نوع مقاله : مقاله پژوهشی
عنوان مقاله [English]
The main objective of this study was to understand of genetic association between quantitative and qualitative traits and also to quantify effects of selection for one trait on others in safflower. The study took place in the experimental field of the faculty of agriculture, Isfahan University of Technology, Isfahan, Iran, in the 2001 and 2002. To find the best variable for selection to improve evaluated traits the correlated response to selection was calculated. The correlated response to selection indicates when selection takes place for a trait what happens for other related traits. The results indicated that, among all correlated traits with seed yield per plant the highest progress can be obtained by selection for increasing the number of head per plant, because correlated response to selection for number of head per plant was bigger than others. After the number of head per plant, the most important trait for improving of seed yield was number of seed per head. Also, if shortening of time to maturity be the aim of breeding project, the best choice is selection of dwarf genotypes, because the highest increasing of correlated response to selection was observed for plant height. On the results of this study it was proposed that evaluation of casual association between yield components and designing of ideotype in safflower must be done with considering of genetic correlations between all related traits and for selection of one trait much attention should be pay to its correlations with other traits. Our findings shows that increasing the number of head per plant improve seed yield and additionally decrease number of seed per head, and this can reduce amount of yield improving. Therefore, it seems necessary that for improving seed yield in safflower a suitable selection index would be defined. Plant height was the best variable for increasing oil content in indirect selection, because the results showed that its correlated response to selection was higher than other traits.