Perform analysis of variance for significance of different traits and multivarite analysis using Mhalanobis D2 Statistics unitless and scale-invariant, and takes into account the correlations of the data set. It was possible to group the examined genotypes into different clusters.

I worked with screening to remove traits that had no or little importance in the study of genotype dissimilarity (PCA and (SINGH, 1981)) (in press). First, i would do this analysis in both environments to verify the stability of these traits regarding the participation in the clustering of the genotypes.

You can do individual anova for each trait, test for heterogeneity of error variances between environments, calculate pooled mean over environments, then go for calculating Mahalanobis distance or euclidean distance, then follow tocher's method or heirarchical clustering methods respectively