Nitrogen limitation is an important factor for the improvement of crop water production potential in rain-fed areas of the Loess Plateau. The reasonable deep application of nitrogen fertilizer is a promising method to increase yield of rain-fed crop. Based on APSIM model, this study simulated spring wheat yield under different nitrogen application rates and depths, by using meteorological observation data from 1990 to 2020 in the semiarid areas of central Gansu Province, aiming to provide theoretical reference for optimizing wheat fertilization strategy. The results showed that the determination coefficient of simulated spring wheat yield, biomass and soil water content in 0-200 cm soil profile was greater than 0.80, the normalized root mean square error was less than 0.2, and the model validity index was greater than 0.5. These results indicated that the model had good fitting and adaptability in the test area. Across all the levels within the experimental design, increasing nitrogen application rates could significantly increase the yield of spring wheat in different precipitation years, and increasing nitrogen application depth could significantly increase spring wheat yield in wet and normal years, but had no effect in dry years. The rate and depth of nitrogen application had significant interaction effects on spring wheat yield in wet and normal years, but not in dry years. According to the binary quadratic regression fitting equation, when the potential maximum yield reached 2749 kg·hm-2 in wet year, nitrogen application depth was 22.7 cm, and nitrogen application rate was 245 kg·hm-2. When the maximum potential yield reached 2596 kg·hm-2 in normal year, nitrogen application depth was 20.6 cm, and nitrogen application rate was 235 kg·hm-2. Integrating the effects of nitrogen application rate and depth on yield, biomass and agronomic efficiency of nitrogen fertilizer, and farmer's fertilizer application habits, the recommended nitrogen application depth was 20-23 cm, and nitrogen application amount was 120-150 kg·hm-2, which could further improve water productivity and nitrogen use efficiency of spring wheat in arid areas of central Gansu Province.
氮是限制黄土高原旱农区作物水分生产潜力提升的重要因素,而氮肥适度深施是旱地作物提效增产的有效措施。本研究利用甘肃省陇中地区1990—2020年气象观测数据,基于APSIM模型模拟了不同施氮量和施氮深度的春小麦产量,以期为优化陇中旱农区小麦施肥策略提供理论依据。结果表明: 模型模拟的春小麦产量、生物量和生育期0～200 cm土层土壤水分含量的相关性决定系数在0.8以上,归一化均方根误差均小于0.2,模型有效性检验高于0.5,表明APSIM模型在研究区域具有较好的适用性和模拟精度。在试验设计水平范围内,不同降水年型增加施氮量均能显著提高春小麦产量;增加施氮深度,可显著提高丰水年和平水年春小麦产量,对干旱年没有影响;施氮量和施氮深度对丰水年和平水年春小麦产量有显著交互效应,对干旱年没有影响。根据二元二次回归拟合方程可知,丰水年获得潜在最高产量(2749 kg·hm-2)时,施氮深度为22.7 cm、施氮量为245 kg·hm-2;平水年获得潜在最高产量(2596 kg·hm-2)时,施氮深度为20.6 cm、施氮量为235 kg·hm-2。综合考虑春小麦产量、生物量、氮肥农学利用效率和实际习惯施肥量,本研究推荐施氮深度为20～23 cm、施氮量为120～150 kg·hm-2,可进一步提升陇中春小麦水分生产力和氮肥利用效率。.
Keywords: APSIM; nitrogen application depth; nitrogen application rate; nitrogen use efficiency; spring wheat; yield.