The accurate estimation of evapotranspiration () and its components are vital for water resource management and irrigation planning. This study models tree transpiration () and for grapefruit, litchi, and mango orchards using light gradient-boosting machine (LightGBM) optimized using the Bayesian hyperparameter optimization. Grounds and for these crops were measured using the heat ratio method of monitoring sap flow and the eddy covariance technique for quantifying . The Sentinel 2 satellite was used to compute field leaf area index (). The modelled data were used to partition the orchard into beneficial () and non-beneficial water uses (orchard floor evaporation—). We adopted the 10-fold cross-validation to test the model robustness and an independent validation to test performance on unseen data. The 10-fold cross-validation and independent validation on and models produced high accuracy with coefficient of determination () 0.88, Kling–Gupta efficiency () 0.91, root mean square error () 0.04 mm/h, and mean absolute error () 0.03 mm/h for all the crops. The study demonstrates that LightGBM can accurately model the transpiration and evapotranspiration for subtropical tree crops using Sentinel 2 data. The study found that which combined soil evaporation and understorey vegetation transpiration contributed 35, 32, and 31% to the grapefruit, litchi and mango orchard evapotranspiration, respectively. We conclude that improvements on orchard floor management practices can be utilized to minimize non-beneficial water losses while promoting the productive water use ().
In South Africa, as in other semi-arid countries, sustainable production of high-value crops requires precise management of limited water resources. We investigated daily and seasonal changes in stem and fruit growth as indicators of water stress in pear (Pyrus communis L.) trees, in the Western Cape Province. Stem and fruit growth data were collected hourly throughout the 2022–2023 growing season on 2 cultivars commonly planted in South Africa – Packham’s Triumph and Forelle. Soil water content, tree sap flow, and orchard microclimate were also monitored. Fruit maximum daily shrinkage (MDS) was highly sensitive to soil water deficit and more sensitive than stem size changes. However, the patterns of fruit MDS for both cultivars changed as the season progressed. Early in the season (October–December), there was a strong correlation between fruit MDS and soil water deficit (R2 ~ 0.72). The fruit shrunk with increasing soil water deficit as water loss through transpiration exceeded gains through xylem and phloem inflows. In contrast, daytime fruit size swelled from late December until harvest (February/March), likely because of the dominance of phloem inflows and decreased peel transpiration as the fruit matured. Correlation between fruit expansion and soil water deficit was weaker (R2 ~ 0.32) during the later stage even though fruit growth continued until harvest. Stem MDS consistently showed midday shrinkage throughout the season in response to soil water deficit, but with more scatter (R2~ 0.37). Seasonal total transpiration was greater for Forelle (733 mm) than Packham’s Triumph (539 mm) because of the higher leaf area index of the Forelle and the longer growing season. This study suggests that pear fruit growth data can provide accurate estimates of tree water status, but only during the early stages of growth. Towards maturity, fruit size changes respond indirectly to water deficit, possibly through reductions in photosynthesis.