Bread wheat (Triticum aestivum L.) germplasm consisting of 45 genotypes were clustered phenotypically using ten morphological traits and Area Under Disease Progress Curve (AUDPC) as measure of stripe rust resistance. The clustering was ratified by using twenty three molecular markers (SSR, EST and STS) linked to stripe rust (Puccinia striiformis f. sp. tritici) resistant QTLs. The aim was to asses the extent of genetic variability among the genotypes in order to select the parents for crossing between the resistant and susceptible genotypes with respect to stripe rust. The Euclidian dissimilarity values resulted from phenotypic data regarding morphological traits and AUDPC were used to construct a dendrogram for clustering the accessions. Using unweighted pair group method with arithmetic means, another dendrogram resulted from the similarity coefficient values was used to distinguish the genotypes with respect to stripe rust. Clustering based on phenotypic data produced two major groups and five clusters (with Euclidian dissimilarity ranging from 2.44 to 16.16) whereas genotypic data yielded two major groups and four clusters (with percent similarity coefficient values ranging from 0.1 to 46.0) to separate the gene pool into highly resistant, resistant, moderately resistant, moderately susceptible and susceptible genotypes. With few exceptions, the outcome of both type of clustering was almost similar and resistant as well as susceptible genotypes came in the same clusters of molecular genotyping as yielded by phenotypic clustering. As a result seven genotypes (Bakhtawar92, Frontana, Saleem 2000, Tatara, Inqilab91, Fakhre Sarhad and Karwan) of diverse genetic background were selected for pyramiding stripe rust resistant genes as well as some other agronomic traits after hybridization.
Keywords: Stripe rust, genetic diversity, molecular markers, selection. Triticum aestivum