基于APSIM模型分析播期和水氮耦合对油葵产量的影响

播期调控和补充灌溉是保障北方农牧交错带油葵稳产和增产的有效措施,然而播期和水氮管理对油葵产量的耦合效应尚不明确。该文基于农牧交错带武川试验站2 a分期播种试验数据评估了APSIM-Sunflower模型的适应性,应用验证后的APSIM模型分析了播期和水氮耦合对油葵产量的影响。研究结果表明:油葵生育期模拟值与实测值均方根误差(RMSE)小于2.4 d,地上部干物质量和产量模拟的归一化均方根误差(NRMSE)分别为21.9%和5.5%,表明APSIM模型能够有效模拟油葵的生育期、地上部干物质量和产量。在有补充灌溉条件时,仅灌一水时在现蕾期补灌油葵产量最高,灌两水时在现蕾和灌浆期补灌产量最高。油葵最佳施氮量随着灌溉量的增加而上升;干旱年无灌溉、灌一水、灌两水和灌三水时最佳施氮量分别为40、60、60和70 kg/hm~2,正常年分别为50、70、80和90 kg/hm~2,湿润年分别为50、80、80和90 kg/hm~2。在湿润年和正常年时雨养、灌一水和灌两水条件下播期在5月中旬较其他播期产量分别高6.9%和11.6%,9.3%和12.0%,9.3%和16.4%,灌一水的产量变异系数分别低41.9%和8.9%;灌两水的产量变异系数分别低38.5%和12.5%;灌三水条件下播期在5月上旬时产量最高。干旱年时早播可降低产量年际间变异,但调控播期对提高产量作用较小。研究结果可为北方农牧交错带油葵生产播期和水氮管理提供参考。

Analysis of interaction of sowing date, irrigation and nitrogen application on yield of oil sunflower based on APSIM model

Oil sunflower is a staple oil-bearing crop with increasing plant area in recent years in the APE (agro-pastoral ecotone of North China). However, the shortage of water resources is a serious threat to oil sunflower production. Adjusting sowing date and applying supplemental irrigation are effective measures to increase the yield and ensure the stability of the yield of oil sunflower. However, the impacts of interaction of sowing date with irrigation and nitrogen fertilization are still unclear. In this study, the suitability of APSIM-Sunflower in the agro-pastoral ecotone of North China was evaluated based on 2 years serial sowing experiments, trial-and-error method was used for model calibration. The validated APSIM-Sunflower model was used to investigate the impacts of the interaction of sowing date with irrigation and nitrogen fertilization on oil sunflower yield. 4 irrigation scenarios were designed including rainfed, 1 irrigation (60 mm), 2 irrigations (120 mm) and 3 irrigations (180 mm). Nitrogen application rate was set between 0 to 120 kg/hm2 at an interval of 10 kg/hm2 under 4 irrigation conditions. The suitable nitrogen application rate was defined as minimum nitrogen application rate when yield change slowly with increasing of nitrogen application rate. 9 sowing dates between 29-Apr to 8-Jun at an interval of 5 days were used to explore the interaction of sowing dates with 4 irrigation scenarios and the suitable nitrogen application rate. The study results showed that the root mean squared error (RMSE) between simulated and observed growth period was less than 2.4 d, and normalized root mean squared error (NRMSE) between simulated and observed above-ground biomass and yields was 21.9% and 5.5% respectively, which suggested that APSIM-Sunflower model performed well in simulating the growth period, above-ground biomass and yield. However, NRMSE of LAI is more than 30%, which suggested that the precision of simulated LAI needs to be improved. Under the condition of 1 irrigation, irrigating at floral initiation produced higher yield than that at floral initiation and start grain-filling. Under the condition of 2 irrigations, irrigating at floral initiation and grain-filling produced higher yield than that at sowing and floral initiation or at sowing and grain-filling. The suitable nitrogen application rate increased with the increasing irrigation. Under the condition of no-irrigation, 1 irrigation, 2 irrigations, and 3 irrigations condition, the suitable nitrogen application rate should be applied by 40, 60, 60 and 70 kg/hm2 respectively at dry years, 50, 70, 80 and 90 kg/hm2 respectively at normal years, and 50, 80, 80 and 90 kg/hm2 respectively at wet years. Comparing yields of different sowing dates under 3 precipitation year types with suitable nitrogen application rate showed that for wet and normal years, sowing at middle May enhanced yield by 6.9% and 11.6%, 9.3% and 12.0%, 9.3% and 16.4% respectively compared to other sowing dates under the condition of no-irrigation, 1 irrigation and 2 irrigations, and decreased variation coefficient of yield by 41.9% and 8.9% under the condition of 1 irrigation, 38.5% and 12.5% under the condition of 2 irrigations. However, the yield of sowing at early May produced higher yield compared to other sowing dates under the condition of 3 irrigations for wet and normal years, and early sowing reduced variation coefficient of yield but had little effect on average yield for dry years. This study provides references for sowing date, irrigation and nitrogen management in the agro-pastoral ecotone of North China.