δ ¹⁸O_sw estimate for Globigerinoides ruber from core-top sediments in the East China Sea

The paired analyses of the Mg/Ca ratio and oxygen isotopic composition (δ¹⁸O_c) of surface-dwelling planktonic foraminifera have become a widely used method for reconstructing the oxygen isotopic composition of ambient seawater (δ¹⁸O_sw) as a robust proxy for surface salinity. Globigerinoides ruber (G. ruber) is a mixed-layer dweller, and its fossil shell is an ideal archive for recording past sea surface water conditions, such as those caused by variability in the East Asian summer monsoon (EASM). Here, we investigate the validity of shell-derived δ¹⁸O_sw estimates for G. ruber using core-top sediments from the East China Sea (ECS). First, we determined a local δ¹⁸O_sw–salinity equation for the eastern part of the ECS in July [δ¹⁸O_sw = −7.74 + 0.23 × salinity]. Then, we calculated δ¹⁸O_sw from core-top δ¹⁸O_c and Mg/Ca values in G. ruber using the δ¹⁸O_c–temperature equation of Bemis et al. (Paleoceanography 13(2):150–160, 1998) and the Mg/Ca–temperature equation of Hastings et al. (EOS 82:PP12B-10, 2001). The core-top δ¹⁸O_sw and salinity were estimated to be in the ranges of −0.2 to +0.39 ‰ and 33.7 to 34.5, respectively, which fall close to the local δ¹⁸O_sw–salinity regression line. The core-top data showed that the Mg/Ca–temperature calibration by Hastings et al. (EOS 82:PP12B-10, 2001) and the δ¹⁸O_c–temperature equation by Bemis et al. (Paleoceanography 13(2):150–160, 1998) are appropriate for calculating δ¹⁸O_sw in the ECS. Furthermore, we measured core-top Ba/Ca ratios of G. ruber (Ba/Ca_{G. ruber}), which ranged from 0.66 to 2.82 μmol mol⁻¹. There was not a significant relationship between the salinity and Ba/Ca_{G. ruber} ratios due to the highly variable Ba/Ca_{G. ruber} data. Given the seawater Ba/Ca data and the published partition coefficient for Ba (D_Ba = 0.15–0.22), pristine Ba/Ca_{G. ruber} ratios at northern Okinawa Trough sites should be less than 0.84 μmol mol⁻¹. Anomalously high Ba/Ca_{G. ruber} ratios (>0.84 μmol mol⁻¹) might be attributable to contamination by sedimentary barite adherent on fossil shells. Therefore, further evaluation of the Ba/Ca_{G. ruber} ratio as a paleo-salinity proxy requires diethylene triamine pentaacetic acid (DTPA)-cleaned Ba/Ca data that can minimize the influence of barite contamination.