Mg/Ca-ΔCO3porewater2−–temperature calibration for Globobulimina spp.: A sensitive paleothermometer for deep-sea temperature reconstruction

• A sensitive Mg/Ca paleothermometer for deep-sea temperature reconstruction is presented.
• Concentration of ΔCO2−3 in porewater is spatially and temporally variable.
• ΔCO2−3 pore water is highly correlated to the ΔCO2−3 of the overlying bottom water.
• Mg/Ca (Globobulimina) sensitivity to ΔCO2−3 is estimated 0.009 mmol/mol/μmol/kg.

Existing benthic foraminiferal Mg/Ca–temperature calibrations are surrounded by substantial uncertainties mainly due to low temperature sensitivity of Mg/Ca in most benthic foraminifers and the effect of carbonate ion concentration on benthic foraminiferal Mg/Ca. Here we present Mg/Ca analysis of Rose Bengal stained and exceptionally well-preserved tests of the infaunal benthic foraminifer Globobulimina spp. from 39 eastern equatorial Atlantic core top samples. Mg/Ca in Globobulimina   spp. varies between 2.5 mmol/mol and 9.1 mmol/mol corresponding to bottom water temperatures (BWT) between 1.8 °C and 19.1 °C and ΔCO3pore water2− between 33.7±433.7±4 and −34.3±4 μmol/kg−34.3±4 μmol/kg in sediment depths between 1 and 10 cm. Mg/Ca and BWT are linearly correlated with a best fit of Mg/Ca [mmol/mol] = (0.36 ± 0.02) ⁎ BWT [°C] + 2.22 ± 0.19 (r2=0.92r2=0.92, p  -value: 11⁎10−2011⁎10−20, and n=39n=39). Using total alkalinity and pH data of pore water samples from 64 Atlantic multi-corer sites, we obtained ΔCO3pore water2− data from the depth habitat range of Globobulimina   spp. (≥1 cm ≤ 10 cm below sediment surface). We show that ΔCO3pore water2− is significantly lower than and linearly co-varies with the ΔCO2−3 of the overlying bottom water: ΔCO3pore water2−=(0.67±0.05)⁎ΔCO3bottom water2−−(39.84±1.98); r2=0.75r2=0.75, p  -value: 6⁎10−206⁎10−20, n=64n=64. We found a Mg/Ca sensitivity of 0.009±0.0044 mmol/mol0.009±0.0044 mmol/mol per μmol/kg ΔCO3pore water2− and Mg/Ca temperature sensitivity of 0.32±0.06 mmol/mol/°C0.32±0.06 mmol/mol/°C after a correction for the ΔCO3pore water2− effect. This study provides a robust Mg/Ca–temperature calibration, highlights that ΔCO3pore water2− is spatially and most likely temporally variable, and contradicts the notion that infaunal foraminiferal Mg/Ca is relatively immune from ΔCO2−3 changes in the overlying bottom water. Furthermore, comparison of down core Mg/Ca data of Cibicides pachyderma and Globobulimina spp. demonstrates that the high temperature sensitivity of Mg/Ca in Globobulimina spp. presents a more robust paleothermometer to reconstruct past changes in the thermal state of the deep ocean.