. Milliman JD (1993) Production and accumulation of 9. Heinze C, Maier–Reimer E, Winn K (1991) Glacial pCO2 tems to climate warming. Global Biogeochem Cycles calcium carbonate in the ocean: budget of a nonsteady reduction by the world ocean: Experiments with the 18:GB3003, 10.1029/2003GB002134. state. Global Biogeochem Cycles 7:927–957. Hamburg carbon cycle model. Paleoceanography 37. Raven J, et al. (2005) Ocean acidification due to increas- 65. Heinze C (2004) Simulating oceanic CaCO3 export pro- 6:395–430. ing atmospheric carbon dioxide. (Royal Society, Lon- duction in the greenhouse. Geophys Res Lett 10. Sarmiento JL, Le Que´re´ C (1996) Oceanic carbon dioxide don) Royal Society Report, Policy Document 12/05. 31:L16308, 10.1029/2004GL020613. uptake in a model of century-scale global warming. 38. Fabry VJ, Seibel BA, Richard AF, Orr JC (2008) Impacts of 66. Gehlen M, et al. (2007) The fate of pelagic CaCO3 ocean acidification on marine fauna and ecosystem production in a high CO ocean: a model study. Bio- Science 274:1346–1350. 2 processes. ICES J Mar Science, Journal du Conseil 2008, geosciences 4:505–519. 11. Joos F, Plattner G-K, Stocker TF, Marchal O, Schmittner A 65:414–432 67. Ridgwell A, Zondervan I, Hargreaves J, Bijma J, Lenton (1999) Global warming and marine carbon cycle feed- 39. Bijma, JH, Spero J, Lea DW, Bemis BE(1999) in Use of T (2007) Assessing the potential long-term increase of backs on future atmospheric CO2. Science 284:464–467. Proxies in Paleoceanography: Examples from the South oceanic fossil fuel CO2 uptake due to ‘‘CO2-calcification 12. Matear RJ, Hirst AC (1999) Climate change feedback on Atlantic, eds Fischer G, Wefer, G. (Springer, Berlin), pp feedback’’. Biogeosciences 4:481–492. the future oceanic CO2 uptake. Tellus 51B:722–733. 489–512. 68. Hofmann M, Schellnhuber HJ (2009) Oceanic acidifica- 13. Plattner GF, Joos F, Stocker TF, Marchal O (2001) Feed- 40. Riebesell U, et al. (2000) Reduced calcification in marine tion affects marine carbon pump and triggers ex- back mechanisms and sensitivities of ocean carbon up- plankton in response to increased atmospheric CO2. tended marine oxygen holes. Proc Nat Acad Sci USA take under global warming. Tellus 53B:564–592. Nature 407:634–637. 106:3017–3022. 14. Schmittner A, Latif M, Schneider B(2005) Model projec- 41. Sciandra A, et al. (2003) Response of coccolithophorid 69. Langer G, et al. (2006) Species-specific responses of tions of the North Atlantic thermohaline circulation for Emiliania huxleyi to elevated partial pressure of CO2 un- calcifying algae to changing seawater carbonate chem- the 21st century assessed by observations. Geophys Res der nitrogen limitation. Mar Ecol Progr Ser 261:111–122. istry. Geochem Geophys Geosyst 7:Q09006, 10.1029/ Lett 32:L23710, 10.1029/2005GL024368. 42. Feng Y, et al. (2008) Interactive effects of increased 2005GC001227. 15. Schmittner A, Oschlies A, Matthews HD, Galbraith ED pCO2, temperature and irradiance on the marine coc- 70. Armstrong RA, Lee C, Hedges JI, Honjo S, Wakeham SG (2008) Future changes in climate, ocean circulation, colithophore Emiliania huxleyi (Prymnesiophyceae). (2002) A new, mechanistic model for organic carbon ecosystems and biogeochemical cycling simulated for a Eur J Phycol 43:87–98. fluxes in the ocean based on the quantitative associa- business-as-usual CO2 emission scenario until 4000 AD. 43. Delille B, et al. (2005) Response of primary production tion of POC with ballast minerals. Deep-Sea Res Global Biogeochem Cycles GB1013: doi:10.1029/ and calcification to changes of pCO2 during experimen- 49:219–236. 2007GB002953. tal blooms of the coccolithophorid Emiliania huxleyi. 71. Klaas C, Archer DA (2002) Association of sinking 16. Russell JL, Dixon KW, Gnanadesikan A, Stouffer RJ, Global Biogeochem Cycles 19:GB2023, 10.1029/ organic matter with various types of mineral ballast Toggweiler JR (2006) The southern hemisphere west- 2004GB002318. in the deep sea: Implications for the rain ratio. Global erlies in a warming world: Propping open the door to 44. Engel A, et al. (2005) Testing the direct effect of CO2 Biogeochem Cycles 16:1116, 10.1029/2001GB001765. the deep ocean. J Clim 19:6382–6390. concentration on a bloom of the coccolithophorid 72. Passow U (2004) Switching perspectives: Do mineral Emiliania huxleyi in mesocosm experiments. Limnol fluxes determine particulate organic carbon fluxes or 17. Greenblatt JB, Sarmiento JL(2004) in The Global Car- Oceanogr 50:493–504. vice versa? Geochem Geophys Geosyst 5:1–5. bon Cycle, eds Field CB, Raupach MR (Island Press, 45. Iglesias-Rodriguez MD, et al. (2008) Phytoplankton cal- 73. Oschlies A, Schulz KG, Riebesell U, Schmittner Washington, D.C.), pp 257–275. cification in a high-CO world. Science 320:336–340. A(2008) Simulated 21st century’s increase in oceanic 18. Wolf–Gladrow DA, Zeebe RE, Klaas C, Ko¨rtzinger A, 2 46. Riebesell U, et al. (2008) Phytoplankton calcification in suboxia by CO2-enhanced biotic carbon export. Dickson AG (2007) Total alkalinity: The explicit conser- a high CO2 world. (technical comment) Science 322, Global Biogeochem Cycles, 22:GB4008, 10.1029/ vative expression and its application to biogeochemical 10.1126/science.1161096. 2007GB003147. processes. Mar Chem 106:287–300. 47. Riebesell U, Wolf-Gladrow DA, Smetacek V (1993) Car- 74. Engel A, et al. (2004) Transparent exopolymer parti- 19. Orr JC, et al. (2005) Anthropogenic ocean acidification bon dioxide limitation of marine phytoplankton cles and dissolved organic carbon production by over the twenty-first century and its impact on calcify- growth rates. Nature 361:249–251. Emiliania huxleyi exposed to different CO2 concen- ing organisms. Nature 437:681–686. 48. Hein M, Sand-Jensen K (1997) CO2 increases oceanic trations: A mesocosm experiment. Aquatic Microb 20. Steinacher M, Joos F, Fro¨licher TL, Plattner GK, Doney primary production. Nature 388:526–527. Ecol 34:93–104. SC (2009) Imminent ocean acidification in the Arctic 49. Zondervan I, Rost B, Riebesell U (2002) Effect of CO2 75. Engel A, Thoms S, Riebesell U, Rochelle-Newall E, projected with the NCAR global coupled carbon cycle- concentration on the PIC/POC ratio in the coccolith- Zondervan I (2004) Polysaccharide aggregation as a climate model. Biogeosciences 6:515–533. ophore Emiliania huxleyi grown under light-limiting potential sink of marine dissolved organic carbon. Na- 21. Takahashi T, et al. (2002) Global air-sea CO2 flux based conditions and different daylengths. J Exp Mar Biol ture 428:929–932. on climatological surface ocean pCO2, and seasonal Ecol 272:55–70. 76. Schartau M, et al. (2007) Modelling carbon overcon- biological and temperature effects. Deep-Sea Res II 50. Leornardos N, Geider RJ (2005) Elevated atmospheric sumption and the formation of extracellular particu- 49:1601–1622. CO2 increases organic carbon fixation by Emiliania hux- late organic carbon. Biogeosciences 4:433–454. 22. Takahashi T, et al. (2009) Climatological mean and leyi (Haptophyta) under nutrient-limited, high-light 77. Arrigo KR (2007) Marine manipulations. Nature conditions. J Phycol 41:1196–1203. 450:491–492. decadal change in surface ocean pCO2, and net sea-air 51. Engel A (2002) Direct relationship between CO uptake 78. Legendre L, Rivkin RB (2002) Pelagic food webs: Re- CO2 flux over the global oceans. Deep-Sea Res II 2 56:554–577. and transparent exopolymer particles production in nat- sponses to environmental processes and effects on the 23. Ko¨rtzinger A, Send U, Wallace DWR, Karstensen J, De- ural phytoplankton. J Plankton Res 24:49–53. environment. Ecol Res 17:143–149. 52. Burkhardt S, Zondervan I, Riebesell U (1999) Effect of 79. McNeil BI, Matear RJ (2008) Southern Ocean acidifica- Grandpre M(2008). The seasonal cycle of O and pCO 2 2 CO concentration on the C:N:P ratio in marine phyto- tion: A tipping point at 450-ppm atmospheric CO . Proc in the central Labrador Sea: Atmospheric, biological 2 2 plankton: A species comparison. Limnol Oceanogr Natl Acad Sci USA 105:18860–18864. and physical implications. Global Biogeochem Cycles 44:683–690. 80. Langdon C, et al. (2000) Effect of calcium carbonate sat- 22:GB1014, 10.1029/2007GB003029. 53. Riebesell U, et al. (2007) Enhanced biological carbon con- uration state on the calcification rate of an experimental 24. Ko¨rtzinger A, et al. (2008). The seasonal pCO2 cycle at sumption in a high CO2 ocean. Nature 450:545–549. coral reef. Global Biogeochem Cycles 14:639–654. 49° N/16.5° W in the northeast Atlantic Ocean and 54. Hutchins DA, et al. (2007) CO2 control of Trichodes- 81. Reynaud S, et al. (2003) Interacting effects of CO2 par- what it tells us about biological productivity. J Geophys mium N2 fixation, photosynthesis, growth rates, and tial pressure and temperature on photosynthesis and Res 113:C04020, 10.1029/2007JC004347. elemental ratios: Implications for past, present and calcification in a scleractinian coral. Global Change Biol 25. Ito T, Follows M (2005) Preformed phosphate, soft tissue future ocean biogeochemistry. Limnol Oceanogr 9:1660–1668. pump and atmospheric CO2. J Mar Res 63:813–839. 52:1293–1304. 82. Guinotte JM, et al. (2006) Will human-induced changes 26. Brzezinski MA, et al. (2002), A switch from SI(OH)4 to 55. Levitan O, et al. (2007) Elevated CO2 enhances nitrogen in seawater chemistry alter the distribution of deep-sea NO3 depletion in the global Southern Ocean, Geophys fixation and growth in the marine cyanobacterium scleractinian corals? Frontiers Ecol Environm 4:141– Res Lett 29:1264, 10.1029/2001GL014349. Trichodesmium. Global Change Biol 13:531–538. 146.

Riebesell et al. PNAS ͉ December 8, 2009 ͉ vol. 106 ͉ no. 49 ͉ 20609 Downloaded by guest on September 25, 2021