Genetic diversity in a large collection of wheats (Triticum spp.)

Three hundred genotypes of bread and durum wheats including Indian and exotic collections were evaluated for 10 yield and other related characters. Five characters (grain and biological yields, tiller number, grain weight/ear and grains/ear) exhibited high variability. Of the remaining five traits, flag leaf area and 100-grain weight showed moderate variability while plant height, peduncle length and harvest index showed relatively low variability. Following the non-hierarchical euclidean cluster analysis, all the 300 genotypes were grouped into 16 clusters with variable number of genotypes. Genotypes of heterogeneous originl place of release, and of different ploidy levels (bread and durum wheats) often grouped together in the same cluster, suggesting some degree of ancestral relationship between the genotypes. This also suggested a lack of relationship between the genetic diversity and the ploidy level of genotypes. On the basis of the data on genetic divergence and mean performance of yield and other traits, five diverse and superior genotypes, namely MUW 109, CPAN 3064, CPAN 1556, MUW 104, and CPAN 1998 were selected. Each of these genotypes was exceptionally good for one or more characters and was reasonable for other characters relatives to those of the best check varieties. Therefore, these genotypes may be involved in multiple crossing programme to recover transgressive segregates. Further, on the basis of character associations, it is argued that selection of plants with high biological yield coupled with optimum harvest index (50%) should result in progenies with high grain yield potential in wheat.