Grey correlational and SSR analyses of cotton hybrids

Ten upland cotton strains exhibiting 3 fiber quality traits and 8 yield traits, were grown for two years in an investigation of the correlation between grey relational analysis （GRA） and genetic identity in heterosis of cot- ton hybrid. The aim was to establish the optimal approach for heterosis prediction and parent selection. Plant traits data were collected and analyzed for GRA. In addition, 72 simple sequence repeat （SSR） markers were examined and 148 polymorphisms were detected. Correlation analysis of GRA, genetic identity, Ft fiber quality and yield heterosis was conducted. Significant differences were observed between the two analytic methods, whereas compa- rable predictions were given for yield heterosis. GRA for yield exhibited slightly higher correlation than genetic identity analysis, with a correlation coefficient of 0.49. GRA and genetic analysis exhibited overlapping yet dis- tinct advantages in heterosis prediction. Therefore, these analytical methods should be integrated to achieve the op- timal heterosis prediction and parent selection.