Direct and Indirect Selection for Small Seed of Soybean in Temperate and Tropical Environments

Some food items made from soybean [Glycine max (L.) Merr.] require small seeds. The wild soybean, G. soja Sieb. & Zucc., has been used as a source of genes for small seed size for development of high‐yielding cultivars for specialty markets. The objective of this study was to determine the effectiveness of direct and indirect selection for small seed in three interspecific crosses of soybean (G. max ✕ G. soja) based on seed weight and pod width measurements. Indirect selection was based on pod width. Forty random BC2F2‐derived lines from each of three crosses were evaluated for seed weight and pod width at two environments in Iowa and two environments in Puerto Rico. Broad‐sense heritability estimates among the three crosses based on variance component estimates for seed weight were 35% on a plant, 52% on a plot, and 89% on an entry‐mean basis, and for pod width were 35% on a plant, 51% on a plot, and 87% on an entry‐mean basis. The actual genetic gain averaged across environments and crosses for direct selection of seed weight was 6 mg seed−1 on a plant, 6 mg seed−1 on a plot, and 7 mg seed−1 on an entry‐mean basis. Indirect selection by pod width resulted in an actual genetic gain of 6 mg seed−1 on a plant, 7 mg seed−1 on a plot, and 8 mg seed−1 on an entry‐mean basis. Selection in Puerto Rico was as effective as selection in Iowa. The results indicate that both direct and indirect selection for small seed are effective in the temperate and tropical environments that were studied.