摘要 : The Mg/Ca ratio of the infaunal benthic foraminifer Uvigerina peregrina is a commonly used proxy for reconstructing bottom water paleotemperatures because it is hypothesized to be insensitive to changes in bottom water carbonate c... 展开 The Mg/Ca ratio of the infaunal benthic foraminifer Uvigerina peregrina is a commonly used proxy for reconstructing bottom water paleotemperatures because it is hypothesized to be insensitive to changes in bottom water carbonate chemistry. This study presents core-top U. peregrina samples from the southwest Pacific from 0.6 to 4.4 km water depth, corresponding to water temperatures of 1-9 degrees C. Samples from New Zealand's Bay of Plenty and Chatham Rise were compared to previous calibrations to assess Mg/Ca temperature sensitivity. Published core-top temperature sensitivies can explain U. peregrina Mg/Ca at a majority of our study sites, with the exception of sites at intermediate depths of similar to 2.4-3.3 km, where the Mg/Ca ratios are substantially higher than expected from these calibrations, possibly reflecting diagenetic or non-thermal effects. Stable isotope measurements (delta O-18 and delta C-13), laser ablation analyses, and scanning electron microscope imagery complement trace element measurements to assess variables potentially affecting U. peregrina Mg/Ca at these sites. Morphotype variability, contamination, dissolution, and recrystallization all failed to provide satisfactory explanations for anomalously high Mg/Ca observations. We infer that a non-temperature control, perhaps related to carbonate chemistry, may be affecting the Mg incorporation in some U. peregrina specimens, though no factor has yet been clearly identified. It is unclear whether the factor affecting these sites is constant through time or could vary. For this reason, we recommend that Mg/Ca of recent specimens at each core site should be checked against established calibration curves prior to pursuing down-core paleotemperature reconstructions using U. peregrina. Where possible, paleotemperature estimates should also be validated using other independent proxies. Existing core-top calibrations effectively predict U. peregrina Mg/Ca at the majority of our study sites, but our findings underscore the need for a more thorough understanding of non-temperature factors that can influence Mg/Ca in U. peregrina. (C) 2021 Elsevier Ltd. All rights reserved. 收起