The effect of warming on grassland evapotranspiration partitioning using laser-based isotope monitoring techniques

The proportion of transpiration (T) in total evapotranspiration (ET) is an important parameter that provides insight into the degree of biological influence on the hydrological cycles. Studies addressing the effects of climatic warming on the ecosystem total water balance are scarce, and measured warming effects on the T/ET ratio in field experiments have not been seen in the literature. In this study, we quantified T/ET ratios under ambient and warming treatments in a grassland ecosystem using a stable isotope approach. The measurements were made at a long-term grassland warming site in Oklahoma during the May-June peak growing season of 2011. Chamber-based methods were used to estimate the δ²H isotopic composition of evaporation (δ_E), transpiration (δ_T) and the aggregated evapotranspiration (δ_ET). A modified commercial conifer leaf chamber was used for δ_T, a modified commercial soil chamber was used for δ_E and a custom built chamber was used for δ_ET. The δ_E, δ_ET and δ_T were quantified using both the Keeling plot approach and a mass balance method, with the Craig-Gordon model approach also used to calculate δ_E. Multiple methods demonstrated no significant difference between control and warming plots for both δ_ET and δ_T. Though the chamber-based estimates and the Craig-Gordon results diverged by about 12‰, all methods showed that δ_E was more depleted in the warming plots. This decrease in δ_E indicates that the evaporation flux as a percentage of total water flux necessarily decreased for δ_ET to remain constant, which was confirmed by field observations. The T/ET ratio in the control treatment was 0.65 or 0.77 and the ratio found in the warming treatment was 0.83 or 0.86, based on the chamber method and the Craig-Gordon approach. Sensitivity analysis of the Craig-Gordon model demonstrates that the warming-induced decrease in soil liquid water isotopic composition is the major factor responsible for the observed δ_E depletion and the temperature dependent equilibrium effects are minor. Multiple lines of evidence indicate that the increased T/ET ratio under warming is caused mainly by reduced evaporation.