Altitude-specific differences in tree-ring δ2H records of wood lignin methoxy in the Qinling mountains, central China

The Qinling Mountains (Qinling Mts.) are characterized as a distinct transition of bioclimatic zones along the north-south geographic boundary in China. Although ongoing global warming may affect the growth of most tree species across elevation gradients, the effect of modern warming on the stable hydrogen isotopes of tree-ring wood is still unknown. In this study, we developed early- and late-wood δ2H chronologies from 1900 to 2018 CE of lignin methoxy groups (δ2HLM) at 700-m intervals from 1700 to 3100 m a.s.l. along the elevation gradient in Mount Taibai (Mt. Taibai), the peak of the Qinling Mts. The δ2HLM values of the earlywood were relatively enriched in 2H compared to the latewood, and an elevation-dependent depleting trend along the altitudinal gradient was observed. Comparing the δ2HLM values with modelled δ2H values of precipitation (δ2HPre) of the sample sites, similar isotopic patterns can be obtained. The reconstructions of δ2HPre values confirm the strong coherence between the isotopic composition of the source water and the methoxy groups in the growing season. The consistency of δ2HLM chronologies between the early- and late-wood at the higher sites was stronger than that at the lower sites. The additional lagged effect and climate response revealed significant seasonal and altitudinal differences. At the timberline site, δ2HLM values of earlywood were mainly related to the January and February temperatures, whereas δ2HLM values of latewood correlated most strongly with temperature from July to September, and both values were controlled by the amount of total precipitation. In addition, the combined new δ2HLM chronologies, based on the lag-1 autocorrelation and the weighted average of the current year earlywood and previous latewood, could be used to merge the corresponding temperature signals from the middle- and low-elevation sites. Altitudinal correlations of elevation offset between the annual δ2HLM chronologies and interpolations of precipitation and temperature suggested that the δ2HLM values at higher elevations better represent regional climatic changes. Therefore, topographic effects such as elevation differences should be considered in large-scale applications of δ2HLM values in future studies.