The objective of this work was to evaluate the clustering pattern consistency in common bean cultivars (Phaseolus vulgaris L.) obtained from the combination of the eight dissimilarity measures and eight clustering methods. Thirteen common bean cultivars were evaluated in nine experiments conducted in the growing seasons between 2000/2001 and 2004/2005 and grouped according to productive (grain yield and number of pods per plant), phenological (number of days between seedling emergence and flowering, and number of days from emergence to harvest) and morphological (height of first pod insertion and height of the final pod insertion) characters. Analyses of variance, correlation, diagnoses of multicollinearity, cluster and comparison of means were carried out. Clustering pattern consistency was evaluated by the cophenetic correlation coefficient. There is variability in the clustering pattern consistency of common bean cultivars, derived from the combination of different dissimilarity measures and clustering methods. Greater clustering pattern consistency in common bean is found with the average linkage between groups method, based on the euclidian distance.
Phaseolus vulgaris; clustering algorithms; cophenetic correlation coefficient; dissimilarity measures
