Application of benthic foraminiferal Mg/Ca ratios to questions of Cenozoic climate change

We investigate the evolution of Cenozoic climate and ice volume as evidenced by the oxygen isotopic composition of seawater (δ ¹⁸O _sw) derived from benthic foraminiferal Mg/Ca ratios to constrain the temperature effect contained in foraminiferal δ ¹⁸O values. We have constructed two benthic foraminiferal Mg/Ca records from intermediate water depth sites (Ocean Drilling Program sites 757 and 689 from the subtropical Indian Ocean and the Weddell Sea, respectively). Together with the previously published composite record of Lear et al. [Science 287 (2002) 269-272] and the Neogene record from the Southern Ocean of Billups and Schrag [Paleoceanography 17 (2002) 10.1029/2000PA000567], we obtain three, almost complete representations of the δ ¹⁸O _sw for the past 52 Myr. We discuss the sensitivity of early Cenozoic Mg/Ca-derived paleotemperatures (and hence the δ ¹⁸O _sw) to assumptions about seawater Mg/Ca ratios. We find that during the middle Eocene (∼49-40 Ma), modern seawater ratios yield Mg/Ca-derived temperatures that are in good agreement with the oxygen isotope paleothermometer assuming ice-free conditions. Intermediate waters cooled during the middle Eocene reaching minimum temperatures by 40 Ma. The corresponding δ ¹⁸O _sw reconstructions support ice growth on Antarctica beginning by at least 40 Ma. At the Eocene/Oligocene boundary, Mg/Ca ratios (and hence temperatures) from Weddell Sea site 689 display a well-defined maximum. We caution against a paleoclimatic significance of this result and put forth that the partitioning coefficient of Mg in benthic foraminifera may be sensitive to factors other than temperature. Throughout the remainder of the Cenozoic, the temporal variability among δ ¹⁸O _sw records is similar and similar to longer-term trends in the benthic foraminiferal δ ¹⁸O record. An exception occurs during the Pliocene when δ ¹⁸O _sw minima in two of the three records suggest reductions in global ice volume that are not apparent in foraminiferal δ ¹⁸O records, which provides a new perspective to the ongoing debate about the stability of the Antarctic ice sheet. Maximum δ ¹⁸O _sw values recorded during the Pleistocene at Southern Ocean site 747 agree well with values derived from the geochemistry of pore waters [Schrag et al., Science 272 (1996) 1930-1932] further highlighting the value of the new Mg/Ca calibrations of Martin et al. [Earth Planet. Sci. Lett. 198 (2002) 193-209] and Lear et al. [Geochim. Cosmochim. Acta 66 (2002) 3375-3387] applied in this study. We conclude that the application of foraminiferal Mg/Ca ratios allows a refined view of Cenozoic ice volume history despite uncertainties related to the geochemical cycling of Mg and Ca on long time scales.