Recurrent Selection for Seed Size in Soybean: II. Indirect Effects on Seed Growth Rate

Abstract

Seed growth rate (SGR) and effective filling period (EFP) have been associated with both seed size and seed yield in soybean [Glycine max (L.) Merr.]. Three subpopulations (designated SS1, SS2, and SS3) the intermating population N79‐1500, which segregates for male sterility (ms1ms1), were subjected to recurrent selection for increased seed size of male‐sterile (MS) plants. The objective was to examine the relationship between seed size and development, and their effects on seed yield in the materials derived from selection. After four cycles of selection in SS1 and SS2, and three cycles in SS3, a composite of the parents of each cycle was tested at two locations in North Carolina in 1987 and 1988. A random sample of pods was collected on three dates ≈ 10 d apart from the upper third of the canopy beginning at growth stage R5. The SGR was estimated by linear regression of seed weight on sampling day after 31 August, and EFP was calculated as mature seed size divided by SGR. A significant linear increase in SGR occurred in SS2 and SS3 across cycles of selection. Differences in EFP within and among subpopulations were not significant. The SGR apparently was more important than EFP in producing the increases in seed size of male‐fertile (MF) plants. Seed size increased significantly in all three subpopulations and seed yield increased significantly in SS2 and SS3, but there were no significant differences across cycles of selection for seed number (seeds ha−1).