Environmental constraints on lake sediment mineral compositions from the Tibetan Plateau and implications for paleoenvironment reconstruction

Inorganic minerals form a major component of lacustrine sediments and have the potential to reveal detailed information on previous climatic and hydrological conditions. The ability to extract such information however, has been restricted by a limited understanding of the relationships between minerals and the environment. In an attempt to fill in this gap in our knowledge, 146 surface sediment samples have been investigated from 146 lakes on the Tibetan Plateau. The mineral compositions derived from these samples by X-Ray Diffraction (XRD) were used to examine the relationships between mineral compositions and the environmental variables determined for each site. Statistical techniques including Multivariate regression trees (MRT) and Redundancy Analysis (RDA), based on the mineral spectra and environmental variables, reveal that the electrical conductivity (EC) and Mg/Ca ratios of lake water are the most important controls on the composition of endogenic minerals. No endogenic minerals precipitate under hyper-fresh water conditions (EC lower than 0.13 mS/cm), with calcite commonly forming in water with EC values above 0.13 mS/cm. Between EC values of 0.13 and 26 mS/cm the mineral composition of lake sediments can be explained in terms of variations in the Mg/Ca ratio: calcite dominates at Mg/Ca ratios of less than 33, whereas aragonite commonly forms when the ratio is greater than 33. Where EC values are between 26 and 39 mS/cm, monohydrocalcite precipitates together with calcite and aragonite; above 39 mS/cm, gypsum and halite commonly form. Information on the local geological strata indicates that allogenic (detrital) mineral compositions are primarily influenced by the bedrock compositions within the catchment area. By applying these relationships to the late glacial and Holocene mineral record from Chaka Salt Lake, five lake stages have been identified and their associated EC conditions inferred. The lake evolved from a freshwater lake during the late glacial (before 11.4 cal. ka BP) represented by the lowest EC values (<0.13 mS/cm), to a saline lake with EC values slightly higher than 39 mS/cm during the early and mid Holocene (ca. 11.4–5.3 cal. ka BP), and finally to a salt lake (after 5.3 cal. ka BP). These results illustrate the utility of our mineral-environmental model for the quantitative reconstruction of past environmental conditions from lake sediment records.