Selection response and genetic variances in three replicates of reciprocal full-sib recurrent selection in maize (Zea mays L.)

This study was focused to evaluate three Reciprocal Full-Sib (RFS) selection replicates applied to the NBS and NSS populations. The selection criterion was a multiplicative index involving the grain yield adjusted by the grain moisture, upright plants and undropped ears. Because each replicate started in a different year, genetic drift and environmental effects together are expected on the selection response in each selection replicate. The objectives of this study were the estimation of the change of the population means for the index and related traits, estimation of the genotypic variance before and after eight cycles of RFS selection, and comparison of the RFS selection replicates based on these estimations. The RFS selection method was evaluated based on the performance of the inter-population crosses at two locations in 1995 and one location in 1996, and the testcross populations at one location in 1995 and 1996. The results of the combined analyses are reported in two separated chapters. The population means for the index and grain yield increased in both types of population crosses. The increments differed among RFS selection replicates. In the inter-population crosses, the RFS-3 selection replicate presented the highest increments for the index and component traits. In the NBS x T1 testcross populations, the RFS-1 and RFS-3 selection replicates presented the highest increments in the means for the index and grain yield. And in the NSS x T2 testcrosses, the highest increments were obtained in the RFS-2 and RFS-3 selection replicates. In both types of population crosses, the results indicated that the genotypic variance for the index was affected by selection. The initial variance for the index was found narrow since the beginning of the selection process, but large enough to get some selection response per cycle. Although, a reduction in the genotypic variance for the grain yield was detected, this reduction was not different between the Cycle 0 and the Cycle 8 of the three selection replicates in both types of population crosses