Study Of The Genetic Variation Within Iranian Local Bread Wheat Lines Using Multivariate Techniques

Authors

Abstract

Genetic and geographical variation in 600 Iranian local bread wheat lines were studied in Agricultural College at Karaj in 1998/99 cropping season. A systematic design and multivariate techniques were used. To study morpho-thecnological traits. Principal component analysis reduced 22 variables down to eight components with 80.5% proportional cumulative variance. Factor analysis using maximum likelihood estimation method with varimax rotation was performed on 21 variables and reduced them down to six common factors with 63.6% proportional cumulative variance which were economic yield production and partitioning (17.4%), dry matter production (15.2%), tillering (10.3%), productivity (8.2%), height-related factors (6.9%) and preheating phase length (5.6%). Common and specific variances of revealed that proportional common variances of grain yield, 100-kernel weight, No. of kernels/ Spike and No. of fertile tillers/ plant were 98%, 93%, 93% and 84%, respectively. According to the results, consideration of the 6-common factor model is very suitable and determinable for yield and yield components. With cluster analysis using UPGMA method and based on 65% within group similarity, locations were classified into six clusters. Cluster analysis using WARD method classified the 600 landraces into six clusters. One-way ANOV A on all traits was performed using clusters membership as subscript to prove 6-cluster model statistically. The results showed that these 6 clusters had significant differences (?= 0.001). Genetic variability did not follow the same pattern as geographical variability.

Keywords