Study on the possibility of wheat fusarium head blight predicting based on weather and cultural conditions

Fusarium head blight (FHB) is one of the most important diseases of wheat that is a suitable disease for forecasting. In order to determine the statistical relationship among disease occurrence and weather and crop variables, this study was conducted in Golestan, Mazandaran and Ardabil provinces during 3 years (crop years 2016-17, 2017-18 and 2018-19). Crop characteristics and disease severity were recorded in wheat fields around meteorological stations. Nine weather data was received from the stations as daily data and converted to 30 periodic variables. Integrating flowering start date and anthesis length in different years and regions calculated as 11 time and meteorological calendar variables and used in statistical analyses. Results of correlation tests of these variables with disease showed that the best window for disease prediction was B020 (20 days before anthesis) and ATX (average of maximum temperature), RD (number of rainy days), DD (number of dry days), ASHTR (average of special hydrothermal ratio) and NP80 (number of days with rain and relative humidity > 80%) showed the highest correlation with disease. Results of logistic regression and discriminant analysis for the selected variables as one, two and three showed that ATX was the best individual variable, and the best combination of two variables was ATX + NP80 and adding ASHTR to them did not cause significant increase in prediction efficacy. Therefore, logistic regression and discriminant analysis model of ATX and NP80 in 20 days before anthesis, introduced as the best model for forecasting wheat FHB.