This research aimed to estimate the appropriate number of S1 progenies that should be used in recurrent selection programs in maize (Zea mays L.). Two-hundred S1 progenies from BR-105 and BR-106 maize populations were evaluated in two environments in randomized complete blocks designs with three replications per environment. The progenies from each population were random split in groups of twenty-five progenies, supplying sample sizes from twenty-five to two-hundred progenies. For each sample, genetic and phenotypic variances, coefficients of heritability as well as their respective confidence intervals were estimated for the following traits: ear weight (EW), plant height (PH), ear height (EH) and prolificacy (PR). As the number of progenies increased, the width of the confidence intervals of the estimated parameters decreased for all traits evaluated and for both populations. For the main trait, ear weight, the estimates of the genetic and phenotypic variances, the coefficients of heritability and the confidence intervals stabilized around 175-200 progenies for both populations. For the other traits, PH, EH, and PR the estimated parameters stabilized around 125, 125, and 175 progenies, respectively, for both populations. Hence, for recurrent selection programs, at least 200 S1 progenies should be used regardless of the genetic type of the population.
Zea mays; sample size; genetic variance; heritability
