Selecting for increased weight per seed among male-sterile soybean pla
nts [Glycine mar (L.) Merr.] increased seed weight and seed yield for
male-fertile progeny but had minimal effects on seed number per ha. As
similate availability to the seed during selection was essentially non
limiting for an extended period, since male-sterile plants have few se
eds per plant. For the observed range, seed weight has shown no associ
ation with Seed yield potential. This study tested two alternate hypot
heses to explain the increased seed yield potential: (i) selection ide
ntified genotypes with increased seed sink activity per plant or (ii)
selection identified genotypes with reduced seed maturation rates. Dry
matter accumulation rate (DMAR) per unit seed dry weight, DMAR per un
it seed coat area, and rate of increase in seed density were determine
d for greenhouse-grown plants using a 7-d period during mid seed devel
opment. The following observations led to the acceptance of the reduce
d seed maturation rate hypothesis: (i) When assimilates were made esse
ntially nonlimiting to the seed by partial pod removal, selected and n
onselected populations had similar DMAR per unit seed coat area and pe
r unit seed dry weight. (ii) Selection for seed weight on male-sterile
plants reduced the rate of increase in seed density. (iii) DMAR per u
nit seed coat area did not increase with selection. Selected and nonse
lected populations, had similar DMAR per unit seed dry weight. Apparen
tly, selection had occurred against the reduction in rate of dry matte
r uptake by the seed. This reduction is normally associated with reduc
ed seed maturation rate. Selection for seed weight on male-sterile pla
nts can be completed during the early generations of a breeding progra
m and used to identify genotypes that have longer seed filling duratio
ns and more efficient systems for assimilate utilization.