Vol: 34 Page: 251

Authors:
Wiesław Mądry
Paweł Krajewski
Leszek Sieczko

Title:
 An application of canonical variate analysis for the multivariate assessment of combining ability in blackcurrant (RIBES NIGRUM L.) variates

Language: English

Keywords:
canonical variate analysis (CVA)
MANOVA model
Griffings's model
Mahalanobis's distance
black currant
diallel cross design
combining ability

Summary:
The aim of the paper was to characterize multivariate GCA effects of five blackcurrant (Ribes nigrum L.) cultivars, estimated in a diallel cross design. The GCA effects were estimated for fruit yield, yield-contributing and resistance characters. Genetic diversity of the parents, as revealed using multivariate GCA effects, was evaluated. Canonical variate analysis (CVA) based on a MANOVA model for 25 offspring (F1 single-cross hybrids) that resulted from a five-parent full-diallel cross (Griffing’s method 1) was used. F1 plants were tested in a randomized complete-block design with four replicates (plots of 15 plants). It was shown that the multivariate characterization of breeding value of blackcurrant parents (as based on GCA effects for fruit yield and important yield-related traits) was much more proper criterion of identifying parental genotypes that could produce superior hybrids than genetic GCA-based distances between parental genotypes. Canonical variate analysis was a useful statistical tool for clear identifying multivariate genetic variation of parents contributed in a diallel cross design and the most influential traits affecting genetic diversity of these parents. This procedure was also effective for multivariate characterization of parent’s GCA effects as a attractive measure of their complex breeding value.