四川盆地冬繁区小麦条锈病气象等级预测模型

四川盆地冬繁区是常年受小麦条锈病危害最重的地区之一.本研究利用盆地冬繁区代表站点1999—2016年的气象资料和条锈病资料,根据发病面积比将小麦锈病发生的气象条件划分为5个等级,采用多种分析方法确定了影响小麦条锈病发生的具有明确生物学意义的气象因子,并建立了小麦条锈病气象等级预测模型.结果表明: 四川盆地小麦条锈病的发生与平均(最高、最低)气温、降水量及距平百分率、相对湿度及距平百分率、平均风速、日照时数等多种气象因子显著相关,其中,平均气温和相对湿度距平百分率起主导作用.历史回代检验中,区(县)级样本准确率64%,市级样本准确率89%.对2017年盆地冬繁区小麦锈病发生气象等级进行预报,预测结果完全正确的样本占总样本量的62.8%;误差1个等级的样本占27.9%,误差2个或2个以上等级的样本仅占9.3%,预测效果较好,能达到从气象角度对小麦锈病发生进行预报的目的.

Prediction model of meteorological grade of wheat stripe rust in winter-reproductive area,Sichuan Basin,China

The winter reproductive areas of Puccinia striiformis var. striiformis in Sichuan Basin are often the places mostly affected by wheat stripe rust. With data on the meteorological condition and stripe rust situation at typical stations in the winter reproductive area in Sichuan Basin from 1999 to 2016, this paper classified the meteorological conditions inducing wheat stripe rust into 5 grades, based on the incidence area ratio of the disease. The meteorological factors which were biologically related to wheat stripe rust were determined through multiple analytical methods, and a meteorological grade model for forecasting wheat stripe rust was created. The result showed that wheat stripe rust in Sichuan Basin was significantly correlated with many meteorological factors, such as the ave-rage (maximum and minimum) temperature, precipitation and its anomaly percentage, relative humidity and its anomaly percentage, average wind speed and sunshine duration. Among these, the average temperature and the anomaly percentage of relative humidity were the determining factors. According to a historical retrospective test, the accuracy of the forecast based on the model was 64% for samples in the county-level test, and 89% for samples in the municipal-level test. In a meteorological grade forecast of wheat stripe rust in the winter reproductive areas in Sichuan Basin in 2017, the prediction was accurate for 62.8% of the samples, with 27.9% error by one grade and only 9.3% error by two or more grades. As a result, the model could deliver satisfactory forecast results, and predicate future wheat stripe rust from a meteorological point of view.