Examining evapotranspiration in a semi-arid region using stable isotopes of hydrogen and oxygen

Abstract

Daily variation of the isotopic composition, both deuterium and oxygen-18, of water in foliage of a mixed Eucalyptus foecunda - Eucalyptus socialis, ‘mallee’ community was studied in relation to the isotopic composition of soil water and atmospheric water vapour, transpiration and evapotranspiration, and environmental conditions. The isotopic composition of water in foliage and the soil surface were enriched over adjacent stem and deeper soil water respectively, characteristic of water subject to evaporation. The composition of foliar water varied systematically over the day, being highest in the late afternoon, in response to lower relative humidity. The foliar water composition fitted a steady-state model of leaf water fractionation. It was found that the fraction of leaf water equilibrated with the atmosphere was at a minimum during the middle of the day, fitting earlier observations of an inverse correlation of this fraction with transpiration rate. The equilibrated fraction was relatively high compared with previous transpiration studies, consistent with low transpiration rates. The soil water isotope-depth profiles had not achieved steady state. Using existing models, we conclude that the foliage and soil compartments are exchanging water vapour, but that some complicated evaporation processes are occurring in the soil, probably as a result of daily temperature cycling. Examining the relationship of the two isotope tracers in the different compartments, the vapour transport fits a transport model in which it is strongly diffusion-limited, both within the soil and leaf. Because of some uncertainty as to the transport mechanisms involved, it is not clear what the relative contributions from soil and foliage are. The vegetation may be having a considerable influence on the total evaporation from the community, since the effects of the root system on the soil structure and soil water transport are not clear.