Validation of a simple canopy conductance model for estimating transpiration of different citrus species under non-limiting soil water conditions

Improved estimates of transpiration (T) in citrus orchards are important to sustain production, especially in areas with limited water availability. Citrus trees exhibit stomatal control over T; with transpiration mainly modulated by canopy conductance (G_c) and vapour pressure deficit (VPD), suggesting that these would be important parameters in any citrus water use model. A study was therefore conducted to calibrate and validate a simple G_c model that estimates transpiration as a function of total daily radiation intercepted and VPD, together with derived parameters that represent radiation use efficiency and the response of G_c to VPD. The study was conducted in different citrus species with varying canopy sizes, grown in summer and winter rainfall regions of South Africa. The species used in the study were: ‘Star Ruby’ grapefruit, ‘Midknight’ Valencia orange, ‘Valley Gold’ mandarin and ‘Nadorcott’ mandarin. The aim of the study was to validate a G_c approach for estimating T of a wide range of citrus orchards to address the uncertainty of using a demand-limited model, such as the crop coefficient approach, in a species that is supply limited. In all the experimental orchards, T was measured with the heat ratio sap flux density method. The model was found to be more reliable for estimating monthly transpiration than for daily estimates in all orchards. On a monthly time, scale, acceptable statistical criteria were observed, with Wilmott index of agreement (D) >0.8, mean absolute percentage error (MAPE) <15%, root mean square error (RMSE) <2.5 mm day^‑1 and R²>0.70. However, discrepancies were observed on a daily time scale, particularly under conditions of low atmospheric evaporative demand. The good estimates of monthly T suggest that the model could be very useful for making strategic decisions regarding water management practices and planning. Improvement is needed for better daily estimates of T, as this will be important for tactical decisions, such as irrigation scheduling.