Effects of wood degradation on tree-ring stable isotopes used in climate reconstructions

Tree-ring stable isotopes from relict wood samples play an important role in climate reconstructions, but fire exposure, waterlogging, and post-excavation storage may bias the isotopic signal. Here, we test how thermal alteration and storage conditions affect holocellulose δ 13 C and δ 18 O values in archaeological oak ( Quercus spp.) samples from southern Germany, focusing on the 1430–1444 CE period and considering a range of preservation states. We differentiate (i) dry-stored wood across four thermal-alteration categories (unaffected, intact, moderately charred, fully charred) and (ii) waterlogged wood kept wet after excavation, sampled at the outer surface and in the interior. We found no significant differences in holocellulose δ 13 C and δ 18 O values among unaffected, intact and moderately charred rings. Fully charred rings yielded no extractable holocellulose; however, their bulk-wood δ 13 C values were systematically shifted to lower values while preserving the interannual variability. In waterlogged wood, surface holocellulose δ 18 O values were lower and showed weaker coherency with the interior series, whereas holocellulose δ 13 C showed no meaningful change. We conclude that moderate charring and waterlogging have limited effects on holocellulose-based tree-ring isotope signals in archaeological wood. For δ 18 O in waterlogged wood, shallow surface samples should be avoided where possible, or a correction should be applied based on the mean offset across overlapping years. Our results support the inclusion of partially degraded archaeological and subfossil oak wood in isotope-based dating and paleoclimate reconstructions.