An ostracod-based calibration function for electrical conductivity reconstruction in lacustrine environments in Patagonia, Southern South America

Abstract

In the Patagonian region (∼37–56°S) E of the Andes, the salinity and solute composition of lakes is strongly related to their location along the marked W-E decreasing precipitation gradient that is one of the main climatic features of the area. A calibration function (n = 34) based on 12 ostracod species (Ostracoda, Crustacea) was developed by WA-PLS to quantitatively reconstruct electrical conductivity (EC) values as a salinity proxy. The selected one component model had a r2 = 0.74 and RMSEP and maximum bias equal to 16% and 31% of the sampled range, respectively, comparable to other published ostracod-based calibration functions. This model was applied to the ostracod record of the closed lake Laguna Cháltel (49°58′S, 71°07′W), comprising seven species and dominated by two species of the genus Limnocythere. In order to evaluate the calibration function’s robustness, the obtained EC values were compared with qualitative lake level and salinity variations inferred through a multiproxy hydrological reconstruction of the lake. Both reconstructions show good overall agreement, with reconstructed EC values in the oligo-mesohaline range (average: 11 060 ± 680 μS/cm) between 4570 and 3190 cal BP, corresponding to the ephemeral and shallow lake phases, and a marked decrease in EC concurrent with a lake level rise, reaching an average EC of 1140 ± 90 μS/cm during the deep lake phase (1720 cal BP to present). The variability in the reconstructed EC values for the ephemeral lake phase showed some inconsistency with the expected trend, which was attributed to time-averaging effects; for its part, the pace of the decrease in EC during the medium-depth phase (3190–1720 cal BP) differed from the expected, which could be due to autigenic effects (redissolution of salts) at the onset of this phase. This comparison not only lends support to the adequacy of the calibration function, but also suggests that its application in the context of a multiproxy study can greatly contribute to distinguish between autigenic and climatic-related controls of paleosalinity in closed lakes, allowing performing more accurate paleoenvironmental inferences on the basis of paleohydrological reconstructions.