Factor analysis of yield, yield components and someeffective characters for drought resistance in lentil (Lens culinaris Medik)

Factor analysis is one of the multivariate statictical methods which can reduce a large number of correlated variables in small number uncorrelated factors. In order to determain the morphological and physiological effective characters on drought resistance of lentil, twenty lentil gnotypes evaluated on the basis of 8 morphological and physiological characters in a split plot design with replications at the Zanjan University research farm during the spring of 004. Factor analysis based on mean values over two stressed and non-stressed environments distributed 8 variables in four factors. For easy in explanation, only those factor coefficients that were greater than 0.8, were selected as signification without considering their sign. The first factor could explain 0.8 percent of total variance. The most important coefficients in this factor were related to plant dry weight, leaf dry weight, steam dry weight, leaf area and days to flowering. Significant and postive correlations were observed for all these traits. The second factor explain 5.6 percent of total variance. In this factor, plant hight at before flowering time, harvesting time, canopy width and number of primery branches were the most important traits.There were negative and significant correlations between plant height in flowering, before flowering and harvesting time with canopy width and number of primery branches. There was also positive and significant correlation between canopy width and number of primery branches. The third factor explained 4.0 percant of total variance.The most important traits in this factor were extracted as harvest index, seed yield, pods per plant, water content of leaves and seeds/plant. The fourth factor explained 0.66 percant of total variance.The most important traits in this factor were selected as biological yield, leaf temperature and 00 seed weight. There was a negative and significant correlation between seed yield and leaf temperature, also a positive and significant correlation between 00 seed weight and seed yield was seen.