The soil water balance of rainfed and irrigated oats, Italian rye grass and rye using the CropSyst model

Crop growth models have been used in simulating the soil water balance for purposes of irrigation management and yield predictions. The application of CropSyst, a cropping systems simulation model, was evaluated for Cedara, South Africa. Simulations included soil water balance of fallow land and rainfed and irrigated winter crops [oats (Avena sativa), Italian ryegrass (Lolium multiflorum) and rye (Secale cereale)]; and irrigation scheduling of the winter crops. Soil, plant, weather and management inputs were used for the soil water balance simulations. Model crop parameters were used from past experiments or obtained from model documentation, with a slight modification to account for varietal differences. The fallow land soil water simulations were more accurate for dry than for wet soil. For all three winter crops, the model consistently over-estimated the soil water content in the upper layers, with a good agreement for the deeper layers until a large precipitation event occurred to which the model responded more slowly than that observed. Simulations using model-scheduled irrigation based on 0.4 and 0.6 maximum allowable depletion criteria indicated that the observed applied irrigation in the field was more than that required. Soil water depletion and accumulated transpiration simulations were similar in both the observed and model-scheduled irrigations, but total soil evaporation and percolation were greater in the case of the observed than the model-scheduled irrigations. Irrigation scheduling using crop growth models may assist in avoiding over- or under-application of irrigation applications by ensuring efficient utilization of rain and irrigation.