Disease Detection and Losses
Assessment of Methods of Determining Powdery Mildew Severity in Relation to Grain Yield of Winter Wheat Cultivars in Ohio. P. E. Lipps, Associate professor, Department of Plant Pathology, The Ohio State University, Ohio Agricultural Research and Development Center, Wooster, 44691; L. V. Madden, Associate professor, Department of Plant Pathology, The Ohio State University, Ohio Agricultural Research and Development Center, Wooster, 44691. Phytopathology 79:462-470. Accepted for publication 31 October 1988. Copyright 1989 The American Phytopathological Society. DOI: 10.1094/Phyto-79-462.
Powdery mildew severity was determined by estimating the percentage of leaf area covered by lesions on the upper three leaves of three winter wheat cultivars in 1985 and six cultivars in 1986 and 1987. Data obtained were used to evaluate five different disease assessment systems for estimating powdery mildew severity: a three-leaf additive system, a three-leaf weighted system, a two-leaf additive system, a two-leaf weighted system, and a 0–10 scale. The 0–10 scale was designed to account for leaf position and the percentage of leaf area covered by lesions, where 0 represents trace or no lesions on any leaf, and 10 represents lesions covering more than 15% of the flag leaf area. Disease severity evaluations using the 0–10 scale were highly correlated with those using the other four systems; e.g., r was 0.81–0.97 at growth stage (GS) 10.3. Linear regression analysis was used to determine the relationship between grain yield and powdery mildew severity. Depending on the growth stage, assessments of disease severity based on all assessment systems were significantly related to grain yield, although the coefficient of determination (R2) varied with cultivar (e.g., at GS 10.3 in 1986, the highest R2 was 0.87 to 0.90 for Becker, and the lowest was 0.15 to 0.32 for Caldwell), indicating that the assessment systems were nearly equal for assessing powdery mildew severity. Over the three years, yield was rarely correlated with disease severity before GS 10 (P < 0.05); disease at GS 10.3 was most consistently correlated with yield for all cultivars and years (R2 values were 0.16 to 0.87). Further regression results were thus based on the disease severity from the 0–10 scale at GS 10.3 and the area under the disease progress curve (AUDPC) based on the 0–10 scale. Slopes and intercepts varied among cultivars within and across years, but the more susceptible cultivars had higher slope values each year (Adena, Becker, and Hart were susceptible; Caldwell, Cardinal, Scotty, and Tyler were less susceptible), indicating greater yield reduction per increase in disease severity. R2 values for regression equations calculated for five of the six cultivars studied in 1986 and 1987 were relatively high (0.66 to 0.87 in 1986 and 0.50 to 0.83 in 1987). R2 values for regression equations calculated from AUDPC data were marginally higher than those from the single assessment at GS 10.3 for some cultivars, but for others R2 values from AUDPC were about the same as those from the GS 10.3 assessment or slightly lower. This was due to the high correlation between disease severity at GS 10.3 and AUDPC (r was 0.87–0.98). Covariance analysis indicated that fungicide treatment did not alter the relationship between yield and disease for any year or cultivar. Results indicated that linear regression equations calculated from disease assessments taken at GS 10.3 using the 0–10 scale adequately predicted the yield of the cultivars studied.
Additional keywords: epidemiology, Erysiphe graminis f. sp. tritici, Triticum aestivum, yield loss assessment.