Ecosystem-level water use efficiency and evapotranspiration partitioning in conventional till and no-till rainfed canola

Accurately determining productive water use (i.e., transpiration, T, component of evapotranspiration, ET) is crucial to improve ecosystem water use efficiency (EWUE) and develop water-saving management practices. High frequency (10 Hz) eddy covariance data for the 2016?2017 growing season (November 2016 ? May 2017) for conventional till (CT) and no-till (NT) canola (Brassica napus L.) fields were analyzed to partition ET into T and evaporation (E) using Flux Variance Similarity method. The objectives of this study were to determine and compare seasonality of T and T/ET ratios, and compare EWUE based on different metrics of carbon (gross primary production, GPP and net ecosystem production, NEP) and water vapor (ET and T) fluxes in CT and NT canola fields. Canopy stands, fluxes, and EWUE were similar until winter in both fields, but were substantially lower in NT field from February to April due to differences in recovery of canola stands after winter dormancy. However, differences were much smaller for water loss than carbon gain. Larger difference was observed for T than ET due to significant amounts of E during rainy periods under both dense (CT field) and sparse (NT field) canopies. The growing season average T/ET ratio was identical (-0.70) in both fields. The EWUE was -23% higher in CT and -26% higher in NT field based on T than ET for a dry period during peak growth. In comparison, EWUE was -36% higher in CT and -58% higher in NT field based on T than ET for a wet period during peak growth due to relatively higher loss of soil E from sparse canopy. The results not only illustrated discrepancies in EWUE based on ET and T for cross-site comparisons but also highlighted the added value of T estimates over ET for more meaningful EWUE estimates during the periods of higher E.