Use of crop simulation models to evaluate limited irrigation management options for corn in a semiarid environment

Increasing competition for land and water resources due to increasing demands from rapid population growth calls for increasing water use efficiency of irrigated crops. It is important to develop location-specific agronomic practices to maximize water use efficiency (WUE). Adequately calibrated and validated agricultural systems models provide a systems approach and a fast alternative method for developing and evaluating agronomic practices that can utilize technological advances in limited irrigation agriculture. The objectives of this study were to (1) calibrate and validate the CERES-maize model under both dryland and irrigated corn (Zea mays L.) production in northeastern Colorado and (2) use the model with a long-term weather record to determine (1) optimum allocation of limited irrigation between vegetative and reproductive growth stages and (2) optimum soil water depletion level for initiating limited irrigation. The soil series was a Rago silt loam, and the initial water content on 1 January of each year was equal to field capacity in the upper 300 mm and half of the field capacity below this depth. Optimum production and WUE with minimum nitrogen (N) losses were found when (1) a water allocation ratio of 40: 60 or 50: 50 (uniform) between vegetative and reproductive stages for irrigations up to 100 mm, and a ratio of 20: 80 for irrigations above 100 mm was used; and (2) irrigation was initiated at 20% plant-available water (PAW) (80% depletion). When available water for irrigation is limited to 100 mm, irrigating 50% of the area with 200 mm of water at 20:80 split irrigations between the vegetative and reproductive stages produced greater yield than irrigating 100% of the area with 100 mm water. Concepts developed in the study can potentially be adapted to other locations, climates, and crops. However, precise site-specific recommendations need to be developed for each soil-climate zone using the validated system model.