Determining diversity and genetic distance of 31 genotypes of Agropyron desertorum for grain yield and its components using multi-variate statistical methods.

In order to investigate diversity and explain the relations of grain yield with its components, 31 genotypes of Agropyron desertorum were densely planted in randomized complete blocks design with four replications in climatic conditions of Arak city (Markazi Province, Iran). Some of the most important morphological and physiological attributes were measured. The data collected in two years were analysed; principal components analysis and cluster analysis were used to determine genetic distance and genotype classification. In principal components analysis, the specific values obtained from 1 to 5 components were larger than one. Totally, 75 percents of all variances justified the variables. In the first component, attributes of ear emergence time, pollination time, plant height, and forage yield, and in the second component, attributes including number of stems, grain yield, and harvest index had greater specific vector coefficients besides major attributes for genotype classification. In cluster analysis, genotypes were categorized in seven groups. According to the obtained results, clusters 4, 6 and 7 - in term of grain yield - and clusters 5, 6 and 7 - in terms of forage production - were significantly different from other groups. Genotypes 3477P4 and 341P11 belonging to cluster 7 were known as superior genotypes in terms of forage production and two genotypes 341M and 287P8 belonging to cluster 4 were proved to be the best genotypes in terms of grain production. Good agreement was observed between the results obtained from cluster and principal components analyses in scatter diagram representation of 31 genotypes based on two principal components (first and second) using fourteen studied attributes.