Using AquaCrop Model to simulate irrigation water use efficiency of potato crop under semi-arid conditions of Central Tunisia

In Tunisia, water scarcity forces producers to face stress conditions. In this study, AquaCrop was used to reproduce the dynamic of water contents, vegetative growth, yield production and water use efficiency under a non-stressed and water stressed treatments. Calibration procedure aimed to use in maximum default parameters of AquaCrop. Since, the paper presented only the parameters that have to be adjusted to obtain similar results of field measurements. Root mean squared error, RMSE, values were always lower than 0.04 cm3.cm-3 for water contents lower than 0.06 for vegetation cover estimation. Moreover, results from Nasch Coefficient, E, were almost equal to one. RMSE and E justified that the model was well assessed to predict the soil water contents and vegetation development under the study area. However, the model presented a greater performance in the case of full irrigation strategy. When evaluating different values of water productivity, it was showed that a WP of 32 g.m-2 produced the lowest estimation error. Regarding yield productions, statistical indictors, computed for a water productivity value of 32 g.m-2 show in general RMSE values lower than 0.4 t/ha. In addition, E was closer to one for the non stressed treatment, T1. For irrigation water use efficiency, it was depicted that the model underestimated field IWUE. Moreover, the discrepancy between simulated and estimated irrigation water use efficiency rose for treatment T2, implying that the model calibration should be improved, especially for stressed conditions. The model, after being calibrated, could be used for simulating the response of the crop to different irrigation management aiming to optimize water use efficiency.