摘要
为掌握汉中市小麦条锈病流行规律,提高病害测报的科学性和准确性,采用DPS数据处理系统,对多年来汉中市小麦条锈病测报资料进行了主导因素分析,筛选出影响发病程度的主要因子包括小麦感病品种种植比例(x1)、秋苗病田率(x2)、秋苗单位面积平均病叶数(x3)、1月份平均气温(x6)、上年11月份降雨量(x11)、早春病田率(x18)、3月中旬病田率(x19),与发生程度(y)进行逐步回归分析,建立了发病程度预测模型:y=-6.354 7+0.084 0x1+0.022 8x3+0.662 8x6+0.020 9x11,R2=0.968 7。拟合预测符合率为92.31%,相对误差8.96%。利用预测模型对2014-2016年汉中市小麦条锈病进行预测,预测的病级分别为2.93、3.21、1.92,与2014、2015、2016年发生的实际病级3、3、2相吻合。此模型可应用于生产中小麦条锈病的测报。
In order to verify the epidemic law of wheat stripe rust in Hanzhong city,and to make the crop disease prediction more scientific and accurate,the DPS data processing system was used to ana- lyze the dominant factors of wheat stripe rust in Hanzhong, according to the local records of recent years. As the consequence,the dominant factors affecting the extent of the disease were screened,including the proportion of wheat susceptible varieties, percentage of diseased fields and the average number of susceptible leaves per 667.7 m2 in last autumn,mean air temperature of last January, rainfall of last November, and percentage of diseased fields in early spring and mid-March. Based on the regression analysis of main factors and the occurrence degree,the prediction model of disease severity was established： y=-6. 354 7＋0. 084 0x1 ＋0. 022 8x3 ＋0. 662 8x6＋0. 020 9x11 ,R2 =0. 968 7. Tested by both fitting predictive feedback and actual observation, the historical coincidence rate of this model is 92.31% ,and the relative error is 8.96%. Furthermore, the model was used to conduct the prediction of H anzhong wheat stripe rust during 2014-2016. The predicted results are 2.93,3.21 and 1.92, respectively,which coincide with the actual records of 3,3,2. Thus, the model can be applied into practice.
出处
《麦类作物学报》
CAS
CSCD
北大核心
2017年第12期1640-1644,共5页
Journal of Triticeae Crops
基金
陕西省技术创新引导专项(2017CGZH-HJ-01)
关键词
汉中市
条锈病
流行因子
预测模型
Hanzhong
Stripe rust
Epidemic factor
Prediction model
