Alpine Holocene tree-ring dataset: age-related trends in the stable isotopes of cellulose show species-specific patterns

Stable isotopes in tree-ring cellulose are important tools for climatic reconstructions even though their interpretation could be challenging due to nonclimate signals, primarily those related to tree aging. Previous studies on the presence of tree-age-related trends during juvenile as well as adult growth phases in δD, δ18O, and δ13C time series yielded variable results that are not coherent among different plant species. We analyzed possible trends in the extracted cellulose of tree rings of 85 larch trees and 119 cembran pine trees, i.e., in samples of one deciduous and one evergreen conifer species collected at the tree line in the Alps, covering nearly the whole Holocene. The age trend analyses of all tree-ring variables were conducted on the basis of mean curves established by averaging the cambial-age-aligned tree series. For cambial ages over 100 years, our results prove the absence of any age-related effect in the δD, δ18O, and δ13C time series for both the evergreen and the deciduous conifer species, with the only exception being larch δD. However, for lower cambial ages, we found trends that differ for each isotope and species; i.e., mean δ13C values in larch do not vary with aging and can be used without detrending, whereas those in cembran pine show a juvenile effect, and the data should be detrended. Mean δ18O values present two distinct aging phases for both species, complicating detrending. Similarly, mean δD values in larch change in the first 50 years, whereas cembran pine changes between 50 and 100 years. Values for these two periods of cambial age for δD and δ18O should be used with caution for climatic reconstructions, ideally complemented by additional information regarding mechanisms for these trends.