The European Science Fondation (EuroCLIMATE Programme) sponsored a workshop on Atmospheric CO2, ocean acidification, and ecological changes in planktonic calcifying organisms with 45 international participants both from the EuroCLIMATE programme as well as externally invited experts. The event was co-sponsored by PAGES (Past Global Changes) with participation from SOLAS and IMBER. The workshop was hosted in Barcelona, Spain at the Cosmo Caixa Science Museum and Hotel Eden Roc Conference Center (Sant Feliu de de Guíxols) in Barcelona, 26-28/09/2007. In the afternoon of September 26th a public symposium on ocean acidification (OA) was organized at, and co-sponsored by, the Science Museum of Barcelona. Public lectures were given by four participants of the workshop: Richard Feely (NOAA Seattle) Global Warming and Ocean Acidification: Double Trouble for Marine Ecosystems; Victoria Fabry (California State University San Marcos) A global geochemical experiment with unknown ecological consequences; James Zachos (University of California at Santa Cruz) Long-term Consequences of Ocean Acidification: A Paleoperspective; and Carol Turley (Plymouth Marine Laboratory) Taking the Science of Ocean Acidification to Policy Makers, Stakeholders and Society. These lectures provide an authoritative but accessible overview of the issue. A summary of the lectures will be available as video at the ESF website of the workshop (http://www.esf.org/acidification-workshop.html).

Background

Changes in ocean chemistry due to anthropogenic CO2 emissions affect marine life, nutrient cycles and biocalcification. Increasing atmospheric CO2 levels results in increased seawater CO2 concentrations and consequently alters the seawater carbonate equilibrium, decreasing pH. This effect can effect marine ecosystems and especially marine calcifying organisms. For given CO2 emission scenarios the gross chemistry of this process can be modeled very well while it is more difficult to predict the impact of OA on the biota and the feedback of the biosphere to OA. The main purpose of this workshop was to understand the consequences of future CO2 emissions for the marine environment and to establish research priorities that would most effectively fill in current gaps in our fundamental understanding.

Marine carbonate precipitation today is a biologically controlled process with main contributions by pelagic planktonic organisms (such as coccolithophores, foraminifera, and pteropods). Planktonic calcifiers are particularly significant since they are the main contributors to the export of carbonates from the surface ocean to the sea floor. Their continuous fossil record provides the opportunity to observe past calcification and ecological responses to a range of scenarios including both rapidly rising and persistently high atmospheric CO2 levels.

Calcidiscus leptoporus (coccolithosphores) images from Gerald Langer, illustrating the effects of artificially elevated CO2 on calcification. The speciment is about 18 ìm in diameter.

Workshop structure

The structure of the workshop included 5 core-topic sessions: 1) Biocalcification mechanisms and their vulnerability to OA, 2) Genetics and Physiology - investigating organismal responses to OA, 3) Ecology and Biogeography - predicting the effects of population responses to OA, 4) Lessons from the fossil record - past responses to OA, and 5) Case study: the likely impact of OA in the Mediterranean.

Some of the key questions addressed included: How will rising CO2 levels affect the calcifying taxa - will they simply produce lighter skeletons or will there be significant reductions in their gross productivity leading to selective extinctions? What will be the secondary ecological and biogeochemical consequences of a reduction in biocalcifying plankton? Will a decrease in calcification significantly reduce the export flux of organic carbon by a reduction in their role to ballast organic carbon, and will this significantly reduce the buffering of atmospheric CO2 by a consequently pH lowering? How will changes in the biogeography of key planktonic calcifiers affect overall marine carbonate export production? How will functional gene-diversity respond to changes in ocean chemistry such as OA and nutrient distribution? What is known about the evolution of these genes, and their potential to adapt? What have we learned from past and rapid OA events in relation to the above questions and points? What is known about the impact of anthropogenic CO2 in the Mediterranean on biocalcification and ecology?

Excellent overview talks on each theme, plenty of time for discussion during the specific workshop sessions, and a very wide range of expertise contributed to a powerful format for the open exchange of concepts and knowledge.

Main outcome

Biocalcification mechanisms and their vulnerability to OA

Although we are just at the beginning of understanding how OA affects calcifying organisms new data presented at the workshop improved our state of knowledge. Colin Brownlee (Marine Biological Association, UK) and Jonathan Erez (Hebrew University, Isreael) presented overviews on coccolithophore and foraminifer biocalcification. Foraminiferal calcification occurs only in membrane-bound compartments, while coccolithophore calcification occurs in unique endomembrane compartments. The sensitivity of foraminifera to OA can be readily explained by the fact that they are utilizing seawater as their direct source of Ca2+ and CO32- for calcification. In foraminifera this use is mediated by vacuolization of seawater and direct channeling of the seawater vacuoles to the site of biomineralization. The main modification that the organisms exert on the seawater is elevation of pH by roughly 1 unit relative to ambient levels in order to increase the dissolved inorganic carbon pool (DIC) available for calcification. It can readily be understood that lowering the ambient pH will impede the DIC “concentrating mechamism”. For coccolithophores one of the key questions that was addressed was: If calcification occurs in an intracellular buffered compartment, why should it be affected by increased ocean acidity? A possible mechanism was related to the metabolic effects of calcification on these organisms.

The regulation of intracellular pH during calcification under conditions when photosynthesis may vary rapidly in fluctuating light may include specific pathways that remove protons from the cell. Evidence for such pathways has been gained from electrophysiological studies which are beginning to indicate that the ability of the cell to eliminate excess protons may be compromised at lower external pH.

This session also heard presentations (Silke Thoms and Gerald Langer, AWI Bremerhave, Germany) which showed the importance of the use of isotope fractionation experiments to inform both the mechanisms of calcification and how coccolithophores are able to exert “vital effects” on calcite isotope signatures.

Genetics and Physiology - investigating organismal responses to OA

The session was very much linked to the one on biocalcification and primarily concerned with coccolithophores (Colomban de Vargas and Ian Probert, Roscoff Biological Station, France). Calcification rate and coccolith morphogenesis of these algae are influenced by the carbonate chemistry of seawater, with different effects on different species (Riebesell et al., 2000; Langer et al., 2006). Up to now short term perturbations were used to assess the response of coccolithophores to OA. Preliminary data presented by Marius Müller (Leibniz Institute Kiel, Germany) from a long-term experiment including an acclimatization phase show that the calcification response is much smaller if the organisms has the chance to adapt to the change in chemistry. Strain- and species- specific responses also need to be considered and there is a need to combine long term experiments with analyses of genetic changes. A future combination of long term experiments with genetic work may identify the parts of the genome involved in calcification.

Besides producing calcareous shells, two coccolithophore genera, Emiliania and Gephyrocapsa, also produce alkenones, the unsaturation of which depends on the growth temperature of these species. Fred Prahl (Oregon State University, USA) showed new data on specimens exporting an unsaturation signal to the seafloor, which records water temperature at the depth of production. Production may not necessarily occur in surface waters, but rather at a subsurface depth where nutrient limitations can be overcome. The temperature signal recorded by unsaturation patterns might also be biased by nutrient or light stress. The further assessment of the alkenone unsaturation signal as an SST proxy will require the understanding of alkenone biosynthesis and its physiological control and the environmental control on organic matter export to the seafloor. The following two questions provide interesting ideas to ponder in future research endeavors. What would be the consequence on export and how the sedimentary record is written if a phenomenon such as anthropogenically imposed OA prevented the coccolithophores from calcifying? And, is the efficiency of alkenone export dependent upon a ballasting effect due to biomineralization?

Ecology and Biogeography - predicting the effects of population responses to OA

The ecology of coccolithophores and foraminifera was reviewed by Jeremy Young (Natural History Museum London, UK) and Ralf Schiebel (National Oceanography Centre Southampton, UK) and by Toby Tyrrel (National Oceanography Centre Southampton, UK) from the ecosystem modelling point of view. Tyrrel specified the need for more large scale surveys, both of the open ocean to constrain the large-scale biogeochemical processes and of shelf seas and semi-enclosed seas which may have more unusual carbon chemistries and as such, possibly be more informative. He also presented recently published data on the Baltic Sea, which revealed an unusual undersaturation for calcium carbonate in wintertime surface waters (Tyrrell et al, 2007).

The workshop emphasized the additional knowledge which can be gained studying ecology and biogeography. Laboratory experiments play a valuable role but have well-acknowledged shortcomings, including: (1) the difficulties of running them for long enough to allow the organisms to adapt via evolution to the different chemistry (2) usually the restriction to a monoculture rather than a species assemblage, and hence no possibility for replacement of acid-sensitive by acid-insensitive (or more tolerant) species or strains, and (3) the lack of an ecosystem, and hence inability to examine trophic effects such as variation in shell thickness on grazing susceptibility. It is therefore essential to supplement laboratory experiments with field observations, including studying the chemical conditions of environments in which calcifiers do and do not live, to see which ranges of pH and CaCO3 saturation state (?) they can tolerate in-situ.

Pelagic calcifiers are not normally dominant components of marine ecosystems, but they do play major biogeochemical roles both via export of calcium carbonate to the sediments and perhaps most importantly by ballasting particulate organic carbon (POC). Transport of POC to the deep sea in marine snow and fecal pellets is dependant on ballasting by mineral matter and particularly by the shells of pelagic calcifiers (Armstrong et al. 2002, Klaas & Archer 2002). Hence, a major impact of reduced calcification would be reduced flux of food to the deep benthic community and a reduction in CO2 sequestration into sediments and the deep ocean. A first compilation of coccolith data shows a relationship between coccolith carbonate and POC flux (Ziveri et al., 2007).

The effect of climate change on the ecology and biogeography of foraminifera and coccolithophores are plentiful and interdependent. For instance, additional to an increase in temperature of the ocean, global warming is predicted to reduce the depth of the mixed layer and to increase the stability of surface waters. These hydrographic changes will have consequences for nutrient supply and hence affect the basis of the oceanic food web. Equally, increased uptake of CO2 by the ocean will not only affect its carbonate chemistry but also the bioavailability of micro nutrients such as Fe. However, new data were presented at the workshop showing ecological and predicted biogeographical changes in our present and future high-CO2 oceans. For instance, Patrizia Ziveri (Universitat Autonoma Barcelona, Spain, and Vrije Universiteit Amsterdam, The Netherlands) reported a clear ecological change in coccolithophore distribution in the annual varved sediment of the Santa Barbara Basin since the 1970’s associated with increased stratification and annual temperature.

Lessons from the fossil record - past responses to OA

The overview lectures by Daniela Schmidt (Bristol University, UK) and Heather Stoll (Oviedo University, Spain) reported examples of OA events in the geological past focusing on some key periods such as the PETM (Paleocene-Eocene Thermal Maximum) at around 55 Myr ago. Andy Ridgewell (Bristol University, UK) discussed the issues related to computer simulations to model past changes in OA and any potential biological responses. Schmidt pointed out that changes in species composition and biogeography could be as important, if not more so, than actual extinctions during OA events. Particular focus was given to the K-T (K-Pg) Boundary (approx. 65 Myr ago). A decrease in carbonate accumulation at this time is related to reduction in carbonate production (i.e. calcification) rather than increased dissolution at the sea floor. Foraminiferal tests decrease in average size while the carbonate producing plankton shifted from coccolithophore to foraminifera domination. Recovery of carbonate production took several millions of years. Stoll focused on the PETM showing evidence from a site in the Southern Ocean (a region particularly sensitive to the effects of OA) of distinct shifts in plankton assemblage at this time, with an increase in the presence of species more suited to warmer conditions although these also tend to be those less susceptible to dissolution. Further evidence of assemblage changes at this time comes from other regions.

Coccolith Sr/Ca measurements are used as a paleo proxy for algal carbonate production and provide some evidence that coccolithophore production may have peaked during the PETM in the Southern Ocean region. Ridgwell discussed some key areas for attention when using computer simulations to model past changes in OA and any potential biological responses. Although it is relatively straightforward to predict the inorganic response of carbonate sediments to changes in ocean acidity, it is extremely important to constrain the associated changes in atmospheric CO2 and possible biological calcification responses before we will be able to reduce the uncertainties in our models. Richard Zeebe (University of Hawaii, USA) stressed the importance of major events in the past as the nearest candidate for a direct analogue to current OA. Particular emphasis was given to considering the geographical distribution of any event in terms of response in the system. For example the extent of carbonate dissolution during the PETM was not globally equal. None of the past OA analogues will be able to depict the true extend of future acidification since the injection of carbon into the ocean / atmosphere system during the PETM was most likely not as rapid and intense as the modern situation. A more gradual release of carbon would have been more efficiently buffered by deep sea carbonate dissolution which would have meant a significantly reduced effect on surface ocean pH. An important point made by Carol Turley (Plymouth Marine Laboratory, UK) was that any effect on marine biocalcification during the PETM would represent the very minimum response that might be expected in the future. A problem with using the PETM and the KPg to understand effects and recovery time is the very different ecosystem structure so deep in the geological record. The youngest geological example of deep sea acidification (causing dissolution) is the Mid-Brunhes interval (around 0.5 Myr ago). Stephen Barker (Cardiff University) used this interval to highlight a key area of uncertainty for predicting future consequences of OA: the cycling of organic and inorganic carbon. Dissolution during the Mid-Brunhes interval is possibly driven by changes in the calcifying planktonic ecosystem that can influence the balance between the organic and inorganic carbon pumps (and hence atmospheric CO2). The recent nature of these changes allows a direct comparison to be made with modern ecosystems and can bridge the gap between geology and modern biogeography.

What is known about the impact of anthropogenic CO2 and OA in the Mediterranean?

Most climate models are today restricted to the major oceans, as the oceanographic processes in semi-enclosed or relatively shallow basins, such as the Mediterranean Sea, the Caribbean, the North Sea and the continental shelf of Indonesia are too complex and small-scale to be resolved in global climate models. From a European perspective, this excludes one of the major economically important regions from future predictions.

Sebastian Meier (CEREGE, France) gave an introduction to the oceanography of the Mediterranean, providing implications about past and future changes. The Mediterranean is an area of high anthropogenic carbon input and extreme carbonate supersaturation, so no signs of dissolution are expected. However, the workshop offered the unique opportunity to share unpublished results that showed something different.

Catherine Goyet (University of Perpignan, France) presented the calculation of anthropogenic carbon from the long time series (1994-2006) DYFAMED site and the extrapolation of the results to the whole Mediterranean. The analysis reveals a big difference between the eastern and western basin. A change is noted between the years 2001 and 2006, related to water circulation.

Despite the fact that the Mediterranean region has been identified as a “hot spot” for future climate change with predicted temperature increase of up to 6°C on land within the next 100 years, only very limited and patchy information is available so far for the Mediterranean Sea so far. Taking into account that the Mediterranean Sea is known to sink for anthropogenic carbon, the combined effects of OA and temperature increase on climate and ecosystem may be larger than in any other European region. Due to the short residence time in the Mediterranean Sea future changes will take place relatively fast and should therefore be a primary target for future research.

In particular, the carbonate system of the Mediterranean Sea and the response of the highly adapted organisms is poorly understood. Generally, supersaturation with respect to calcite and aragonite is observed throughout the entire basin, which may be the reason for overcalcification observed in coccolithophores in the eastern Mediterranean (Maria Triantaphyllou, University of Athens, Greece). At the same time, however, effects that are ascribed to OA can already be observed in other parts of the Mediterranean, like e.g. thinning and malformation of coccolithophores (S. Meier) and massive blooms of jellyfish (Dror Angel, Haifa University, Israel). From the little information available, it seems that recent oceanographic changes like the Eastern Mediterranean Transient, i.e. a shift in the site of deep water formation, coincide with changes in the uptake and sequestration of anthropogenic carbon. Due to this high environmental variability within a relatively restricted space and the resulting steep physicochemical gradients, the Mediterranean Sea could serve as a small-scale model for studying the impact of future OA and warming.

Final considerations

The stimulating cross-disciplinary thinking of this workshop provided a unique opportunity to trace the state of knowledge and prioritize research agendas to understand future impacts of fossil-fuel CO2 on planktonic calcifying organisms and ecosystems.

We are now beginning to understand mechanistically why the physiology and the biocalcification of these organisms are affected by high CO2. This understanding makes the issue of OA and the rate and magnitude projected for the coming decades an even larger risk. There are already notable field observations on ecological changes and effects on calcification.

The workshop was entitled “Atmospheric CO2, Ocean Acidification, and Ecological Changes in Planktonic Calcifying Organisms”, but it is noteworthy that most marine plankton are not calcifiers and several of the participants pointed out that we should place greater emphasis on the effects of elevated oceanic CO2 and acidification on planktonic communities as a whole since we may assume that these changes will affect many other cellular processes and organismal interactions, in addition to calcification.

There is the need of an international network of observations and process studies investigating potential OA impacts. Considering the potential socio-economic impacts of OA, it seems timely to join forces and share resources in an international effort guided by an international scientific steering group similar to that carried out in the 1980-90’s by the international marine programs JGOFS and WOCE. It is fundamental to agree on standardized protocols (e.g. calcification rate measurements, CO2 system parameters, manipulation of seawater CO2 chemistry, determination of abundances and distributions of planktonic calcifiers, respiration measurements). The organizers of the Scoping Workshop on Ocean Acidification Research (Scripps, San Diego, October 9-11, 2007) Victoria Fabry and Richard Feely who were actively involved in this ESF workshop felt very strongly about this.

Along with climate change, ocean acidification shows the need for urgent and substantial reduction of CO2 emissions and might provide tipping points for mitigation or adaptation strategies. Therefore it is important that both the climate change and energy policy communities are aware of the problem and that scientists with knowledge of ocean acidification engage with these communities. Carol Turley noted that if the marine science community wants to influence future CO2 emission strategies we have to catch–up with the climate change policy community and need to address questions such as “what constitute a dangerous pH change?” and “where are the tipping points?” Most importantly we will assess within our own community the level of certainty of our understanding, with regard to ocean acidification, of the distant past and our predictions of the future, their impacts and feedbacks and find effective way to voice this information clearly and without ambiguity to those that will be making future policy decisions. As a follow up, several of the participants will meet in January 2008 to discuss socio-economic impacts of ocean acidification and produce a policy briefing.

Authors:

• Patrizia Ziveri, Universitat Autonoma de Barcelona, Institute of Environmental Science and Technology (ICTA) Bellaterra, Spain, and Department of Paleoclimatology and Geomorphology, Vrije Universiteit, Amsterdam, Netherlands
• Victoria Fabry, California State University San Marcos, Department of Biological Sciences, San Marcos, United States
• Jelle Bijma, Alfred-Wegener Institut für Polar- und Meeresforschung, Carbon Group. Bremerhaven, Germany
• Daniela Turk, European Science Foundation, Life, Earth and Environmental Sciences, Strasbourg, France
• Jeremy Young, The Natural History Museum, Department of Palaeontology, London, United Kingdom
• Daniela Schmidt, University of Bristol, Bristol, United Kingdom
• Colin Brownlee, Marine Biological Association, Plymouth, United Kingdom.
• Jonathan Erez, The Hebrew University of Jerusalem, Jerusalem, Israel
• Thorsten Kiefer, PAGES International Project Office, Bern, Switzerland
• Carol Turley, Plymouth Marine Laboratory, Plymouth, United Kingdom
• Anton Eisenhauer, Kiel University, Leibniz Institute of Marine Sciences (IFM-GEOMAR), GEOMAR Research Center for Marine Sciences, Kiel, Germany

Corresponding author: Patrizia Ziveri, Universitat Autonoma de Barcelona, Institute of Environmental Science and Technology (ICTA) Bellaterra, Spain, and Department of Paleoclimatology and Geomorphology, Vrije Universiteit, Amsterdam, Netherlands