Sampling strategy and climatic implications of tree-ring stable isotopes on the southeast Tibetan Plateau

We explore the potential of tree-ring cellulose δ18O and δ13C records for reconstructing climate variability in the southeast Tibetan Plateau. Our sampling strategy was designed to investigate intra and inter-tree variability, and the effects of the age of tree on δ18O variation. We show that intra-tree δ13C and δ18O variability is negligible, and inter-tree coherence is sufficient to build robust tree-ring δ18O or δ13C chronologies based on only four trees. There is no evidence of an age effect regarding δ18O, in contrast with tree-ring width. In our warm and moist sampling site, young tree δ13C is not clearly correlated with monthly mean meteorological data. Tree-ring δ18O appears significantly anti-correlated with summer precipitation amount, regional cloud cover, and relative humidity. Simulations conducted with the ORCHIDEE land surface model confirm the observed contribution of relative humidity to tree cellulose δ18O, and explain the weak correlation of δ13C with climate by the non-linear integration linked with photosynthesis. Altogether, the tree-ring cellulose δ18O is shown to be a promising proxy to reconstruct regional summer moisture variability prior to the instrumental period.