Modeling Soil–Water–Disease Interactions of Flood‐Irrigated Mandarin Orange Trees: Role of Root Distribution Parameters

Abstract

Core Ideas Soil–water–disease interactions were analyzed using field and numerical studies. ERT was able to image the soil moisture front separated by active roots. HYDRUS overpredicted RWU from diseased trees, requiring a compensation mechanism. Parameters defining the root distribution were optimized considering plant health. Variably saturated flow models that simulate soil–water–plant interactions within the rhizosphere largely ignore the effect of plant health. This makes it difficult for them to effectively simulate root water uptake (RWU) and implement alternate management practices for diseased trees. This research was aimed at understanding the hydrological and plant controls on RWU simulations dominated by the health of the tree using experimental and numerical frameworks. Two research plots, one around a healthy mature and the other around a diseased mature (Phytophthora spp. affected) orange tree (Citrus reticulate Blanco), in the Vidarbha region of India were considered for our analysis. Three‐dimensional electrical resistivity tomography (ERT) performed at the two locations revealed that the soil moisture profiles following irrigation are different between the two plots. A two‐dimensional axisymmetric form of Richards' equation was then solved using HYDRUS (2D/3D) by incorporating a root distribution function. A global sensitivity analysis was performed to identify the root distribution parameters that influence soil moisture simulations. These parameters were then optimized using a genetic algorithm for healthy and diseased conditions. We observed that the diseased orange tree had consumed less water, leaving high soil moisture in the rhizosphere, a condition favorable for the further growth of disease‐causing fungi. Hence, managing irrigation water in accordance with RWU patterns is essential for diseased trees. To understand the error propagation in RWU estimation, we simulated RWU from a diseased tree with the optimal parameters derived for a healthy case. Results show that the error in estimating RWU from the model cells progressively increases with radial distance and soil depth.