Multi-site investigation of the determinants of evapotranspiration partitioning with a mobile water isotope laboratory

Quantifying and partitioning evapotranspiration (ET) of agricultural ecosystems in various environmental settings enable studying the site-specific determinants of plant water use. The aim of the study was to conduct reliable and reproducible field-scale partitioning of ET into its component fluxes soil evaporation (E) and plant transpiration (T) from water stable isotope analysis (δ²H and δ¹⁸O). Isotope-based partitioning methods, because of their methodological independence to other, traditional, experimental or data-driven approaches, are useful for intercomparison. Campaigns were carried out at two agricultural field sites of the ICOS ecosystem thematic network, differing in their hydroclimate and crop settings. Isotopic partitioning was achieved by simulating (i) the isotopic composition of ET (δ_ET) from atmospheric water vapor measurements and (ii) δ_E and δ_T from simultaneous destructive sampling of soil water and plant xylem water. Campaigns were carried out from June 22 to July 19, 2022 (sunflower crop, mean air temperature and relative humidity: 24 °C and 65 %) and from March 27 to April 18, 2023 (winter wheat, 12 °C and 75 %) in the Mediterranean site FR-Aur (Auradé, France), and from June 5 to August 26, 2024 (winter wheat, 19 °C and 73 %) in the temperate site DE-RuS (Selhausen, Germany). Up to three measurements per day of isotope-based partitioning results were confronted against estimates of ET obtained from on half-hourly eddy-covariance data. Non-isotopic ET partitioning was calculated based on simulations using half-hourly sap flow- (T) and daily microlysimeter measurements (E) during the 2022 campaign in Auradé. Both the non-isotopic and isotopic data showed an increase in daily T/ET ratios during the 2022 campaign. Daily mean T/ET ratios were 0.79 from sap flow/EC data, 0.75 from sap flow/microlysimeter data, and mean sub-daily isotopic T/ET ratios of 0.66 for sunflowers in 2022. T/ET ratios of winter wheat in Auradé 2023 showed a mean value of 0.92. The differences between the isotopic and non-isotopic T/ET ratios in 2022 might be a result of differences in measurement footprint, as field-scale EC-based partitioning was compared to sub-field scale isotopic partitioning. Estimation of T/ET uncertainty, calculated as from propagation of errors of the individually conducted measurements, was provided. While errors of daily sap flow/EC partitioning were lower compared to microlysimeter/EC and isotopic partitioning, errors of sub-daily EC/sap-flow T/ET exceeded the errors of the other two approaches. In addition, values of sap flow/EC T/ET increased over 100% from the late afternoon, showing a limitation of the sap flow/EC-based partitioning method on the sub-daily timescale. During the 2024 campaign, isotopic measurements were performed at an hourly resolution, and analysis of isotopic and non-isotopic T/ET ratios for the 2024 campaign is pending. The aim of future campaigns is the continuation of intercomparison between partitioning methods and the identification of differences and fit among T/ET partitioning approaches specifically to the considered temporal and spatial scales.