Investigation of genetic diversity in alfalfa ecotypes collection using multivariate analysis based on morphological traits

In order to study genetic diversity in alfalfa ecotypes collection, 140 ecotypes collected from various regions of Iran and the world were investigated during 2006-2007. Mean of the studied traits calculated for both growing seasons was used for analysis. Simple correlation coefficients between traits, multiple regression, stepwise regression and principal component analysis (PCA) were carried out. Dendrogram of alfalfa ecotypes was created based on Euclidean distances. Traits correlation analysis revealed that weight of herb, number of stem in the herb and plant height had significant positive correlations with green matter yield. The stepwise regression analysis showed that only weight of herb among the other independent variables had a significant positive effect on green matter yield. Alfalfa ecotypes were grouped into 8 clusters. The PCA analysis showed that the first three principal components explained 78.2% of the total variability among the alfalfa ecotypes. Cluster analysis in this study did not separate Iranian ecotypes from foreign ecotypes. It is probably related to traits which did not show high variability in the studied ecotypes. Almost Iranian and French ecotypes were placed in one cluster. Alfalfa is believed to belong to Caucasus region: northeastern Turkey, Turkmenistan and north western Iran; thus, it can be supposed that the French ecotypes have been introduced from Iran and adapted to the climate conditions of France.