The aridity influence on oxygen isotopes recorded in tree rings

Abstract The stable isotopes of oxygen in wood cellulose (δ18Ocell) have been widely used to reconstruct historical source water use in trees or changes in atmospheric humidity. However, in many cases, the δ18O of source water use is assumed to reflect that of precipitation, which is often not the case in semi-arid to arid ecosystems where trees use deeper and older water from previous precipitation events (or even groundwater). Furthermore, the degree to which δ18Ocell reflects source water and atmospheric aridity depends on pex, normally defined as the proportion of oxygen atoms that exchange between isotopically enriched carbohydrates from the leaf and unenriched xylem water during cellulose synthesis. Many studies treat pex as a constant. However, pex can only be estimated with direct measurements of δ18Ocell and the δ18O of tree source water and sucrose. Additionally, other physiological mechanisms (e.g. photosynthate translocation) can alter the isotopic signal before cellulose is produced. Thus, determining this "apparent pex" (apex; which includes photosynthate translocation plus the exchange of oxygen atoms during cellulose synthesis), can be difficult. In this study, we collected δ18O of xylem water and δ18O of wood cellulose from seven stands of Ponderosa pine situated at the northern boundary of the North American Monsoon climate system to assess how potential variability in apex influenced how source water and aridity were recorded in δ18Ocell. We compared measured and modeled values of δ18Ocell and found that more arid sites under-represented the vapor pressure deficit signal in cellulose while wetter sites over-represented the vapor pressure deficit signal in cellulose. We also found that apex varied as a function of site aridity, where low precipitation and high vapor pressure deficit led to high apex, while high precipitation and low vapor pressure deficit led to low apex. Future studies can use our emerging understanding of the aridity-apex relationship in different portions of the annual ring to better disentangle the source water and VPD signals in cellulose, particularly for regions such as the North American Monsoon region.