GENOTYPE BY ENVIRONMENT INTERACTIONS AND SELECTION RESPONSES OF TWO MAIZE POPULATIONS TO RECIPROCAL RECURRENT SELECTION IN A HIGH-YIELD ENVIRONMENT (CORN)

A selection program using reciprocal recurrent selection procedures in maize (Zea mays L.) involving intra- and interpopulation selection were used for BS10 and RSSSC. Intrapopulation selection included phenotypic mass selection for multiple leaf and stalk-rot diseases resistance and S(,1) family selection for grain yield in a high-yield environment. Interpopulation improvement involved selection of testcrosses of S(,1) family x inbred tester in a high-yield environment.;During the first six years of selection experiment (1978-83), the same 100 S(,1) and testcross families were grown in a high-yield and normal environment. Comparison of S(,1) and testcross data showed estimates of genetic variance component and heritability to be greater for the S(,1) families than for the testcrosses in each cycle. Genetic correlations between S(,1) families and testcrosses within the high-yield and normal environments were not very high. Estimates of genetic variance component and heritability were greater in high-yield environment than in normal environment. Genetic correlations between high-yield environment and normal environment were high for both the S(,1) families and testcrosses. Partition of genotype by environment interaction component into differences in genetic variance and differences in genetic correlations from one environment to another, were inconsistent among populations. RSSSC but not BS10 showed a consistent increase in differences in genetic variance in three cycles of selection.;Evaluations of populations per se, population crosses and populations x testers for response to two and three cycles of selection for grain yield were conducted in 1984 and 1985, respectively. Evaluation of populations per se and population crosses for response to three cycles of selection for multiple leaf disease and multiple stalk-rot resistance were conducted in 1985. The results indicate that selection response was consistent among environments. Gain from selection was satisfactory in populations per se, population crosses and populations x testers for grain yield, multiple leaf disease resistance and multiple stalk-rot resistance. Stalk lodging resistance was improved in BS10 and the population crosses. Correlated changes in other agronomic traits were mostly desirable.