Reconstructing pCO2 in depth: Organic and inorganic proxies for past seawater carbon chemistry

Past changes in marine inorganic carbon chemistry can be used as analogues for the current disturbance of the carbon cycle and hence, predict future impacts of the rising atmospheric CO2 levels. The marine inorganic carbon system consists of six parameters – pCO2, [CO32-], [HCO3-], dissolved inorganic carbon (DIC), pH, and total alkalinity – of which at least two must be known to derive the rest of the system. Such a reconstruction of the complete inorganic carbon system including dissolved [CO2] allow for estimating atmospheric pCO2 as the latter two are in equilibrium. Current approaches often involve combining organic and inorganic proxies, however, isolating the primary control on any of these proxies is often complicated by the natural covariance among the carbon system parameters. Boron isotopic composition (δ11B) in foraminiferal shells and the carbon isotopic composition (δ13C) of alkenones are considered as the most robust proxies to reconstruct pCO2. Still, these proxies rarely agree completely due to their inherent complications. Therefore, resolving discrepancies between independent proxies requires refinement in their applications and a mechanistic understanding of how these proxies function. Field surveys using core-top sediment samples and inorganic calcite growth experiments are used here to determine and decouple the parameters influencing element concentrations in foraminiferal shells, such as those of sulfur and boron. These two approaches allow to calibrate element-to-calcium ratios to parameters of the marine carbon system and show that more than one of them affect the incorporation of sulfate and boron in calcite. While application of these proxies is traditionally based on the assumption that boron and sulfur are incorporated as single anions, results of this thesis suggest that incorporation in the form of ion-pairs may explain some of the discrepancies observed in proxy calibration studies. Therefore, exploring cooperative incorporation of elements into calcite along with alternative incorporation pathways in the form of ion-pairs shows a promising approach to improving these proxy-based pCO2 reconstructions. Upwelling regions are dynamic systems that play an important role in carbon cycling as relatively cold, DIC- and nutrient-rich waters are returned from depth to the sea surface. The behavior of these regions over glacial-interglacial periods and their contribution to atmospheric pCO2 is not fully disclosed, but potentially important and can be evaluated using the aforementioned CO2-proxies. Foraminiferal δ11B and the δ13C of alkenones are applied to investigate the CO2-history of the Benguela Upwelling System and Canary Current Upwelling System during the last glacial cycle. Using multiple species with different living depths as proxy signal carriers, depth-dependent differences in carbon chemistry, and consequently ocean-atmosphere interactions can be reconstructed. While these reconstructions suggest enhanced carbon storage at intermediate depth, the outgassing of CO2 was modulated by changes in upwelling strength and the efficiency of the biological carbon pump to draw down CO2. High primary productivity likely prevented the outgassing of CO2 in the Benguela Upwelling System but was not sufficient to confine the outgassing in the Canary Current Upwelling System during the last glacial maximum.