Farmers are challenged to maintain yield and economic productivity with declining water resources and climatic variability in semi-arid regions worldwide. Growth stage-based deficit irrigation has been suggested as a feasible approach to maintain yields with less water. Experiments were conducted in 2012, 2013, and 2015 in which maize (Zea mays L.) was irrigated under twelve treatments with varied levels of deficit irrigation during the late vegetative (Lveg) and maturation (Mat) growth-stage periods in Northern Colorado. The Root Zone Water Quality Model 2 (RZWQM2)-CERES-Maize model was used to simulate the effects of growth stage-based deficit irrigation on maize production and yield components. The results showed that RZWQM2 could simulate the impact of temperature on maize phenology but did not simulate the impact of water stress on maize maturity. Both simulated and observed aboveground biomass, grain yield, and kernel weight decreased with the decrease of irrigation water amount during Lveg and Mat periods. In general, the simulated aboveground biomass and grain yield showed larger errors in terms of root mean squared error (RMSE), relative RMSE, and Nash–Sutcliffe efficiency, than those reported in the previous modeling studies where deficit irrigation was applied uniformly throughout the growing seasons in the same field. Future efforts to improve the effects of deficit irrigation on kernel development will likely make RZWQM2 a better tool for optimizing irrigation management in semi-arid regions.

在全世界半干旱地区，随着水资源的减少和气候的变化，农民面临着维持产量和经济生产力的挑战。有人提出基于生长期的亏缺灌溉是用较少的水维持产量的可行方法。在2012年，2013年和2015年进行了玉米（Zea maysL.）在北科罗拉多州的植物生长（Lveg）和成熟期（Mat）的后期，采用十二种处理方法进行了不同程度的亏缺灌溉。根区水质模型2（RZWQM2）-CERES-玉米模型用于模拟基于生长期的亏缺灌溉对玉米产量和产量构成的影响。结果表明，RZWQM2可以模拟温度对玉米物候的影响，而不能模拟水分胁迫对玉米成熟度的影响。在Lveg和Mat时期，随着灌溉水量的减少，模拟和观察到的地上生物量，谷物产量和籽粒重量均下降。一般而言，模拟的地上生物量和谷物产量在均方根误差（RMSE），相对RMSE和Nash–Sutcliffe效率方面显示出较大的误差，比以前的建模研究中报道的那些研究（在同一田间的整个生长季节均均匀地应用亏缺灌溉）。未来为改善缺水灌溉对玉米籽粒发育的影响所做的努力可能会使RZWQM2成为优化半干旱地区灌溉管理的更好工具。