干旱区生态环境知识资源中心

Irrigation in semi-arid areas can be optimally scheduled using models that maximize accuracy while minimizing data requirements. In this work, the validation of the one-dimensional transient-state SALTIR-SOIL_M model able to monthly simulate the soil water content (theta), downward water flux (D), and the electrical conductivity in the saturation extract (ECe) is presented. The theta and the ECe were monthly monitored at 15 and 45 cm depth in a plot in commercial production of globe artichoke under semi-arid climate and drip-irrigation in SE Spain during the 2011-2012 growing season. Data on water quality, weather, irrigation management, crop development and soil were also collected and used to simulate the theta, D, and ECe throughout the growing season at both depths with the SALTIRSOIL_M model. Reference and simulations of theta and D reasonably agreed with one another at both depths. ECe at 45 cm depth was estimated correctly enough by the model but was underestimated at 15 cm depth. The higher concentrations of nitrate, potassium, and hence other cations, which were observed at 15 cm as a consequence of fertilization could explain this. The model was subsequently used to estimate the optimum water management in the plot. Use of 325-450 mm yr(-1) instead of 694 mm yr(-1), and by means of evenly distributed pulse irrigations instead of continuous ones, would have met the crop evapotranspiration requirements while avoiding excessive drainage, and maintaining soil salinity well below damaging values. (C) 2014 Elsevier B.V. All rights reserved.