The objective of this work was to evaluate the adjustment of the yield for the stand in common bean experiments and to propose a procedure for the adjustment through covariance analysis. Data from 33 environments of the Common Bean Assay of Rio Grande do Sul State, Brazil, were used. Variance and covariance analyses of yield and stand were proceeded for each environment. Four alternative models were considered: without adjustment for stand; with adjustment for the average stand; with adjustment for the genotypes average stand; and with adjustment for the groups of genotypes average stand. Significant linear or quadratic effects of stand on yield occurred in 70% of the environments and of genotype on stand in 85%. The compensatory effect was manifested with different intensities in the environments. The adjustment of the yield for stand, considering the effect of genotypes on stand, is important in common bean experiments. The record of information to discriminate the origins of the variation of stand is necessary.
Phaseolus vulgaris; covariance analysis; experimental precision; validity of the experiment
