Investigation of correlation and causal relationships affecting yield of 14 cotton genotypes in normal conditions and imposing drought stress after flowering

The purpose of this study was to identify the most important traits affecting yield in 14 cotton genotypes in normal conditions and imposing drought after flowering. Two separate experiments were conducted in a randomized complete blocks design with three replicates at the Agriculture Research Farm, University of Birjand during 2013. Phenotypic correlation analysis showed that the correlation between cotton yield with fiber weight, biological yield, harvest index, 1000-seed weight, number of bolls per plant, days to second harvest, oil percent in normal conditions and with fiber weight, biological yield, harvest index, 1000-seed weight and number of bolls per plant in stress condition was positively significant. The stepwise regression analysis showed that the boll number per plant and fiber weight was the most important traits that affected the cotton yield in normal and stress conditions, respectively. The path analysis showed that the fiber weight and biological yield in normal and the biological yield and harvest index in stress conditions had the greatest positive impact on cotton yield. For two conditions the genotypes were grouped into five classes by cluster analysis. Factor analysis was detected that five factors in normal conditions and four factors in stress conditions, showed % 90.03 and % 81.28 of total of variations. In normal and stress conditions yield factor selected as the first and secend important factors, respectively.