Comparisons of agronomic traits in the initial and advanced cycles of IAP3BR(M) random-mating grain sorghum population

Abstract

<p>The random-mating sorghum Sorghum bicolor (L.) Moench population IAP3BR(M) was constituted in 1976 by crossing 30 large-seeded grain sorghum lines onto IAP1R(M)Cl. Half-sib (HS) bulk populations from selection cycles 1, 2, 3, and 4 along with S(,1) bulks from cycles 0, 2, 3, and 4 were evaluated in Experiment I. Sixty S(,1) families chosen randomly from both the initial (C0) and the fourth (C4) cycle of the population were evaluated in Experiment II. Both experiments were grown in two years near Ames, Beaconsfield, and Sutherland, Iowa;Four cycles of gridded mass selection for heavy 100-seed weight increased mean 100-seed weight, days to midbloom, plant height, and openness of panicle, while decreasing mean grain yield, seeds/panicle and panicles/plant. Estimates of the genotypic variance among S(,1) families were not significantly different in the C0 and C4 for 100-seed weight, seeds/panicle, and plant height, but they decreased for panicles/plant and increased for grain yield, days to midbloom, and panicle type. Means and ranges among genotypes indicated the population should be useful as a germplasm source for large-seeded genotypes with a diversity of agronomic characters;Estimates of inbreeding depression at 100% homozygosity were significant (P < 0.05) only for grain yield (-29.6%). The estimates for 100-seed weight (-14.0%), seeds/panicle (-10.2%), and plant height (-22.7%) were not statistically significant;Heritability of 100-seed weight on a progeny mean basis decreased slightly from the C0 to C4 (0.90 vs 0.85). Individual plant estimates of the heritability of 100-seed weight in the C0 (0.35) and C4 (0.39) were higher than the realized heritability (0.14);Phenotypic correlations between 100-seed weight and grain yield decreased from the C0 to C4 (0.33 vs. 0.06). The correlations between 100-seed weight and seeds/panicle were strong in both cycles (-0.72 and -0.78);Gridded mass selection was effective for increasing 100-seed weight, but S(,1) family testing seems necessary for obtaining the greatest gains in grain yield. The results indicate that continued selection for 100-seed weight in IAP3BR(M) would not produce appreciable increases in grain yield.</p>