www.-.org

Laurent Bopp*, The Ocean: (LSCE, Gif-sur-Yvette) Chris Bowler*, (ENS, Paris) a Lionel Guidi, (CNRS UPMC, Villefranche-sur-Mer) Pump Éric Karsenti, (EMBL) Colomban de Vargas (CNRS UPMC, Roscoff)

*lead authors

The ocean contains 50 times more carbon than the atmosphere and is exchanging large amounts of CO2 with the atmosphere every year. In the past decades, the ocean has slowed down the rate of by absorbing about 30% of human emissions. Whereas this absorption of anthropogenic CO2 is today the result of physical-chemical processes, ma- rine biology is playing an important role in the ocean by sequestering carbon in the deep ocean. Changes in any of these physical, chemical and biological processes may result in climate feedbacks that either increase or decrease the rate of climate change, although knowledge of such interconnections is today still limited. The feedbacks between climate, the ocean, and its need to be better understood in order to predict the co-evolution of atmospheric CO2 and climate change more reliably and also to understand the characteristics of a future ocean.

A MAJOR ROLE FOR THE OCEAN IN than any other that has happened during the THE EVOLUTION OF ATMOSPHERIC past 65 million years (Portner et al., 2014; Rhein et al., 2014).

CO2 Since the beginning of the industrial era, the ocean The cycling of carbon involves a wide range of has played a key role in the evolution of atmos- physico-chemical and biological processes contri- pheric CO2 by absorbing a significant fraction buting to a series of interconnected carbon re- of CO2 emitted into the atmosphere by human servoirs in the Earth System. A schematic diagram activities, deforestation and burning of fuels. of the global carbon cycle showing the relative During the past decade (2004-2013), the global importance of each of these processes is shown in ocean has absorbed 2.6 billion tonnes of carbon Figure 1. The global cycle was roughly balanced per year, representing nearly 30% of anthropogenic before the industrial era. During the past 200 years, emissions over this period. Since 1870, the amount atmospheric CO2 has increased from less than of carbon absorbed by the ocean has reached 0.03% to more than 0.04%, as a result of 150 billion tonnes – also representing 30% of anthro- burning, production, deforestation and pogenic emissions over this period. By absorbing other changes in land use. It is considered that this , the ocean thus contributes such a rapid change is at least ten times faster to slow down human-induced climate change.

12 www.ocean-climate.org

A NATURAL OCEAN CARBON organic carbon pool (~3PgC), but they are ca- CYCLE INVOLVING PHYSICO- pable of generating large amounts of (DOC: ~700PgC) to sustain the CHEMICAL AND BIOLOGICAL food chains because their turnover is very rapid, PROCESSES from a few days to several weeks. A fraction of produced organic material exits the surface layer as sinking particles, thus trans- Anthropogenic carbon absorbed by the ocean ferring the surface carbon towards the deep feeds a considerable natural carbon reservoir. layers of the ocean (Figure). Before being se- The ocean contains about 40,000 billion tonnes questered to the deep the atmospheric carbon of carbon (40,000PgC), mainly in the form of fixed by photosynthetic undergoes a inorganic carbon dissolved in . This series of transformations: can be amount represents 50 times the size of the at- directly consumed by , or indirectly mospheric reservoir. Each year, the ocean natu- by heterotrophic , which will in turn be rally exchanges with the atmosphere almost a eaten by larger organisms. During this process, hundred billion tonnes of carbon as CO2. a fraction of the biomass (average value of 10%) ends up as detrital matter, fecal In the ocean, this carbon, which prevails essen- pellets or dead cells which compose the stock of - tially in the form of ions (HCO3 ), marine particles. In turn, a fraction of these par- is not evenly distributed, as dissolved carbon ticles (in suspension or sinking) also undergoes concentrations are higher at depth than at the a series of transformations before reaching the surface. The spatial distribution of carbon with base of the mesopelagic layer (typically 1000m depth controls atmospheric CO2 levels, as only depth), thus sequestering atmospheric CO2 for the inorganic carbon from the surface is in thousands of years.2 It is generally believed that contact with the atmosphere and contributes 0.1 to 1% of the carbon-containing material at to the exchange of CO2 between the atmos- the surface finally reaches the base of the me- phere and the ocean. This vertical gradient of sopelagic zone, then the where it can carbon can be explained by both physico-che- turn into fossil fuel deposits. The remaining orga- mical and biological processes. nic matter is remineralized through respiration,

and CO2 returns to the atmosphere. Each year, • Biological Processes nearly 10 billion tonnes of carbon are exported Phytoplankton living in the sunlit layer of the from the surface layer and are responsible for ocean use light to perform photosyn- most of the carbon vertical gradient. All of these thesis. They take up as well as dissolved inorganic car- atmospheric CO bon to produce . atmospheric 2 CO CO The production of these car- 2 2 carbon dissolved bon-based materials supported at the surface by solar energy is called primary deep production. It represents the base euphotic zone phytoplankton carbon of the trophic chains from which zooplankton upwelling 100m dissolved in deep waters other non- photosynthetic or- sinking organic ganisms can feed on. Photosyn- thetic activity is therefore an ef- ficient mechanism for extracting 400m dissolved carbon of organic detritus CO2 from the atmosphere and transferring the carbon into living organisms. Surprisingly, the orga- nisms that contribute to primary Natural carbon cycle and representation of biological and physical production represent only a small pumps (Bopp et al. 2002).

13 www.ocean-climate.org

processes that contribute to the governing role from further contact with the atmosphere. This of on the carbon cycle in the process that contributes to the vertical gradient ocean are part of the so called biological car- of ocean carbon is known as the physical pump bon pump (Figure). or pump (Figure).

Although only a small fraction (~ 0.2PgCyr-1) of Despite the fact that biological processes are the carbon exported by biological processes responsible for the majority of the vertical gra- from the surface reaches the sea floor, the fact dient of natural carbon in the ocean, the phy- that it can be stored in for millennia sico-chemical processes can nevertheless ex- and longer (Denman et al., 2007; Ciais and al., plain the anthropogenic observed

2014) means that this is the today. Indeed, excess CO2 in the atmosphere most important biological mechanism in the will lead to a net carbon flux to the ocean due

Earth System allowing CO2 to be removed from to the disproportion induced between atmos- the carbon cycle for very long periods of time. pheric and oceanic CO2 concentrations. Sub-

sequently, once the anthropogenic CO2 enters Over geological time-scales, the biological car- surface waters, it is transported by ocean cur- bon pump has formed oil deposits that today rents and progressively mixed with the sub-sur- fuel our economy. In addition, biochemical se- face waters. dimentary rocks such as limestone are derived principally from calcifying corals, molluscs, and IS THE OCEANIC CARBON SINK , while the considerable reserves of hydrates (or clathrates) are GOING TO SATURATE? similarly the result of hundreds of millions of years of activity of methanogenic microbial consortia. To date, and since the beginning of the indus- Considering that, each day, large amounts of trial era, the ocean has continuously absorbed

CO2 that have been trapped for millions of years a relatively constant part of the amount of CO2 are discharged into the atmosphere (the order emitted by human activities. However, many of magnitude is now probably about a million studies based on theoretical considerations, in years of trapped carbon burned by humankind situ observations, controlled laboratory expe- each year), it is easier to understand the rapi- riments, or supported by models, suggest that dity at which present climate change is taking several processes may lessen or slow-down this place. Consequently, there is a dramatic diffe- natural carbon sink. rence between the rate of CO2 sequestration by and rate of CO2 discharge into The first series of processes is related to the che- the atmosphere. The anthropogenic emissions mistry of (exchanges between CO2, 3- 2- will therefore need to be redistributed by the HCO and CO3 ) and can eventually lead to a global carbon cycle until a new steady state is saturation of the oceanic carbon sink. Indeed, reached. the dissolution of anthropogenic decreases the ocean ion content and therefore the buffer effect of the ocean,

• Physico-Chemical Processes which in turn increases the proportion of CO2 in A second series of processes, comprising phy- comparison to the other forms of dissolved inor- sico-chemical activities, also contributes to the ganic carbon species and thus may reduce the increasing carbon distribution with depth. The efficiency of the natural carbon sink. This- phe cooling of surface waters at high latitudes fa- nomenon occurs in parallel with the process of vours their ability to dissolve atmospheric CO2 , and could potentially have (mainly by increasing the solubility of the gas) serious impacts on life in the ocean. as well as increasing their density. These heavy surface waters plunge down to great depths, The second series of processes is related to the in this way exporting the CO2 and preventing it feedback between climate and the carbon

14 www.ocean-climate.org

cycle. This concerns the feedback between sity-dependent stability of the surface mixed anthropogenic climate change and different layer (González-Taboada and Anadón, 2012; carbon absorption phenomena. As mentioned Portner et al., 2014) are all likely to affect photo- earlier, climate change leads to modifications synthetic activity, as are , pH, and salinity. in water temperature, ocean currents, and pro- Environmental variability and the displacement duction of organic matter in the ocean. If these of organisms by ocean currents cause variabi- changes should boost the carbon sink, they lity in phytoplankton productivity, competitive- would curb climate change and induce a ne- ness, and natural selection, which are also likely gative feedback. On the contrary, in the event to result in changes in . of a weakening of the carbon sink, the changes It is therefore crucial to estimate how the pro- would lead to a that would in duction of organic material by phytoplankton turn accelerate the phenomenon. is going to be affected by changes in environ- mental conditions of surface water: for example Once more, different processes are involved. rising water temperature, melting of sea ice and For example, the increase in the temperature changes in dissolved availability (ni- of the ocean weakens the ocean carbon sink. trates, ). An increase by 2 or 3°C in sea surface tempe- rature decreases the solubility of CO2 by a few Modelling approaches predict an overall reduc- percent, and thus the capacity of the ocean to tion in global mean NPP as a result of climate absorb carbon dioxide. Another effect could change, albeit with significant latitudinal varia- accentuate this saturation of the carbon sink: tions. One of the factors leading to this reduction in response to rising temperatures, climate mo- is the predicted expansion of oligotrophic gyres dels predict an increase in vertical as nutrient availability decreases with the inten- of the ocean. In other words, vertical mixing, sification of stratification.2 Predictions indicate which tends to homogenize the surface waters increasing NPP at high latitudes (because the with the deep, would diminish and the resulting amount of available should increase as stratification would reduce the present penetra- the amount of water covered by ice decreases). tion of anthropogenic CO2 towards the ocean However this would be counterbalanced by a depths. decrease of NPP in temperate and tropical lati- tudes (because of reduced nutrient supply). The The future of the biological pump is difficult to types of species that would dominate predict. Even a qualitative estimate of the ef- the in altered conditions should also fect of changes in marine ecosystems on the be estimated, as the composition of plankton ocean carbon sink remains highly speculative. can significantly affect the intensity of CO2 ab- More specifically, because the activity of the sorption. The role of certain phytoplankton po- biological pump is likely to be strongly regulated pulations, such as , can be particularly by net (NPP), it is important significant. They are characterised by relatively to consider the effects of climate change on large cell sizes (tens to hundreds of microme- photosynthetic activity. On land, as the CO2 ters), which allows them to sink rapidly. They supply is generally limiting for photosynthesis, the are therefore responsible for the export of a increase in anthropogenic CO2 tends to stimu- large fraction of carbon to the deep ocean. late plant growth (known as the carbon dioxide Nonetheless, diatoms cannot thrive in nutrient fertilization effect). This does not appear to be depleted conditions. In this case they could be the case in marine systems because Dissolved replaced by other types of smaller (<10 microns) Inorganic Carbon (DIC) is not limiting for carbon phytoplankton cells that are better adapted to fixation by photosynthesis. However, photosyn- poor nutrient conditions. Although such cells are thesis is also strongly affected by temperature, abundant in the ocean, due to their small size and the upper ocean has significantly warmed they are principally recycled within the surface during the last 150 years. In addition to tempe- layer, and thus have a very minor role in carbon rature, light, inorganic nutrients, and the den- export to the deep. A decrease in the /

15 www.ocean-climate.org

small cell community ratio could thus greatly dis- pecially phytoplankton, are able to regulate the rupt the intensity of the biological pump, espe- world climate by releasing the sulphurous gas di- cially in the polar regions. methyl sulphide or DMS. Once in the atmosphere, this gas favours the formation of tiny sulphate par- Despite these multiple levels of uncertainty - the ticles which play a role as condensation nuclei most important being the biological response for clouds, thus contributing to an increase in to climate change - the different predictions cloud cover. This hypothesis, which is still called produced by numerical models that couple the CLAW hypothesis (based on the first letter of the and the carbon cycle all the surname of each of the authors; Charlson point to a declining ocean carbon sink due to et al., 1987), states that the ocean ecosystem global warming. Even though this ocean sink is reacts to an increase in temperature by increa- unlikely to become a source there is no doubt sing productivity. This in turn leads to increased that a decrease will affect the evolution of the emissions of DMS, resulting in a temperature drop

CO2 in the atmosphere and, ultimately, climate due to the enhanced cloud cover. This would change itself. By 2100, the feedback between be a self-regulating loop. It is the climate and the carbon cycle (including the an example of regulation that allowed Lovelock response of the terrestrial biosphere to climate to build the Gaia theory, stipulating that several change) could even be responsible for an ad- self-regulatory processes, including the sulphur ditional increase in atmospheric CO2 of several cycle, allow the planet Earth to be considered tens of ppm! as a living .

The future evolution of the oceanic carbon sink, More than 20 years later, research projects have as predicted by models coupling the climate revealed the complexity of the sulphur cycle in and carbon cycle at a global scale, still remains the ocean, but have neither confirmed nor re- very uncertain. The last IPCC report points to a futed this hypothesis. It is not yet known how, why number of poorly constrained processes that and what species of phytoplankton can release explain the wide range of uncertainties asso- the precursory sulphur compounds for the forma- ciated with the predictions: these primarily in- tion of DMS. Knowledge is therefore still lacking clude biotic responses to climate change and to determine whether anthropogenic climate the changes in the biological pump (the com- change will result in a decrease or an increase plexity of biological processes being extremely in DMS emissions from the ocean. difficult to include in climate models). Other processes related to the representation of small- MANIPULATION OF THE CARBON scale features (eddies) and to the consideration of particularly complex coastal areas are also PUMP TO OFFSET CO2-INDUCED mentioned in this report. CLIMATE CHANGE

A ROLE IN OTHER Humankind has disrupted the steady state ba- BIOGEOCHEMICAL CYCLES lance of the global carbon cycle and has brutally contributed to the modification of the composi- tion of Earth's atmosphere, just as bacteria, protists Besides its role in both the carbon cycle and the and the biosphere in general have played a role evolution of atmospheric CO , it must be empha- 2 in the shaping of the Earth’s atmosphere in the sized that the ocean also plays a key role in other past. As other events have marked the history of major biogeochemical cycles, including , our planet in the past, these present changes pro- and sulphur that are liable to affect voked by human activities will significantly affect the biogeochemical balance of our planet. the Earth System. Our duty as inhabitants of the planet Earth is now to formulate predictions and In the mid-1980s, several scientists including James to react in the best possible way to avoid disaster. Lovelock suggested that ocean ecosystems, es-

16 www.ocean-climate.org

Studies have suggested that an artificial -en To conclude, it remains essential to protect the hancement of the ocean carbon pump might ocean carbon pump that contributes to more improve carbon sequestration in the ocean, than half of the CO2 sequestered each day. This thus counterbalancing CO2-induced climate can only be done by preserving the , change. For example, primary productivity of their and their planktonic ecosys- phytoplankton could be stimulated by adding tems. The carbon balance of the different parts nutrients such as to surface waters where of the carbon cycle also needs to be better they are limiting. There is currently no consensus characterised by carrying out further funda- on the efficiency of such methods, which are mental research in this field. limited to a few field experiments. Moreover, alternative geoengineering approaches focu- sing on solar radiation management are not capable of resolving the issue of ocean acidi- fication.

REFERENCES

• BOPP L., LEGENDRE L. et MONFRAY P., 2002 – La pompe à carbone va-t-elle se gripper. La Recherche, 355, 48-50. • CHARLSON R. J., LOVELOCK J. E., ANDREAE M. O. and WARREN S. G., 1987 – Oceanic Phytoplankton, Atmospheric Sulphur, Cloud and Climate. , 326, 655-661. • CIAIS P., SABINE C., BALA G., BOPP L., BROVKIN V., CANADELL J., CHHABRA2 A., DEFRIES R., GALLOWAY J., HEIMANN M., JONES C., LE QUÉRÉ C., MYNENI R. B., PIAO S. and THORNTON P., 2013 – Carbon and Other Biogeochemical Cycles. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press. • DENMAN K. L., BRASSEUR G., CHIDTHAISONG A., CIAIS P., COX P. M., DICKINSON R. E., HAUGLUSTAINE D., HEINZE C., HOLLAND E., JACOB D., LOHMANN U., RAMACHANDRAN S., DA SILVA DIAS P. L., WOFSY S. C. and ZHANG X., 2007 – Couplings Between Changes in the Climate System and . In Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press. • GONZÁLEZ-TABOADA F. and ANADÓN R., 2012 – Patterns of Change in in the North Atlantic During the Last Three Decades : Beyond Mean Trends. Climatic Change, 115, 419-431. • LE QUÉRÉ C. et al., 2014 – Global Carbon Budget. Earth Syst. Sci. Data Discuss., 7, 521-610. • PÖRTNER H.-O., D. KARL M., BOYD P. W., CHEUNG W. W. L., LLUCH-COTA S. E., NOJIRI Y., SCHMIDT D. N. and ZAVIALOV P.O., 2014 – Ocean Systems. In Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A : Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press. • RHEIN M., RINTOUL S. R., AOKI S., CAMPOS E., CHAMBERS D., FEELY R. A., GULEV S., JOHNSON G. C., JOSEY S. A., KOSTIANOY A., MAURITZEN C., ROEMMICH D., TALLEY L. D. and WANG F., 2013 – Observations : Ocean. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press.

17