Core-top calibration of paleotemperature geochemical proxies: a case study in the Southeast Brazilian continental margin

Reconstructing past sea surface temperature conditions is valuable to observe and evaluate the climate of the past. In the SW Atlantic, so far, few studies have exanimated the applicability of paleotemperature equations, which may reflect in fewer reliable SST estimates. This study aimed to evaluate two marine proxies for SST reconstructions: the alkenones unsaturation index (UK\'37) and the ratio Mg/Ca in different planktonic foraminifera species and size fractions (G. ruber pink; G. ruber white senso stricto and senso lato; between 250 - 300 µm and 300 - 355 μm; G. truncatulinoides (d) crusted and non-crusted forms, between 380 - 620 μm; and G. inflata 300 - 425 μm). The samples were collected in the shelf break of South Brazilian Bight (SBB, 22 °S to 27 °S). The proxies were converted into SSTs and then compared to modern SSTs from the World Ocean Atlas and satellite images. The first chapter describes the UK\'37 signals and the applicability of three different paleotemperature equations, and the second chapter describes the Mg/Ca a proxy for past SSTs and water mass conditions after applying the most used paleotemperature equations for the ratio Mg/Ca. The UK\'37 seems to record mostly autumn conditions at 0m, and the most recent Bayspline equation results into more similar temperature averages when compared to modern annual SSTs (p-value of 0.81, n = 47, ΔSST of -0.03°C ± 0.27), while the most used linear equations result into SSTs similar to autumn conditions. In subtropical regions with low-temperature variations (∼4°C), the function UK\'37 versus SSTs works in an exponential relation, and they are related to seasonal temperatures. For the Mg/Ca in planktonic foraminifera species, the Mg/Ca-inferred temperatures agree well with modern ocean temperatures. Our data suggest that in different size fractions the tested species respond differently to the most used species-specific and general species paleotemperature equations. All the upper water column species agree well with temperatures at 0-meter depth. The G. ruber (p) responds well to the SS equation, where the smallest and the largest size results into annual and autumn estimates, respectively (p-value of 0.81 and 0.71, n = 23); the smallest size of G. ruber (w) s.s. provides summer estimates applying the GE equation (p-value of 0.21, n = 7), while the largest size provides annual estimates (p-value of 0.86, n = 13). The G. ruber (w) s.l. has significant averages when compared the largest size with autumn and summer conditions (p-value of 0.21 for both, n = 16); the deep-dwelling species G. inflata and the G. truncatulinoides (d) crusted form reflect the uplift of the SACW, calcifying at 10 and 20 m depth (p-value of 0.89, n = 10 and p-value of 0.06, n = 14, respectively), while the G. truncatulinoides (d) non-crusted form records annual temperature in deep water layers conditions (p-value of 0.06, n = 16). This is the first study to report Mg/Ca ratios in surface-dwelling and deep-dwelling planktonic foraminifera tests obtained in core-top samples at the SBB. This validation will inform the reconstruction of past environmental conditions of SW Atlantic, especially in the SBB. (AU)