Classification of rice genotypes in Northern Iran using multivariate statistical analysis

Increasing population and limited arable lands has perpetuated the efforts to improve yield per unit area through various methods such as sophisticated agronomical operations, selection of productive and tolerant genotypes and improving germination components. In order to produce desirable and highly adaptable varieties, it is critically important to maintain a good understanding on genetic variations and selection of advanced genotypes. In the present research, 15 landraces of rice were investigated as randomized complete blocks design, with three replications. The study genotypes varied significantly in terms of the most of traits. Cluster analysis was done based on the significant traits from analysis of variance conducted on standardized data using WARD method. This analysis classified the genotypes into 2 groups, which 11 of them including in two groups. Varieties in this group had the highest mean values for all the evaluated traits and so were designated as the best group. Analysis into factors showed that three of the factors accounted for 90.66% of total variations, while the first one was designated as efficient factor among grain germination components by accounting for 40.797% of the variation.