气候变化领域集成服务门户

The Mg, Sr and Ca contents of single shells of the ostracod Eucypris inflata from sediments retrieved from Sugan Lake, on the northern Tibetan Plateau, were analysed in order to reconstruct the salinity changes during the past similar to 850 yr. At present, the lake water has a Mg/Ca ratio > 40 and carbonates precipitated in the lake in the summer are mainly monohydrocalcites. However, X-ray diffraction analysis or core sediments suggests that several carbonate phases, including aragonite and even possibly dolomite, may have precipitated in the past. During the period (before 1620 AD) when there was aragonite precipitation, the Sr/Ca ratios of ostracod shells were at a minimum and showed little variation. When there was no aragonite precipitation (after 1620 AD), the Sr/Ca ratios of ostracods may have reflected salinity changes. The generally low Mg/Ca ratios during most of the past 850 yr may have resulted from decreased Mg partitioning in high-Mg/Ca host water as ostracods physiologically exclude excess Mg. Therefore, the Mg/Ca of ostracods cannot be used as a salinity proxy. The short-lived Mg/Ca maxima after 1300 AD correspond to decreased Sr/Ca and oxygen isotopes, indicating brier periods of freshwater input. The peaks of Mg/Ca ratios showed a similar variation with temperature change during the past 700 yr, indicating possible temperature dependence during times when the Mg/Ca ratio of the water was reduced. Water chemistry, temperature and mineralogical processes may have all affected the trace-element ratios of ostracods. Combined Sr/Ca ratios, carbonate mineralogy and oxygen isotopes allowed the lake's salinity to be reconstructed during the past similar to 850yr. In enclosed lakes in hyper-arid regions, especially in high-Mg/Ca waters, the trace-element chemistry of ostracod shells should be treated with care and the complex controls on trace-element should be examined before paleosalinity or paleotemperature reconstruction is attempted. (C) 2008 Elsevier Ltd and INQUA. All rights reserved.