Forty nine cotton genotypes were evaluated at Samaru, Northern Guinea Savanna Zone of Nigeria during 2013 cropping season. The experimental design used was 7×7 lattice design with three replications. Plant data recorded on twelve characters were subjected to analysis of variance, Principal component analysis, Cluster analysis, Rank summation index and Correlation analysis. Analysis of variance revealed highly significant differences (p<0.01) for days to 50% flowering, fibre length and fibre fineness, while non-significant differences (P>0.01) were observed in the remaining agronomic traits. The phenotypic coefficients of variations were higher than the corresponding genotypic coefficients of variations for the entire traits studied. Seed cotton yield had the highest genotypic coefficient of variation. The first five axes of principal components analysis having eigen value >1 captured 63.60% of the total variability and jointly identified number of sympodia, number of bolls, ginning outturn, seed cotton yield, boll weight, seed index, days to 50% flowering, fiber length, and lint index as characters contributing most to the total variability. Mean performance revealed moderate variability in all the means of the agronomic traits studied. The variability was highest for seed cotton yield, and by plant height. Means of the fiber quality traits showed considerable variability among the genotypes especially for fiber length, and fiber fineness. Correlation analysis revealed that days to 50% flowering, plant height, sympodia per plant, boll weight, seed index, number of bolls and fiber fineness were highly significant and positively correlated with seed cotton yield, while highly significant negative correlations are observed between number of monopodia, ginning outturn, and fiber length with seed cotton yield. From rank summation index genotypes PIMA S4, RSA (79) 4A, ASA (78) 34A, SAMCOT 12 and SAMCOT 13 are the best genotypes for all the viii traits studied. Cluster analysis grouped 49 cotton genotypes into five major clusters with each cluster having distinct features. The present investigation provided considerable information useful in genetic improvement of cotton. Genotypes PIMA S4, RSA (79) 4A, ASA (78) 34A, SAMCOT 12 and SAMCOT 13 obtained from the USA, Mexico and Nigeria were the best materials for all the traits studied. These genotypes are therefore considered the best potential parents use in for the hybridization programs for the improvement of cotton in Nigeria. Important morphological and fiber quality traits like greater seed cotton yield potential, plant height, fibre length, fibre fineness, number of sympodia per plant and number of bolls per plant etc. served as selection criterion to produce promising cotton genotypes. The low level of genetic diversity observed among the genotypes studied indicate the need for breeders to search for novel and diverse material in order to explore the unutilized genetic diversity for future cotton breeding program of Nigeria. |