No oxygen isotope fractionation for subfossil cellulose during diagenesis

Nonpermineralized, subfossil wood samples collected from within Cenozoic aged sediments are an important substrate for reconstructing paleoclimate under different levels of atmospheric carbon dioxide. Stable oxygen isotope composition (δ¹⁸O) of cellulose chemically extracted from these samples have been used to quantify seasonal to interdecadal climate; however, the effect of diagenesis on δ¹⁸O values in subfossil cellulose is not clear. In this paper, (1) we experimentally simulate the factors influencing isotopic fractionation during diagenesis, including hydrothermal conditions and fungal degradation, and demonstrate that the δ¹⁸O values of the reacted residues remained nearly unchanged after cellulose extraction; (2) analyze intra-annual δ¹⁸O variation for a subfossil wood sample exhibiting varying degrees of diagenesis, and showed that δ¹⁸O across different degradation levels follow a similar pattern (p < 0.001). Based on these results from both laboratory experiments and analysis of subfossil samples, we conclude that the δ¹⁸O value of cellulose is not affected by diagenesis. Thus, the δ¹⁸O of cellulose can be used to reliably reconstruct palaeoclimate in deep time.