Coupled DSSAT-SWAT models to reduce off-site N pollution in Mediterranean irrigated watershed

In any agricultural watershed, best management practices (BMPs) are a conservational way to reduce non-point source pollution and, soil and water resources sustainability. The objectives of this study were to calibrate and validate the modified Soil and Water Assessment Tool (SWAT) in the Violada Watershed (VW), Spain and assess the BMPs scenarios, already tested at field scale, using the Decision Support System for Agro Technology Transfer model (DSSAT), and finally, to access the BMPs impact on water quality off-site. To this end, daily streamflow discharge and NO₃^- concentration were measured at VW outlet from October 2015 to September 2017 for model evaluation. The SWAT-CUP was used for sensitivity analysis, calibration and validation for both measured variables after manual calibration of the main crops yield. Three management scenarios were compared to the current conditions (baseline): (i) recommended N fertilization, (ii) optimum irrigation and; (iii) combined recommended N fertilization and optimum irrigation (combined BMPs). The SWAT crop growth model calibration demonstrated that the annual average of crop yield and actual evapotranspiration estimations were satisfactory. Monthly calibration and validation results were satisfactory for streamflow discharge and NO₃-N load, with Nash-Sutcliffe efficiency (NSE) according to the criteria reported in the literature. The two individual scenarios results showed difference in their environmental impact and therefore combined BMPs scenario was considered more efficient in reducing NO₃-N load (51%) than the recommended N fertilization (36%) and the optimum irrigation (12%), while including all additional environment and farmers' benefits of both individual scenarios. Under this combined scenarios, all crops yields were maintained or increased, and the total irrigation water and N mineral fertilizers application reduction were about 5% and 27%, respectively. However, further work is still needed to consider additional BMPs to limit the soil N residual losses during the non-cropped period. The applied methodology can be a good alternative for improving water quality in similar irrigated watersheds.