Appraisal of alkenone- and archaeal ether-based salinity indicators in mid-latitude Asian lakes

Lake water salinity, an important indicator of lake hydrological conditions, is critical to deciphering terrestrial paleoclimatic and paleoenvironmental changes. The proportion of C-37:4 alkenone to total C-37 alkenones (%C-37:4) and the relative abundance of archaeol to caldarchaeol (ACE) are promising salinity indices in lacustrine settings. Developing regional calibrations for both indicators is challenging due to limited datasets and a poor understanding of the factors that influence them. Here we present %C-37:4 and ACE data collected from 55 lakes in mid-latitude Asia with a wide salinity range, to evaluate how well the two proxies can be correlated with salinity and to infer potential influencing factors. The %C-37:4 values only show a weak negative correlation with salinity from all investigated lakes. The occurrence of C-37:3 alkenone isomer and abnormal values in the alkenone C37/C38 ratio might signal species shifts or environmental factors that compromise the %C-37:4-salinity relationship in our data set. Seasonal bias in alkenone production could further obscure this %C-37:4-salinity relationship. A stronger relationship emerges after removing samples influenced by these factors: %C-37:4 = -8.56 * log(10) (salinity) + 80.6, r(2) = 0.62, n = 37; or %C-37:4 = -13.46 * log(10) (salinity) + 101.48, with uncertainty in both variables considered. The ACE values show a strong positive correlation with salinity from all investigated lakes: ACE = 2.27 * 10(-4)* salinity + 25.4, r(2) = 0.75, n = 68; or ACE = 1.86 * 10(-4)* salinity + 38.1, with uncertainty in both variables considered. However, substantial deviations of the ACE values in low salinity range (<60,000 mg/L) and majority of the ACE shift in the range of 60,000-100,000 mg/L suggest that the ACE primarily responds to Euryarchaeota/Archaea community changes when a salinity threshold is crossed. Accordingly, both lake salinity indicators have the potential to reconstruct past salinity changes when their influencing factors could be constrained, and the two independent proxies, when they are used together, could further refine salinity reconstructions. (C) 2020 Elsevier B.V. All rights reserved.