フロリダ綿花における作付け時期、灌漑、窒素が収量と硝酸塩浸出に及ぼす影響

Decision support system for agrotechnology transfer (DSSAT) provides a robust platform for evaluating long-term management effects. A CROPGRO Cotton (Gossypium hirsutum L.) model (a component of DSSAT) was calibrated using 92 data points and evaluated using 232 data points to assess its performance in simulating phenology, growth, nitrogen (N) content, seed cotton yield (SCY), and nitrogen use efficiency (NUE). The calibrated model was then used to analyze the long-term (2003-2023) effects of three planting windows (May 15-30 [planting date 1, PD1], June 1-15 [PD2], and June 15-30 [PD3]), two irrigation strategies (rainfed and rule-based irrigation), and six N rates (0, 50, 101, 151, 202, and 252 kg ha^-1) on SCY, NUE, and N leaching. The calibrated and evaluated model showed a good agreement between simulated and observed data. Long-term seasonal analysis indicated that median SCY in PD1 was 21% and 46% higher than in PD2 and PD3, respectively. Delayed planting reduced NUE by 37% and increased N leaching by 85%. Irrigated cotton resulted in 38% and 29% higher SCY and NUE, respectively, than rainfed cotton, without significantly increasing N leaching. NUE declined by 19%-33% with each N rate increment from 0 to 252 kg N ha^-1, while N leaching increased by 45% at 252 kg N ha^-1 compared to 0 kg N ha^-1. However, leaching increased significantly only when N application exceeded 151 kg N ha^-1. These findings emphasize the importance of adjusting N rates based on PD and irrigation strategy to maximize yield and NUE while minimizing leaching losses.