PREDICTING HETEROSIS IN GRAINYIELD FROM GENETIC DIVERGENCE ANALYSIS IN SUNFLOWER

Genetic divergence analysis was conducted based on performance of 12 quantitative traits in 85 sunflower inbred lines，to study its effectiveness in predicting yield heterosis of the 72 hybrids from them. Results showed that linear regression model fit the relationship between heterosis of seed oil content and genetic distance。and quadratic regression equation the relationship between grain yield heterosis and genetic distance，whichcan be used in predicting heterosis from the parent performances. Cluster analysis was effective to certain extent，but its utilization should be limited. Grain yield and oil content can be improved simutaneously according to the relations between genetic divergence and the heterosis of these two traits.