Epsilon Open Archive

Abstract

Weather factors are driving forces in plant disease development and differ between years and locations. Results from long-term fungicide field trials 1983-2007 and disease surveys 1988-2007 in winter wheat in southern Sweden were used to evaluate relationships between yield, the yield increase obtained by fungicide treatment, thousand grain weight (TGW), disease severity and disease incidence, and the independent variables air temperature and precipitation as monthly means. These two weather variables explained more than 50% of the variation between years regarding yield increase, TGW, LBDs (Leaf Blotch Diseases, including Septoria tritici blotch, Stagonospora nodorum blotch and tan spot), brown rust, yellow rust and eyespot, but less than 50% of the variation in yield and powdery mildew. Precipitation in May was the factor most consistently related to LBD disease intensity, and adding another two weather factors further improved the degree of explanation. Weather factors in the preceding growing season influenced growth stage, powdery mildew and brown rust. Mild winters and springs favoured the biotrophs such as powdery mildew, brown rust and yellow rust. Statistically significant correlations between incidence and severity were found for LBDs, brown rust and eyespot, but not for yellow rust and powdery mildew. Regression models with disease incidence as dependent variable generally had a higher degree of explanation and lower P-value than models with disease severity as dependent variable. Our results confirm that weather data can be successfully used in wheat disease prediction models.