Delphine Dissard

• The trace element composition of calcitic foraminiferal tests has become an important tool by which paleoceanographers reconstruct past oceanic conditions. On timescales much shorter than their residence time, the elements strontium and magnesium occur in seawater with nearly constant ratios to calcium. Variation in Sr/Ca and Mg/Ca in benthic foraminiferal tests can then be explained as a function of environmental parameters that control their incorporation into the tests. For benthic foraminifers, temperature appears to be the dominant parameter controlling the incorporation of Mg, however, a better understanding of the possible impact of other parameters such as pH or [CO32-], and salinity is needed to increase the accuracy of element ratio proxies. My thesis aimed to optimise these reconstructions by developing a mechanistic understanding of the biotic processes that control divalent cation concentration in biogenic carbonate. Towards this goal the benthic foraminifer Ammonia tepida was cultured under three different salinities (20, 33 and 40) and two different pCO2 conditions (230 and 1900ppm), both sets of experiment were run at two different temperatures (10 and 15°C). Weights and elemental composition were determined. Results indicate that both Mg and Sr incorporation are enhanced with increasing temperatures and increasing salinity. Changes in [CO32-] or total dissolved inorganic carbon do not affect the Mg partitioning coefficient. On the contrary, Sr incorporation is enhanced under increasing [CO32-]. Shell weights decrease with decreasing [CO32-], and increase with decreasing temperature. To extend this mechanistic understanding to planktonic foraminifers, a calibration based on field specimen of Globigerinoides sacculifer was developed. This allowed us to evaluate agreement between observed and predicted δ18Ow as a function of Mg/Ca temperature and δ18Oc, and to test the robustness of SSS reconstructions for the Atlantic Ocean.