Abstract
The current breeding program of sugarcane in Pakistan does not fulfill the variety evolution of desired characters. It mostly
depends on import of exotic fuzz-seed of sugarcane varieties, without keeping varietal characters in view. So, it is the dire need
of the time to characterize the sugarcane germplasm not only for saving resources but also for finding genetic relationships
among breeding materials for overall sugarcane crop improvement. In the present study 60 genotypes of sugarcane, belonged
to 3 research stations two of Pakistan and one from Sri Lanka, were characterized using Principal Component Analysis (PCA)
and Cluster Analysis. At maturity data were analyzed for 19 traits and analyzed. PCA showed seven principal components
(PCs) having eigen value more than one and exhibited 72.1% variability in the genotypes. Four groups of PCA were formed
for the traits. Cluster analysis placed genotypes in seven groups. Groups were not formed according to origins of genotypes
which might be due to resemblance in their progenitors. Each cluster was also marked for specific trait improvement related
to that cluster based upon maximum mean value of that trait, either through selection or hybridization. Four traits were found
better in Sugarcane Research Institute, Faisalabad (SRI, FSD) genotypes, five in Shakarganj Sugarcane Research Institute
(SSRI), Jhang genotypes and three in Sugarcane Research Institute (SRI), Udawalawae, Sri Lankan genotypes, on the basis of
mean values. Thus more emphasis could be given on these traits related to specific station for use in variety selection program
for improvement of target trait. Another interesting conclusion drawn from the study was that clusters were not formed
according to genotypes of a specific research station because these local and exotic sugarcane genotypes are allopolyploid
(with aneuploidy) hybrids, and can be traced back to founder clones (S. officinarum × S. spontaneum )
