Multivariate analysis in parental lines and land races of pearl millet [Pennisetum glaucum (L.) R. Br.]

Abstract

Present investigation was carried out with 31 pearl millet genotypes containing R lines, B lines and land races at Department of Millets, Tamil Nadu Agricultural University, Coimbatore during kharif 2019 in order to assess the genetic diversity and to compare different methods of multivariate analysis. Mahalanobi’s D2 cluster analysis resulted in six clusters with the highest inter cluster distance observed between cluster V and cluster VI. Cluster mean showed that cluster III and cluster I genotypes played significant influence on yield and yield attributing characters. In agglomerative hierarchical cluster (AHC) analysis, the highest inter cluster distance was observed between cluster II and cluster VI. Cluster V and cluster I represented the highest cluster mean for yield and yield component traits. Principal component analysis (PCA) resulted four principal components with eigen values more than one explaining 73.2 per cent variability. The biplot revealed five clusters with cluster I am possessing maximum number of genotypes and positively associated to most of the traits. PT 6706, PT 6709, Nattu Cumbu, Cumbu 2, PT 6676 and PT 6067 were the top-ranking genotypes upon PCA analysis with positive PC1 scores. All three multivariate analyses revealed considerable divergence in the experimental material as well as a comparable type of clustering in the diversity of R line, B line and small seeded land races and hence can be used in future breeding programmes.

Keywords

• mahalanobis’ d2
• agglomerative hierarchical clustering
• principal component analysis
• pearl millet