A core-top calibration of the benthic foraminiferal B/Ca proxy for deep water carbonate saturation for Nuttallides umbonifera: Facilitating paleoceanographic reconstructions

Reconstructions of past deep-ocean carbonate saturation states (Δ[CO32-]) can offer significant constraints on atmospheric pCO2 levels by elucidating the CaCO3 cycle’s contribution to pCO2 variations. Yu and Elderfield (2007) recently introduced a novel method for the reconstruction of paleo-Δ[CO32-], which utilizes boron/calcium ratios in benthic foraminifera. In that initial work, they observed significant inter-species variation in B/Ca, highlighting the importance for species specific calibrations. We are near the completion of a new modern B/Ca core-top calibration for two benthic species, Nuttallides umbonifera and Oridorsalis umbonatus. These calibrations will aid in extending the utility of the proxy into deeper time. N. umbonifera is the closest extant relative to Nuttallides truempyi, a long-lived and common Paleogene species. O. umbonatus is also widely distributed in the Paleogene and Neogene and is still extant. We measured B/Ca ratios in N. umbonifera and O. umbonatus from 35 Holocene sediment samples from the Atlantic, Pacific and Southern Ocean. B/Ca ratios in O. umbonatus are significantly lower (35 - 60 µmol/mol) than those observed in Cibicidoides wuellerstorfi (~130 - 250 µmol/mol; Yu and Elderfield, 2007). Further, the sensitivity to Δ[CO32-] appears to be reduced, potentially invalidating the use of B/Ca in O. umbonatus for reconstructing past deep water Δ[CO32-]. B/Ca ratios in N. umbonifera are also offset from C. wuellerstorfi toward lower values (~85 - 175 µmol/mol), but by a constant amount, such that the overall sensitivity to Δ[CO32-] is similar to that observed for C. wuellerstorfi. A linear correlation between benthic foraminiferal B/Ca and deep water Δ[CO32-] holds true, however incorporating both the analytical error associated with the measurement of B/Ca as well as the error associated with estimating Δ[CO32-] is critical. Because hydrographic data are rarely available from the same location as the sediment cores being used, core-top calibrations are hampered by the necessity to estimate core site hydrographic parameters from nearby oceanographic sites (e.g. GLODAP, CARINA, etc.). Quantifying the error in these estimates, though challenging, is therefore essential, as consideration of error in both axes can have a significant effect on the slope of the regression line. References: Yu, J., Elderfield, H., 2007. Benthic foraminiferal B/Ca ratios reflect deep water carbonate saturation state. Earth and Planetary Science Letters 258 (1-2), 73-86.