Evaluation of genetic diversity for drought tolerance in stem cuttings of Rosa damascena Mill using multivariate analyses

Authors

1 Research Institute of Forests and Rangelands, P.O.Box 13185-116, Tehran-Iran

2 Imam Khomeini Center for Higer Education, P.O.Box 13145-498, Tehran-Iran

Abstract

In this study, drought tolerance in Rosa damascena from different parts of Iran, was analyzed in Research Institute of Forests and Rangelands in 1991. Growth pattern of stem cuttings of 30 genotypes were evaluated in recovery environment, after a 21 day water stress, by multivariate analyses. The scored characters were plant survival, branch number, leaf number, leaflet number, leaflet area, longest branch, and plant vigor. The statistical multivariate methods applied included multivariate linear regression, path analysis, principal components analysis (PCA) and cluster analysis. According to the results of regression analysis, plant vigor, leaflet number, and leaflet area showed a significant association with survival, and a good fit was obtained for this association by linear regression model. Path analysis revealed that the greatest direct effect on plant survival being positive was related to leaflet number, and the lowest direct effect which was negative, was related to the leaflet area. The direct effect of plant vigor on survival was negative, but relatively high. Also, indirect effects of plant vigor vialeaflet number and leaflet area were positive and negative, respectively. Leaflet number is therefore the most important trait and it could be possible to increase drought tolerance of Rosa damascena by selection for this variable. Principal components analysis (PCA) presented three components indicating more than 82% of the total variation. The highest special values belonged  to second component , based on which genotype 88B was selected for survival percentage under drought stress. Cluster analysis classified the genotypes in seven groups, based on the traits under study. The obtained result therefore, indicated that the highest variation could be achieved for the traits under study, through combining genotypes in the first and seventh groups.

Keywords