Multivariate analysis for earliness, yield and fibre quality traits in cotton using correlation, PCA and cluster analysis

Abstract

The present study was undertaken to assess the genetic variability, trait associations and genetic divergence among 40 cotton genotypes using correlation, Principal Component Analysis (PCA) and cluster analysis. Analysis of variance revealed significant variability for all the traits studied, indicating ample scope for selection and improvement. Correlation analysis showed that plant height, number of bolls per plant, fibre length and elongation percentage had significant positive associations with seed cotton yield, identifying them as important yield-contributing traits. Earliness traits were positively interrelated, while fibre quality parameters such as fibre length, fibre strength, uniformity index and elongation percentage were positively correlated among themselves and with yield, suggesting the possibility of simultaneous improvement of yield and fibre quality. PCA revealed that seven principal components with eigen values greater than one explained 75.6% of the total variability, with PC-I contributing the highest proportion and largely influenced by yield and fibre quality traits. The bi-plot analysis identified several genetically divergent genotypes positioned away from the origin, indicating their potential as promising parents in hybridization programmes. Cluster analysis grouped the genotypes into five distinct clusters, with Cluster II characterized by superior yield and fibre quality traits, and Cluster III by earliness and higher lint index. The substantial inter-cluster distances suggested the potential for exploiting genetic divergence through hybridization to develop high-yielding, early-maturing and superior fibre quality cotton cultivars. Overall, the combined multivariate approaches provided valuable insights for effective parent selection and genetic improvement in cotton breeding programmes.