Studies On The Models For Prediction Of Main Diseases In Wheat Field In Luohe

This paper analyzed field prevalence trend and factors of main diseases in Wheat in Luohe, predicted the prevalence of diseases based on systematic analysis of the impact elements of main diseases using weighted crosstabs, Fisher's two group discriminant analysis and stepwise regression analysis, and built up the models for forecast.The elements which influence wheat head blight include mainly climate factors, pathogen amount, whole growth period of host plant, varieties and other factors. The paper probed into the traits, occurrence characteristics, prevalence trend and impact factors of this disease, and the results were as follows. Under a certain of rain amount and humidity, the temperature is the important factor which effects on the occurrence and development of wheat head blight. The higher the temperature is, the more severely the head blight occurs, and the bigger the harm is. The number of rainy day and relative humidity are the prerequisites of the head blight occurrence, and the length of consecutive rainy day number after wheat heads is responsible for the degree of wheat head blight occurrence. After heading for 20 or 30 days, the very significantly positive correlation happened between the number of rainy day and relative humidity, and the occurrence rate of the head blight (r=0.76^**and 0.66^**or 0.86^**and 0.81); rain amount has nothing significant to do with the occurrence and development of wheat head blight, and it influences plant growth at the late stage and the resistance to diseases only by the effects on the humidity in wheat field. The negative correlation existed between sunshine duration and the occurrence rate of the head blight, which was influenced by the rainy day number within 30 days after heading. The models for forecast of prevalence degree of wheat head blight were built up by weighted crosstabs. The models were consistent up to 90% with previous situation by validation. The methods can be applied to long-term forecast of prevalence degree of wheat head blight in 50-60 days before optimum time for control and improvement of accuracy of forecast.The results of analysis of prevalence factors of wheat powdery mildew indicated that its pathogen amount had something significant to do with its occurrence degree, namely, when its initial occurrence stage was earlier, and the center was more, the disease was more severely at the year; or vice versa, the disease was more mildly. Besides the pathogen amount, the climate was also the important factor to the disease occurrence. In May, such conditions as lower temperature, much rain water, large relative humidity, and less sunshine, benefit the occurrence and development of the powdery mildew, result in long prevalence period of the disease, and more severe disease index; whereas such conditions as higher temperature, little rain water, small relative humidity, and more sunshine, restrict the occurrence of the disease, caused its earlier ending and more mild occurrence. The model for forecast of the occurrence and prevalence of wheat powdery mildew was built up by Fisher's two group discriminant analysis: y=-0.0017x₁-0.0426x₂-0.0275x₃. The results of validation of previous data showed that the consistency is up to 100%. The model for prevalence structure of wheat powdery mildew was set up by comparison and analysis of 11 years of history data about wheat powdery mildew in the whole Luohe district using three factors of rain amount, rainy day number and average relative humidity from late March to middle April as forecast factors.The results of analysis of field prevalence trend and impact factors of wheat sharp eyespot showed that the occurrence degree of the disease was correlated closely to the climate conditions, and rain amount and relative humidity was positively correlated to diseased plant rate (r=0.9283,r=0.9355), namely, much rain amount, large relative humidity, low temperature and little sunshine caused severe disease; or vice versa, mild disease. Before winter increasing diseased plant rate was the main behavior of the disease development, whereas disease index slowly went up, generally below 5. From regreening to harvest stage, the disease development mainly appeared to be commonly increased diseased plant rate and severity. During the period, from regreening to middle late elongation stage, the former increased quickly; from middle late elongation to grain-filling stage, the latter apparently rose. The results indicated that elongation stage was the main infection period of Rhizoctonia cerealis, covering on estimation 80% of total of diseased plants. The model for forecast of prevalence of wheat sharp eyespot was set up with SPSS statistic software by stepwise regression analysis. The forecast model before winter: y=0.066881+0.334236x₁-0.355563x₂+0.014048x₃; the forecast model after winter: y=-0.00356+0.001553x₁+0.0003987x₂+0.00024499x₃. When given day infection rate (y), thedynamic forecast model at a certain moment was ln(1/(1-X₂)) = ln(1/(1-X₁))×Rs (t₂-t₁)...