Pairing plant-wax H and C isotopes with lake-area – A method for evaluating the local amount effect in northern China during the late Quaternary

The correlation between modern tropical mean annual precipitation (MAP) and its isotopic composition (δ_p) has been extensively used to interpret terrestrial paleoclimate records. However, the existence of this phenomenon in the past and in individual regions remains a matter of debate. Here, we describe a method for evaluating the relationship between local δ_p and local MAP in the past (i.e. the “local amount effect”) and apply this method to a closed-basin lake in northeastern China. We measured compound specific hydrogen isotopes from plant-wax derived long-chain n-alkanes (δ²H_wax) and fatty acids (δ²H_FA) from lake sediments, which primarily record local δ_p, and compared them with the lake level and area history of the lake, which records local MAP. Both the lake level reconstruction and the organic isotope proxies are derived from the same samples and thus enable a direct comparison. In addition, we explore the possibility of using chain-length distribution and δ²H_wax data to assess sediment reworking in the lake basin. Our results show that the n-C₂₉ n-alkane δ²H_wax and C_16:0n-acid δ²H_FA are correlated with lake level over the past 15 to 3 ka from samples where both age and lake level are well constrained (n = 5 and n = 3, respectively). Over the past 2–3 ka this correlation breaks down: lake level remains low, following summer insolation, but the Chinese cave δ¹⁸O and modern δ²H_wax become progressively more negative. The reason for this ‘2 ka shift’, first identified in Chinese cave δ¹⁸O records, is unclear. Our data suggests that δ_p is decoupled from MAP during this time and thus adds an additional constraint to solving this issue. Our results show that combining lake-area reconstructions with organic isotope geochemistry from closed-basin lakes can be used to assess the relation between local rainfall amount and local δ_p in the past, but requires a thorough assessment of sediment dynamics of each individual sample.