Do proxies agree? δ18O, δ13C and Mg/Ca from tests of Neogloboquadrina dutertrei in the eastern equatorial Pacific

Abstract

We present new δ18O and δ13C data from Neogloboquadrina dutertrei tests picked from 30 core tops in the eastern equatorial Pacific (EEP) combined with a compilation of published stable isotope and Mg/Ca data. We studied the effects of foraminifer test size on stable isotope data using four size fractions of N. dutertrei: <250 µm, 250–350 µm, 400–500 µm, and 500–600 µm. We find that δ18O decreases and δ13C increases with increasing test size in the smaller size fractions. For tests > 500 µm these trends plateau for δ13C and are reversed for δ18O. Statistically, the most robust relationships between δ18O and temperature, and between δ13C and apparent oxygen utilization, occur in size fractions <350 µm. We studied the effects of dissolution on all three proxies using the Globorotalia menardii Fragmentation Index (MFI). We find no effect of dissolution on shell stable isotope geochemistry in core tops from the EEP. Shell dissolution has an observable effect on the Mg/Ca of N. dutertrei shells, but this effect is difficult to quantify. We used stable isotope shell geochemistry to reconstruct N. dutertrei habitat depths in the EEP which reflect oceanographic current patterns. Habitat depths estimated with the Mg/Ca paleo-thermometer do not agree with those calculated from δ18O data. This is mostly due to the gross disparity between estimates made with various Mg/Ca-temperature calibration equations, many of which are not calibrated with habitat depth temperatures. The difference of means ranges between 5 and 7 °C for temperature estimates made with two independent Mg/Ca –temperature calibrations for the Pacific. We do not find a discernible effect of test size in the disparity of temperature estimates with Mg/Ca; however calibrating Mg/Ca-temperature equations with δ18O-based habitat depth temperatures yields more robust temperature estimates.