Experimental Constraints on the Impact of Shell Dissolution on the Mg/Ca Temperature Relationship in the Polar Foraminifera Species Neogloboquadrina pachyderma

Mg/Ca and δ¹⁸O in foraminiferal shells are the most commonly used paleoclimate proxies for ocean temperature, salinity, and global ice volume. It is hypothesized that preferential dissolution of more soluble shell features and heterogenous shell chemistry as well as recrystallization can alter original shell chemistry as foraminifera sink below the lysocline into the deep ocean to be deposited as sediments. Core top studies have provided valuable insight into this process using closely grouped core tops with similar calcification conditions at the sea surface and different bottom water carbonate chemistry. They have proposed corrections for downcore Mg/Ca and/or δ¹⁸O based paleoclimate records using size normalized shell weights or past carbonate chemistry. However, it is impossible to unequivocally disentangle the effects of original calcification conditions and dissolution on core top shell chemistry at multiple sites. Recent studies have shown that foraminiferal shell dissolution can be simulated through lab-based experiments and dissolution intensity constrained using X-ray microcomputed tomography (Micro-CT) scans of the dissolved shells. We investigated dissolution effects by simulating dissolution of Neogloboquadrina pachyderma, the dominant polar foraminiferal species. Living N. pachyderma were collected from one station in the Greenland Sea via plankton tows and frozen at -80°C prior to dissolution experiments. Aliquots of 200 and 300 picked shells were dissolved in acidified Labrador Sea bottom water on a shaker table for two and three days respectively. An additional 189 shells were kept undissolved as a control. Micro-CT scans show that the average percentage of low-density calcite increased from 33.3 ± 3.6% for the undissolved shells, to 39.2 ± 4.3% for the two-day dissolved shells, to 44.8 ± 6.5% for the three-day dissolved shells, demonstrating an increase in dissolution intensity with an increase in dissolution time. We present δ¹⁸O and solution and LA-ICPMS based Mg/Ca results from the experiments that constrain the impact of dissolution on foraminiferal shell chemistry and propose a new framework for handling dissolution when generating foraminiferal chemistry-based paleoclimate records.