Modeling water quality to improve agricultural practices and land management in a tunisian catchment using the soil and water assessment tool.

Agriculture intensification has impaired water quality. In this study, the risk of pollution by nitrates was assessed by experimental monitoring, spatial integration of farm census, and modeling of water quality using the Soil and Water Assessment Tool (SWAT), version 2009, over the period of 1990 to 2006 for a catchment located northern Tunisia. Under a semiarid climate, the water quality is influenced by the predominating agriculture activities. The hydrological results are compared with the observed flows derived from measurements at the outlet of the Joumine watershed. Model performance showed good statistical agreements, with a Nash-Sutcliffe efficiency of 0.9 and a value of 0.92 after monthly calibration. The model predicted the timing of monthly peak flow values reasonably well. During the validation period, SWAT simulations were nearly as accurate, with Nash-Sutcliffe efficiency and values of 0.89 and 0.92, respectively. The model was used to simulate NO concentrations. The predicted NO concentration values were compared with in situ measured concentrations. The simulated and measured NO-N concentrations varied in the same range of 0 to 5 mg L at the E3 and E5 locations. The calibrated model was then used for simulating the impact of the best management practice scenarios to reduce NO loads to the river. The first set-up consisted of reducing the N fertilizer application by 20 and 100% from the current state. These two scenarios induced a reduction in NO loads by 22 and 72%, respectively. The second set-up consisted of using vegetation filter strips. The last scenario combined filter strips and a reduction of 20% in N fertilizer application. Results showed NO reduction rates of 20 and 36%, respectively. The SWAT model allowed managers to have several options to improve the water quality in the Joumine watershed.