Predicting rain-fed wheat yellow rust epidemics in Kermanshah province

During ten-year study (2009-2018), rainfed wheat fields were selected to evaluate disease and examine associations of climatic data with wheat yellow rust disease epidemics across four regions, Sarpolzohab and Gilangharb (tropical), Eslamabad Gharb and Mahidasht (temprate). No disease was evident in 2009-2010, 2016-2017 and 2017-18 and there was low and sparce disease levels in remainder years. Principal component analysis of climatic and disease data indicated that number of rainy days from October to May, periods of consequential days with minimum temperature within 6-9˚C and maximum relative humidity > 60%, and longest period with these climatic characters were the best indicators of yellow rust disease epidemics occurrence across four study regions. Two characters of monthly average of maximum relative humidity for February and minimum temperature of March were also identified as important disease epidemics predictors. Two characteristics, number of icy days and days with minimum temperature under -10˚C were recognized as best indicators to predict disease onset time. To test efficiency of developed stripe-rust-predicting model, lack of or low disease occurrence for two years of 2017 and 2018 were predicted correctly.