In order to group winter rapeseed cultivars according to evaluated traits, an experiment was conducted in the Research Greenhouse of Agriculture Faculty, University of Tabriz-IRAN. In the experiment were included 12 cultivars of winter rapeseed and 3 levels of water deficit stress. Gypsum blocks were used to monitor soil moisture. Water deficit stress was imposed from stem elongation to physiological maturity. According to the principal component analysis, five principal components were chosen with greater eigenvalue (more than 0.7) that are including 81.34% of the primeval variance of variables. The first component that explained the 48.02% of total variance had the high eigenvalue. The second component could justify about 13.64% of total variance and had positive association with leaf water potential and proline content and had negative relationship with leaf stomatal conductivity. The third, fourth and fifth components expressed around, 10.18, 4.83 and 4.68% of the total variance respectively. The third component had the high eigenvalue for plant dry weight. The fourth component put 1000-seed weight, seed yield, Silique per Plant and root dry weight against plant dry weight, chlorophyll fluorescence and leaf water potential. The fifth component had the high eigenvalue for root dry weight, root volume and 1000-seed weight.

In order to group winter rapeseed cultivars according to evaluated traits, an experiment was conducted in the Research Greenhouse of Agriculture Faculty, University of Tabriz - IRAN. In the experiment were included 12 cultivars of winter rapeseed and 3 levels of water deficit stress. Gypsum blocks were used to monitor soil moisture. Water deficit stress was imposed from stem elongation to physiological maturity. According to the principal component analysis, five principal components were chosen with greater eigenvalue (more than 0.7) that are including 81.34% of the primeval variance of variables. The first component that explained the 48.02% of total variance had the high eigenvalue. The second component could justify about 13.64% of total variance and had positive association with leaf water potential and proline content and had negative relationship with leaf stomatal conductivity. The third, fourth and fifth components expressed around, 10.18, 4.83 and 4.68% of the total variance respectively. The third component had the high eigenvalue for plant dry weight. The fourth component put 1000-seed weight, seed yield, Silique per Plant and root dry weight against plant dry weight, chlorophyll fluorescence and leaf water potential. The fifth component had the high eigenvalue for root dry weight, root volume and 1000-seed weight.

Abstract:Finding the relationship between physiological traits and seed yield components is an important objective in crop breeding programs. Canonical correlation analysis has been adopted to study the strength of association between the physiological traits and seed yield under water deficit stress and to obtain the physiological traits that have the largest effect on seed yield and its components. This study revealed that leaf water potential, relative water content, leaf osmotic potential and chlorophyll index had the largest influence on seed yield and its components under severe water deficit. Under mild water deficit, leaf water potential and relative water content were also important for improving seed yield. Leaf water potential, relative water content, chlorophyll fluorescene and chlorophyll index were had the largest effect on seed yield and its components under well watered condition.