A GENETIC APPROACH TO THE MULTIVARIATE DIFFERENTIATION OF PERENNIAL RYEGRASS (LOLIUM-PERENNE L) POPULATIONS

Genetic differentiation among perennial ryegrass populations was studied using multivariate procedures incorporating a broad range of seasonal growth, quality and persistency traits. Principal components analysis, based on a genetic variance/covariance matrix, resulted in a correlation of 0.70 between the relative weighting (communalities) of traits in the first four principal components and the corresponding between population heritabilities. This was compared to correlations of 0.43 and -0.82 obtained for principal components, based on phenotypic and environmental variance/ covariance matrices, respectively. Principal component communalities were more evenly related to heritabilities than those from canonical variates analysis which produced a correlation of 0.5 with heritabilities. Cluster analyses, based on principal component scores, produced four groups of populations separated by Mahalanobis distances ranging from 3.0 to 7.8. Considerable heterosis was obtained in crosses between populations from the more widely genetically separate groups.