Multivariate screening among black beans accessions to participation in crossing blocks

Abstract

The multivariate statistical analysis, through measures of dissimilarity, has been helping plant breeders in identifying distant genotypes used to obtain populations segregating desired. Thus, the objective of this study was to evaluate the dissimilarity between 38 accessions of black beans by means of multivariate techniques, and to select parents to compose blocks of crosses. From the Euclidean distance between accessions, grouping by the hierarchical method of Ward was carried out to permit classification of the accessions into different groups, which were then compared using multivariate contrasts. With the use of scores of the first three canonical variables, it was then possible to form six distinct groups that were similar and to corroborate the analysis by clustering methods. The results revealed a wide variability among the accessions studied, where the multivariate analysis was able to add information to the joint range of response variables, and was efficient in determining the difference between such accessions.

Keywords

• Canonical variables
• Phaseolus vulgaris L.
• Techniques of grouping