Comparison of S1 with Testcross Evaluation for Recurrent Selection in Maize1

Abstract

We made S1 and testcross progeny evaluations for two cycles of recurrent selection for grain yield in a 13‐line maize (Zea mays L.) synthetic. In the first cycle we evaluated 60 lines at four locations in 1 year. For the second cycle we evaluated 68 lines in three trials during 2 years. Selection intensity was 17% in both cycles.Genetic variance for grain yield in the first cycle was significantly greater for selfed progeny in comparison with testcross evaluation, but decreased in the second cycle. Variability for ear moisture, broken stalks, and root lodging was increased with both selection procedures. S1 selection for grain yield gave greater improvement in the population per se, and was equal to testcross selection for general combining ability improvement. However, one second‐cycle testcross progeny exceeded the highest‐yielding parental line testcross.Genotypic correlations were greater between related and nonrelated testcrosses than between S1 and either testcross. However, several selections were common to both S1 and testcross procedures. Second‐cycle correlations were significantly lower than first‐cycle values.