Determination of genetic distance among 29 genotypes of perennial ryegrass (Lolium perenne) using cluster analysis of yield and morphological traits

In order to study the genetic variation among local and foreign accessions/ varieties of perennialryegrass (Lolium perenne) in Gene Bank of Natural Resources, an experiment was conducted in Alborz Research Center, Karaj, Iran. Seedlings of 29 genotypes were transplanted in the field using a complete block design with three replications. Each row plot was consisted of five spaced plants. Forage dry matter yield, basal cover, ear emergence date, pollination date, stem number, stem height and persistency at four harvesting dates were recorded. Data was analyzed using analysis of variance, principal components and cluster analysis.
Analysis of variance showed significant differences among accessions/ varieties for all of the characters in each cutting, indicating the existence of genetic variation in (Lolium perenne) germplasm available in the Gene Bank of Natural Resources. Using principal component analysis, the first three components determined 72% of the total variation. Ear emergence date, pollination date, stem number and stem height in cut1 and annual dry matter were the most important traits in first components. Basal cover, stem number and stem height of cut 2 and persistency were the important traits in second components. The 29 genotypes were grouped into 5 clusters based on multivariate analysis of 16 classification variables. Accessions in cluster 1 averaged well above the overall mean for productivity and persistency. Only one accession was in cluster 2 that was late maturity and had lower dry matter yield, tiller number and tiller height. The accessions in cluster 3 were tetraploid and late maturity. They had more dry matter yield in cuts 2 and 3. Cluster 5 characterized by early maturity, higher dry matter yield, higher stem number and higher stem height in cut1. But, it was less persistent than other clusters. Distribution of genotypes based on the first two component scores was in agreement with cluster analysis. Considerable heterosis would be obtained in crosses between populations from more widely genetically separate groups.