Calibration and application of the 'clumped isotope' thermometer to foraminifera for high-resolution climate reconstructions

The reconstruction of past ocean temperatures is fundamental to the study of past climate changes, therefore considerable effort has been invested in developing proxies for seawater temperatures. One of the most recent and promising new proxy is carbonate 'clumped isotope' thermometry, in particular because it is based on thermodynamic equilibrium and not on bio- geochemical proxies. Here, we present a new calibration of the 'clumped isotope' thermometer to foraminifera based on seven species of planktic and benthic foraminifera spanning a growth temperature range of 2-28 C. We used a newly developed technique for the measurements of small samples to improve the applicability of this method to paleoceanography. Our data have a comparable precision (0.005-0.013&) and confirm previous calibration studies based on biogenic and inorganic cal- cite. We discuss possible sources of uncertainty such as over-/underestimation of the calcification temperatures, species-spe- cific vital effects, pH variations between the seawater and the vacuole water of the species and possible kinetic effects on the 'clumped isotope' calibration. To validate our calibration study and test the applicability of our measuring technique to paleoclimate and paleoceano- graphic studies we measured the isotope composition of Globigerinoides ruber (white) at high-resolution in a sediment core covering the last 700 years in the Gulf of Taranto (Mediterranean Sea). The results show that it is necessary to average a rel- atively large number of analyses to achieve a consistent temperature signal for the detection of small sea surface temperature changes. Although with the current analytical system, 'clumped isotope' thermometry is only applicable to the analysis of rel- atively large SST changes in marine sediments, further technical improvements may make this a very powerful technique for paleoceanographic studies.