THE USE OF GENETIC DISTANCE AND GROUPING METHODS TO PREDICT Eucalyptus pellita F.MUELL GENITORS FOR HYBRIDIZATION

Abstract

The objective of this study was to use quantitative traits to estimate the genetic distance among E. pellita provenances and progenies, to inform possible hybridization strategies in a species improvement program. A provenance and progeny test with 118 progenies from seven provenances was evaluated. The following quantitative traits were measured at seven years of age: diameter at breast height (DBH); height; and individual volume. The data were submitted to REML/BLUP analysis to obtain the predicted genetic value (BLUP). From this, the Mahalanobis (D-2) genetic distance was estimated for provenances and progenies, which were then grouped by Tocher's method, the unweighted pair group method using arithmetic averages (UPGMA), and principal component analysis (PCA). In total, 29 divergent groups were obtained among progenies based on Tocher's method, which showed greater reliability according to the cophenetic correlation coefficient than UPGMA. The opposite was found between provenances, where the results for UPGMA demonstrated greater clustering reliability. Based on principal component analysis (PCA), the M. Ray and Tully provenances were the most similar, while Connl. A and Orchard were the most divergent. Height was the most important trait in estimating genetic distance. The results obtained offer important insights for breeding programs; with this information, crosses can be designed between contrasting individuals among and within provenances to obtain E. pellita hybrids, validating the possible heterotic groups identified through the genetic distance and grouping methods.