Parameterization of the AquaCrop model for simulating table grapes growth and water productivity in an arid region of Mexico

Currently, the AquaCrop model has been widely tested for many fruit/grain crops; root and tuber crops; leafy vegetables, or forage crops, but is restricted to annual herbaceous species, while deciduous crops have received less if no attention. In this context, this study aims to test for the first time the ability of the AquaCrop model to simulate canopy cover (CC), actual evapotranspiration (ETa), total soil water content (TWC), biomass (B) and fruit yield (FY) of table grapes vineyards (Vitis vinifera L., cvs. Perlette and Superior) at the Costa de Hermosillo, Sonora in Northwest Mexico. Observed weather and soil physical parameters, with measured crop parameters from an experiment conducted during 2005 were used to develop climate, soil and crop input files for AquaCrop and for calibrating the model. While collected data during the 2006 growing season were used to validate the model. The model adequately simulated CC, ETa and TWC during 2005 and 2006. The Root Mean Square Error (RMSE) between observed and measured CC, ETa and TWC were 5.18%, 0.46 mm/day and 10.11 mm during 2005, and 8.82%, 0.84 mm/day and 9.1 mm during 2006, respectively. The good accuracy of simulations of CC, ETa and TWC by the model have been confirmed by additional statistical parameters like the coefficient of determination (R-2), The Mean Bias Error (MBE), the Willmott's index of agreement (d) and the Nash-Sutcliffe Efficiency (NSE). For the B and FY simulations, the results showed that the model correctly reproduced the B and FY with NRMSE value of 8.8%. The estimated average value of FY (14.56 t/ha) for both seasons are in the range of the potential yield (14-18 t/ha) of table grapes in the irrigated Costa de Hermosillo in northwest Mexico. After the validation of the AquaCrop model, it was used to evaluate the irrigation scheduling by the farmer as well as to assess the water productivity computed as the ratio of crop production to crop water use. The results showed that, the recommended irrigation by the model was about 547 mm and 509 mm, which it is about half of that applied by the farmer (1006 mm and 929 mm) during 2005 and 2006, respectively. This large difference, which represents approximately 54% and 57% of the irrigation supply, is lost through deep percolation and could be saved without vegetation suffering from water stress while maintaining the same yield. The high loss of water by percolation affects significantly the water productivity (WP), which decreases from 3.22 to 1.74 kg/m(3) if we consider the transpiration (WPTr), and the sum of ETa and Percolation (WPETa+Pr) for WP computations, respectively. Consequently, the AquaCrop model can be used as an operational tool by decision makers and growers to improve irrigation management. This is of crucial importance in arid and semi-arid regions where water is becoming increasingly scarce