Multiple Regression Accounting for Wheat Yield Reduction by Septoria nodorum and Other Pathogens

Abstract

During 1973-74 and 1974-75 severe natural infection of wheat [Triticum aestivum] by S. nodorum occurred in Georgia [USA]. Numerous cultivars were evaluated for disease severity and the effect of the pathogen on various yield components. A Septoria disease index (SDI) was derived from these data. The SDI correlated with yield, 1000-kernel weight, test weight and plant height. Multiple regression equations were utilized to predict yield from disease severity caused by S. nodorum, Erysiphe graminis f. sp. tritici, Puccinia recondita f. sp. triticina, and several agronomic factors. When 1000-kernel wt was included in an equation with SDI and E. graminis, 81% of the yield variation could be explained. Correlation coefficients involving SDI and yield were improved when SDI was squared and/or divided by each cultivar''s mean plant height.