Water scarcity in arid and semiarid regions poses problems for agricultural systems, awakening special interest in the development of deficit irrigation strategies to improve water conservation. Toward this purpose, farmers and technicians must monitor soil water and soluble nutrient contents in real time using simple, rapid and economical techniques through time and space. Thus, this study aimed to achieve the following: (i) create a model that predicts water and soluble nutrient contents in soil profiles using electrical resistivity tomography (ERT); and (ii) apply the model to different woody crops under different irrigation regimes (full irrigation and regulated deficit irrigation (RDI)) to assess the efficiency of the model. Simple nonlinear regression analysis was carried out on water content and on different ion contents using electrical resistivity data as the dependent variable. A predictive model for soil water content was calibrated and validated with the datasets based on exponential decay of a three-parameter equation. Nonetheless, no accurate model was achieved to predict any soluble nutrient. Electrical resistivity images were replaced by soil water images after application of the predictive model for all studied crops. They showed that under RDI situations, soil profiles became drier at depth while plant roots seemed to uptake more water, contributing to reductions in soil water content by the creation of desiccation bulbs. Therefore, the use of ERT combined with application of the validated predictive model could be a sustainable strategy to monitor soil water evolution in soil profiles under irrigated fields, facilitating land irrigation management.
Keywords: electrical resistivity; nonlinear regression analysis; regulated deficit irrigation; soil moisture; tomography.