Estimates of repeatability in the evaluation of resistance of soybean genotypes to powdery mildew

Abstract

Soybean is worldwide a crop of great importance, however, with its expansion, diseases causing great yield loss such as the powdery mildew surge (Erisyphe diffusa U. Braun & S. Takam). Objective of the present study was to estimate the ideal number of evaluations of the resistance of soybean genotypes to powdery mildew by means of repeatability studies with the traits: ILA (infection of the leaf area), ILS (infection of the leaf blade surface) and ILU (infection of the leaf underside), using the treatments no control – NC, partial control – PC (fungicide 80% sulphur at ILA > 40%) and total control – TC (presence of symptoms). Only trait ILA presented estimates of repeatability (ANOVA and principal components), with 85- 90% of stability in 3 or 4 measurements (PC and NC methods). These results and the high values found for R2 indicate that selection based on this parameter is possible.