【摘要】 根据河南、河北两省多年多点对玉米小斑病流行进程的观察记载,共积累20个年次的病情系统调查资料,结合相应的气象数据,按照本病的流行规律,设计8个供选因子,用北京大学DJS-18型电子计算机进行逐步回归运算,分别推导出从7月中旬至8月中旬的30天中期预测式和从7月中旬至9月中旬的60天长期预测式。所谓测报时间范围内相应时期的气象因素,在中期预报中为7月份的总雨量、总雨日、日平均温度等,在长期预报中为7、8两个月的上述气象因素。两个预报式的确定系数分别为0.8125和0.6843,经用建立模型以外的多个实际病情对所得方程式进行检验,证明预测式有较高的可靠度。长期预报式可用做流行强度的趋势性预报,中期预报式可供确定防治适期的参考。更多还原

【Abstract】 Systematic data pertaining to the study of corn leaf blight have been accumulated during the pastyears from many counties of Henan and Hebei provinces. The meteorological items of correspondingperiod of predictions were also listed as primary data. According to the principles of epidemiology ofcorn leaf blight, 8-related factors were designed for selection. DJS-18 typed computor were used forstepwise regression calculation. As a result, 4-factors were selected for both long-period （60 days）and mid-Period （30 days）Predictions. The regression equations are as follows: The mid-period predect equation: The long-period predict equation: Where:L_n X₁/1-X₁=The logistic value of disease index of the initial date.L_n X₂/1-X₂= The logistic value of disease index of the middledate.L_n X₃/1-X₃= The logistic value of disease index of the final date.■=The average daily temperature of the corresponding Period of prediction.ΣR_Q=The total quantity of rainfall of the corresponding period of prediction.ΣR_D=The total number of rainy days of the corresponding period of Prediction.ΣR_Q T=The interaction of the factors of rainfall and the average daily temperature during thecorresponding period of prediction.ΣR_D·■=The interaction of the factors of rainy days and the average daily temperature during the corresponding Period of prediction. The regression equations were examined with the actual final disease index used other than in theestablishment of models, a better fittingness were observed between the actual and theoretical values.Through x²-test, the above conclusion was also obtained. The authors believe that if these regression equations for predicting the, amount of disease wereused together with the model of forecasting yield-loss, it would be benefitial to make decision indisease management, such as fungicide application, rate-limiting resistance and various cultivatingpractices.更多还原