ABSTRACT:
The objective of this research was to evaluate the phenotypic variance components and estimate the influence of interaction genotype*environment in grain yield of beans. The variance components were estimated by the method restricted maximum likelihood and best linear unbiased prediction (REML / BLUP), along with the specific space of inference. Data were collected during the crop seasons 2006/07 to 2011/12 in Lages, SC - Brazil. During this period, 104 genotypes were evaluated. The data are unbalanced and 13 genotypes remained presents in all the assays years. Observing the results it was possible to see that the large variation (59.0%) in the performance of genotypes over the years is primarily attributable to the variance of the environment (σ2a=436245). There were significant differences among genotypes for all environments, but the difference between them were constant, ie, the genotypes did not respond differently in environments. Genotype*environment responds primarily by very small changes (0.2%) in the phenotypic variation (σ2ga=1368), with no parameters to discriminate the grain yield of genotypes in Lages, SC - Brazil. This fact makes easier the selection in plant breeding programs, where interaction genotype*environment could hinder distinction and selection of superior genotypic constitutions. In this situation, the processes of recommendation from improved lines (assay value for cultivation and use - VCU) that maintain the same genotypes dispense successive evaluations, since the ranking is parallel over the years.
Key words:
effect of the environment; plant improvement; linear mixed models; Phaseolus vulgaris L.; REML / BLUP
Thumbnail
