Fractionation of δ2H during root water uptake is linked to plant transpiration in saplings of a temperate tree species

The use of the stable isotopes of water to trace the origin of plant-accessed water is based on the assumption that there is no isotopic fractionation during root water uptake. In many field studies, however, xylem water isotopes do not match any of the potential sources considered, although this is rarely explicitly addressed. Recent evidence suggests that discrimination against the heavier isotopes might occur during root water uptake, but the factors affecting fractionation still remain poorly understood. We tested the potential factors driving isotopic fractionation at the soil-root interface under controlled conditions. Fagus sylvatica saplings potted in three different soil types were subjected to a drought treatment. In addition, a subset of plants was subjected to a reduction of vapor pressure deficit by covering them with plastic bags. Soil, root, and stem d²H and d¹⁸O were measured in all the treatments. Overall, the d¹⁸O of root and stem water matched that of pot soil water. In contrast, d²H of root and stem water was more depleted than that of soil water (-7.3±1.1‰). This difference in d²H between plant and soil was only observable when soil gravimetric water content was above 12%. As drought progressed and transpiration ceased, plant water became more enriched than pot soil water. The magnitude of depletion of stem and root water under relatively wet conditions was correlated with the difference between predawn and midday twig water potential, i.e. xylem water became more depleted in d²H with respect to the soil with larger transpiration. In turn, the d²H of soil water was more enriched for those plants that transpired more. This transpiration-driven enrichment of soil water was confirmed by a smaller soil-stem d²H offset in plants experiencing lower vapor pressure deficit. The soil type only affected the isotopic patterns through an early or delayed onset of plant transpiration cessation. We show that at least in F. sylvatica, isotopic fractionation of δ²H between soil and plant can occur under well-watered conditions, so it is not exclusive of xerophytic and halophytic plants. On the other hand, d¹⁸O seems to be a more reliable tracer of plant water sources. We argue that the correct identification of plant water sources should account for possible d²H fractionation during root water uptake.