RESPONSE TO RECIPROCAL RECURRENT SELECTION AND GENETIC DIVERGENCE IN TWO TROPICAL MAIZE (Zea mays L.) COMPOSITES
Average productivity of maize (Zea mays L.) in the moist savanna of West and Central Africa (WCA) is still considerably low. Improving maize populations with recurrent selection methods is important for developing divergent inbred lines with good combining ability, for maintaining genetic gains and enhancing the utility of genetic pools. Two tropical complementary maize composites representing an alternative heterotic pattern, namely TZL COMP3 and TZL COMP4 were formed following a Comprehensive Breeding System (CBS). These populations have undergone four cycles of Reciprocal Recurrent Selection (RRS) for grain yield, resistance to foliar diseases and other desirable agronomic traits using reciprocal S1 testcross evaluation scheme. The objectives of this study were to: (i) evaluate the effects of RRS on the rate of genetic gains in grain yield and other traits as well as the changes in variance and heritability estimates as well as correlations between grain yield and other traits in the two tropical maize composites, (ii) evaluate the effect of selection on genetic gain in heterosis for grain yield, foliar disease resistance and other desirable agronomic traits, (iii) determine the extent of genetic variation within and between selection cycles of the two tropical broad-based composites (TZL COMP3 and TZL COMP4) and to determine the effect of RRS on changes in population structure of randomly selected S1 lines from the original (C0) and advanced (C4) selection cycles of the two tropical composites using Single Nucleotide Polymorphism (SNP) markers. A total of 200 paired S1 line testcrosses, one hundred each from C0 and C4 of TZL COMP3 and TZL COMP4 plus ten commercial hybrid checks were evaluated across eight environments in Nigeria. Also, ten parental populations, five inter-population crosses from selection cycles of TZL COMP3 and TZL COMP4 and one commercial check were evaluated at four locations in 2014 and 2015 in Nigeria. The effects of RRS on the original and advanced selection cycles of the two tropical composites were measured by analyzing the variation at 10,579 SNP loci among 400 S1 lines. Testcrosses differed significantly for grain yield and other agronomic traits except for Curvularia leaf spot. Grain yield increased by 14 % after four cycles of RRS. Selection reduced days to silking, anthesis-silking interval, plant and ear aspect scores, and resistance to foliar diseases. Mean grain yields of the top ten C4 x C4 testcrosses exceeded that of the ten commercial hybrids by 29 %. The top ten advanced selection cycle testcrosses also performed better than the commercial hybrids for other desirable agronomic traits and resistance to foliar diseases. Estimates of genetic variance were significantly larger than zero for most traits before and after selection. Considerable genetic variability and relatively high levels of heritability were still available in the advanced selection cycles to ensure continued genetic gain from RRS. Genetic gain from selection for grain yield was associated with reduction in scores for ear aspects, southern corn leaf rust and southern corn leaf blight in the advanced selection cycle testcrosses. Evaluation of the parental populations and inter-population crosses indicated that RRS was effective. Mean grain yields of the advanced selection cycle inter-population crosses exceeded that of a late maturing commercial check by 30 %. Grain yield increased with selection in the inter-population crosses by 3 % per cycle. Selection also improved anthesis-silking interval, ear aspect, plant aspect and resistance to foliar diseases. A divergence in the frequency of favorable alleles between the parental populations created by selection through the fixation of different alleles at a locus in each population resulted in increased heterosis in the inter-population crosses. Mid-parent heterosis increased from 4 % (C0 x C0) to 24 % (C4 x C4). Grain yield accounted for positive average heterosis effects, suggesting the presence of non-additive gene effects. Dominance was towards higher mean performance of inter-population crosses for grain yield, number of ears per plant and ear height. With the genetic gains from RRS in this study, the advanced selection cycle can be useful for the development of diverse and stable lines with improved combining ability and also in the development of improved inter-population crosses for specific markets, thereby increasing farmers’ options. The results of the molecular study revealed that the proportion of alleles at both low and high frequencies decreased from C0 to C4, whereas those at intermediate frequencies increased at C4 in the two composites. More unique and other alleles were lost at C4 in TZL COMP3 relative to those in TZL COMP4. The changes in different measures of genetic diversity were either small or negligible with selection in the two composites. The proportion of markers departing from Hardy-Weinberg equilibrium (HWE) decreased with selection, whereas the total number of pairs of markers in Linkage Disequilibrium (LD) increased with selection in the two composites. Examination of changes in population structures using a model-based approach as well as cluster and multivariate analyses revealed a high degree of concordance in stratifying the 400 S1 lines into four non-overlapping groups corresponding to the two selection cycles each within the reciprocal composites. The observed molecular-based divergence between cycles within the same composite and the clear differentiation between the complementary composites indicated the importance of RRS for preserving genetic diversity for long-term selection. This increased the potential of the advanced selection cycles to sustain genetic gain in productivity of maize hybrids adapted to the savannas of WCA.