engleski

Comparing estimates of variances and correlations for two recurrent selection methods in the maize synthetic FAOSYNFR1B

The objective of this study was to compare estimates of variances and correlations for S1 and HSI methods in the second cycle of recurrent selection in the FAOSYNFR1B synthetic for stalk rot resistance and grain yield. We compared estimates of variance and correlations for S1 and HSI methods in the second cycle of recurrent selection in the maize synthetic FAOSYNFR1B for stalk rot resistance and grain yield when the components were re-estimated with restricted maximum likelihood (REML) method. There were some differences between variance components estimated by REML analysis and ANOVA due to adjusting of estimates in extended random model and by re-estimating the other variances because of some negative estimates. There was no either phenotypic or genotypic correlation between the two traits for the S1 method, while the both correlations for the HSI method were negative and very weak. The estimates of phenotypic and genotypic correlations were similar for both selection methods indicating that phenotypic correlations could alone accurately reflect expected correlated responses between the traits when sample size is relatively large.

maize; yield; stalk rot; REML; genetic variance

nije evidentirano