Genetic Divergence in Sugarcane Genotypes

Abstract

To assess genetic divergence of sugarcane germplasm, an experiment comprising 25 sugarcane genotypes was conducted at Sugar Crops Research Institute (SCRI), Mardan, Khyber Pakhtunkhwa, Pakistan, in quadruple lattice design during 2008-09. Among the 14 parameters evaluated, majority exhibited significant differences while some showed non-significant mean squares. The initial correlation matrix revealed medium to high correlations. Principal Component Analysis (PCA) showed that there were two principal components accounting for 88% of the total variation in the tested breeding material. The new components were named “Vigor”, and “Quality”. Principal Component Regression (PCR) indicated that these two accounted for 93.64% and 7.36% of variation in the yield, thus signifying the role of the “Vigor” Component. Cluster analysis using Ward’s method on the newly created variables using principal components revealed that there were 3 clusters at a linkage distance of 4.5. Cluster I and III had 11, and cluster II had 3 genotypes. Cluster I showed high mean values for Vigor Component while Cluster II for Quality Component and Cluster III showed genotypes with high mean yield. There was no correspondence of the clustering with the geographic location of the genotypes. It could be concluded from these analyses that there are two main components i.e. vigor, and quality accounting for maximum variation in yield. The genotypes in cluster I and II could be utilized as source for future selection or hybridization program for the improvement of these characters in sugarcane