The dissolution effect on the B/Ca ratio of planktic foraminifers: a potential bias for paleo-pH reconstructions

The B/Ca ratio of planktonic foraminifers is a new paleoceanographic tool to reconstruct past sea-surface pH. However, we show that the dissolution of foraminifer shells may result in the preferential removal of B and represents, therefore, a potential bias for pH reconstructions, similar to that observed for Mg/Ca and paleo-temperature estimates. We determined B/Ca and Mg/Ca of the planktonic foraminifer species Globigerinoides sacculifer in seven core-tops from a depth transect along the Sierra Leone Rise, in the Eastern Equatorial Atlantic. Samples were analyzed independently at two laboratories, the LSCE at Gif-sur-Yvette and the Godwin Laboratory at Cambridge. Both sets of results show a systematic decrease of B/Ca and Mg/Ca along the depth transect, with an overall loss of ∼14 μmol/mol (∼15%) for B/Ca and of ∼0.7 mmol/mol (∼21%) for Mg/Ca between the shallowest (2640 m) and the deepest (4950 m) sites. Because of this dissolution bias, surface water pH reconstructed from B/Ca of G. sacculifer is ∼0.11 unit lower at the shallow water site than at the deep site. This dissolution-driven difference in pH has a similar magnitude than the expected glacial/interglacial surface water pH change. Because planktonic foraminifer dissolution is related to the saturation state of bottom-water with respect to calcite, we propose a correction approach based on benthic foraminifer B/Ca ratios, a proxy of carbonate ion saturation (ΔCO₃^2-). Along our Sierra Leone Rise transect, we found a positive relationship between the B/Ca of the benthic species Cibicides wuellerstorfi (and hence ΔCO₃^2-) and the relative loss of B/Ca and Mg/Ca. This empirical relationship makes it possible to correct estimates of temperature (Mg/Ca) and pH (B/Ca) for dissolution effects, and may be applied to paleoceanographic reconstructions to correct for the effect of past changes in bottom water carbonate ion saturation on these paleoceanographic proxies.