Multivariate analysis from SSR and morphological data in chickpea (Cicer arietinum L.) for breeding purposes

Abstract

In order to enhance genetic potential of chickpea materials from the National University ofCórdoba Breeding Programme and Germplasm collection (Argentina), a study for a comprehensiveunderstanding of the amount and pattern of genetic variation within and betweengenotypes was carried out by applying a multivariate analysis form single simple repeats(SSR) and morphological data. Molecular data were also used to determine the discriminatingpower for genotype identification, and to find the optimal primer combination to ensureunambiguous identification. With the analysis of 15 SSR markers on 53 genotypes, a totalof 58 alleles were detected with individual values ranging from one to nine alleles perlocus. High values of discriminating power (Dj ⩾ 0.7, PIC ⩾ 0.7), and low values of confusionprobability (Cj ⩽ 0.23) were obtained for at least four evaluated markers. The combination ofTA113 + TA114 + H1B09 + TA106 primers was effective for discriminating the 53 chickpeagenotypes with a cumulative confusion probability value (Ck) of 9.60 × 10−4. Except forsome exceptions, individual chickpea genotypes within a cluster in the consensus tree weredefinitely more closely related with each other by the origin or pedigree. The results confirmedthat both multivariate data analysis methods, ordination and clustering, were complementary.In most genotypes, discriminant principal component analysis classification was consistentwith the original clusters defined by molecular data. Differences in results from molecularand morphological data indicate that they provide complementary and relevant informationfor establishing genetic relationships among chickpea materials and a better description andinterpretation of the available variability in the germplasm collection.