Abstract

The presence of genetic variation in plant populations is useful for conservation and use in breeding programs. This study was conducted to estimate the extent and patterns of genetic diversity among 200 sorghum accessions collected from different parts of Ethiopia and preserved in a gene bank.  Using 39 polymorphic simple sequence repeat (SSR) markers, which were previously mapped, 261 alleles were produced with mean 6.7 alleles per SSR. Polymorphism Information Content (PIC) and Dice’s similarity coefficient values ranged from 0.06 to 0.81 and from 0.062 to 0.96, respectively. Hierarchical clustering using UPGMA analysis revealed three major clusters with no clear distinction among geographical origins. Moreover, analysis of molecular variance (AMOVA) indicated that the majority of the variation (99.62%) observed was attributed to differences among accessions and only a small fraction of the total variation (0.38%) was related to regions of original collection, which may indicate that geographical origin is not a useful guide to follow for germplasm collection. Rather, agro-ecological classifications may be better for collection mission. Furthermore, crossing of accessions from the three distant clusters could result in promising genotypes for use as varieties or parents for the future breeding programs.

Key words: Allele, cluster, genetic variability, polymorphism information content, sorghum, SSR.