干旱胁迫对玉米物候及产量组成的影响及模拟研究

为了研究产量关键期干旱胁迫对玉米物候及产量和产量组成的影响，评估作物生长模型对干旱胁迫下玉米物候和产量模拟的效果，基于锦州农业气象试验站2011-2015年分期播种试验玉米产量和产量组成观测资料，尤其是2014年和2015年天然干旱胁迫试验数据和2015年玉米开花、吐丝物候加密观测资料，分析了产量关键期干旱胁迫对玉米物候及产量和产量组成的影响，评估了CERES-Maize模型对不同降水年型玉米产量和产量组成的模拟效果，提出了模型改进的方向。结果表明，2014年和2015年辽宁省西部地区在玉米开花期前后经历了较严重的干旱胁迫过程，干旱胁迫导致玉米吐丝延迟程度大于开花，90%以上的植株能到达开花阶段，但仅有45%~88%的植株能到达吐丝阶段，直接影响株籽粒数（不同播期下的玉米株籽粒数相差32%）及最终产量（产量下降33%~78%）。CERES-Maize模型对正常年玉米产量及产量组成的模拟效果较好，对干旱年的模拟效果较差，部分原因在于模型在模拟玉米物候时不对开花和吐丝期加以区别，只考虑了温度对花期的影响，而没有考虑干旱胁迫下玉米因开花-吐丝间隔增大、雌穗发育异常、物候期推迟而造成的减产作用。因此，玉米产量关键期干旱胁迫直接影响玉米物候（开花-吐丝期），进而影响玉米穗粒数及最终产量；为提高干旱胁迫下作物模型的模拟评估能力，亟待开展干旱胁迫下基于冠层吐丝动态的玉米产量模拟研究。

Effect of drought stress on maize phenology and yield components and its sim-ulation

Drought is one of the main limiting factors for maize production in Northeast China, and drought stress is particularly severe during the seeding and flowering stages of the crop. The occurrence of drought around anthesis affects the temporal separation of male (anthesis) and female (silking) floral maturity (referred to as the anthesis-silking interval, ASI), which leads to a reduction in maize kernel numbers. To examine the effect of drought stress during the "yield critical period" on maize phenology, yield and yield components as well as to assess the modeling results of phenology and yield simulated by a crop growth model under drought stress, a field experiment was conducted. Sowing was performed on different dates at 10-day intervals beginning from April 20 over a 5-year period (2011-2015). Observation data of maize yield and yield components, particularly phonology and yield data in the drought years of 2014 and 2015 were used to analyze the effect of drought stress on maize phenology and yield components. The results of simulation using the CERES-Maize model were assessed in this study. Further improvements for the CERES-Maize model under drought conditions were proposed. The findings revealed that drought during the yield critical period delayed silking much greater than anthesis. More than 90% of the plants reached the anthesis stage, whereas only 45%-88% of the plants reached the silking stage. Drought in 2014 and 2015 reduced the kernel numbers of maize by 32% and maize yield by 33%-78%. Grain yield was significantly (P<0.001) and positively correlated with the number of kernels per ear. The CERES-Maize model showed good performance (normalized root mean square error (NRMSE) of yield simulation was 6.5% for 2012 and 10.8% for 2013) in the normal years, whereas its performance during the drought years was unsatisfactory (NRMSE of yield simulation was 21.6% for 2014 and 76.5% for 2015), which was attributed partly to the neglect of the longer ASI, failure to reach silking, or delayed phenology, causing a decrease in yield. Above all, drought stress during the yield critical period affected maize phonology (anthesis to silking stage) to some extent, affecting kernel number and grain yield. Thus, it is necessary to study the modeling of maize yield under drought stress based on ASI and silking dynamics by coupling the plant biomass framework.