Hydrogen isotope ratios of terrestrial leaf wax n-alkanes from the Tibetan Plateau: Controls on apparent enrichment factors, effect of vapor sources and implication for altimetry

Empirical evidence suggested that the altitudinal dependence of hydrogen isotope ratios of leaf wax n -alkanes ( δ D wax ) can be used to estimate paleoaltitudinal changes. However, the application of δ D wax -based paleoaltimetry remains difficult, as the impacts of evaporative, transpirative and biosynthetic processes on hydrogen isotope fractionations in changing environments and the influence of likely changing water vapor sources are not well explored. For this study, we sampled stream waters, soils and plant leaves along two transects spanning large gradients of altitude, precipitation amount, vapor source, temperature and vegetation type on the Tibetan Plateau (TP). δ D values of stream water (as an approximation for δ D p ), soil water ( δ D sw ) and plant leaf water ( δ D lw ) as well as leaf wax n -alkanes were measured in order to quantify isotopic fractionations in the formation of leaf waxes. Most interestingly, we found a strong negative correlation between the evapotranspirative enrichment of leaf water against precipitation (e lw-p ), which combines the effects of soil evaporation and leaf transpiration, and the biosynthetic hydrogen isotope fractionation (e wax-lw ), which describes isotopic enrichment between leaf wax and leaf water. The relationship yields a steady apparent isotopic enrichment factor (e wax-p ) between leaf wax and precipitation, which is independent from climatic parameters and has an average value of −107 ± 26‰ for grasses (monocotyledons) and −77 ± 22‰ for trees (dicotyledons). Since the terrestrial n -alkanes, especially n -C 27 and n -C 29 , in sediments are derived from trees and grasses, the likely change of the vegetation type in the uplift of mountains can change the isotopic estimates by about ±30‰, which corresponds to an altitudinal change of ∼1600 m. We, therefore, suggest that hydrogen isotope ratio of sedimentary n -C 31 alkane, which is mainly derived from grasses might be better proxies to reconstruct paleoaltitudes. Our large dataset of δ D wax from trees and grasses that aimed to mirror the variability of environmental factors over geological time frames showed the lapse rates were significant, but much smaller than in previous studies. Most importantly our result demonstrated that the lapse rate significantly differed for both transects ( p = 0.0068), i.e. 0.87 ± 0.71‰/100 m ( R 2 = 0.28, p = 0.2841, n = 6) and 2.28 ± 0.82‰/100 m ( R 2 = 0.34, p = 0.0135, n = 17) for Indian monsoon and Westerly dominated areas, respectively. This suggests that different moisture sources might strongly affected the observed lapse rates. In consequences altitude reconstructions are strongly complicated in areas with likely changing air masses like the Tibetan Plateau.