The number of seeds per unit area is an important yield component in soybean [Glycine max (L.) Merr.]; however, the mechanisms responsible for the regulation of this yield component are not well understood. Field experiments were conducted at Lexington, KY (3 yr), and at Taian, China (1 yr), to investigate the relationship between net canopy photosynthesis and seeds per unit area using genotypes with differences in individual seed growth rates (SGR). At Lexington, shades (30 and 63% reduction in insolation) were placed over plots from growth stage Rl until maturity to create differences in canopy photosynthesis. Planting dates (early and late) and row spacing (wide and narrow) were used at Taian to create differences in canopy photosynthesis. Crop growth rate (CGR) was measured between growth stage Rl and RS as an estimate of net canopy photosynthesis. Yield, seeds per m2, and SGR were also measured. Within each genotype, there was a linear relationship between CGR and seeds per m2 across treatments and years. Within an experiment, seeds per m2 at a constant CGR was inversely related to genotypic differences in SGR. A partitioning coefficient (γ) was estimated by dividing the total sink demand (seeds per m2 ✕ SGR) by CGR. There were no apparent genotypic differences in γ however, γ decreased linearly as CGR increased. The data suggest that the model proposed by Charles‐Edwards, which describes seeds per m2 as a direct function of canopy photosynthesis and a partitioning coefficient and an inverse function of assimilate flux to individual seeds, accurately describes the regulation of seeds per m2 in soybean.

Planting date is a critical aspect of all soybean [Glycine max (L.) Merr.] production systems, but the response of yield to planting date fl uctuates widely among environments. A combined analysis of many planting date experiments will provide a better estimate of the average response. Data from 28 nonirrigated planting date experiments that were conducted for more than 1 yr with at least three planting dates (a minimum of one in June) were combined by regions (Midwest, 9; Upper South, 10; Deep South, 9) for analysis. Experiments using the early soybean production system (ESPS) were not included. A segmented-linear regression model with two segments was used to determine when yield began a rapid decline as planting was delayed. Th e rapid decline began on 30 May in the Midwest, 7 June in the Upper South, and 27 May in the Deep South. Th e rate of decline was larger in the Upper South (1.1) and Deep South (1.2) than it was in the Midwest (0.7% points per day). Yield trends before the rapid decline began showed no signifi cant advantage for early plantings. Only 23% of the April or early (fi rst week) May plantings exhibited higher yields than later plantings and the average advantage was 7%. Th e average response to planting date was remarkably similar across all regions and there seems to be no consistent advantage for ultra-early planting dates with traditional cultivars, but there was a signifi cant penalty for planting aft er the critical date in late May or early June.

Seed viability and vigor directly affect the performance of seeds planted to regenerate the crop. Although seed quality can influence many aspects of performance (e.g., total emergence, rate of emergence), the objective of this review was to examine the relationship of seed vigor to one aspect of performance: crop yield. Reductions in yield can be indirectly related to low seed vigor if plant populations are below a critical level. Thus, we have reported only on the direct effects of seed vigor on yield in the absence of population differences. Only those references where seed vigor was measured or where seed lots were evaluated following natural or artificial storage (both reduce seed vigor) were included. Annual crops were subdivided into those harvested during vegetative growth (eight species), early reproductive growth (four species) or full reproductive maturity (nine species). Seed vigor affects vegetative growth and is frequently related to yield in crops that are harvested vegetatively or during early reproductive growth. However, there is usually no such relationship in crops harvested at full reproductive maturity, because seed yields at full reproductive maturity are usually not closely associated with vegetative growth. The use of high‐vigor planting seed can be justified for all crops, however, to ensure adequate plant populations across the wide range of field conditions that occur during emergence.

A gronomy J our n al • Volu me 10 0 , I s sue 4 • 2 0 0 8 ABSTRACT Th e widespread adoption of glyphosate [N-(phosphonylmethyl)-glycine]-resistant soybean [Glycine max (L.) Merr.] and the increased cost of soybean seed have generated interest in determining the minimum plant population needed for maximum yield. Th e objective of this study was to determine yield and economic return responses to plant population for normal and late planting dates. Cultivars with relative maturities of 2.8 to 4.9 were planted at fi ve seeding rates (43,000 to 560,000 seeds ha -1 ) in May and/or June in 38-cm rows during 2003 to 2005. Th e eff ect of plant population on both yield and economic return was explained with a variation of a Mitscherlich equation. Optimum plant population (OPP) and economically optimum plant population (EOPP) were defi ned as those resulting in 95% of the estimated yield or estimated economic return, respectively, at the maximum plant population. Optimum plant population ranged from 108,000 to 232,000 plants ha -1 for May planting dates and 238,000 to 282,000 plants ha -1 for June planting dates. Economically optimum plant populations were 7 to 33% less than OPPs. Complete canopy cover at R1 produced maximum yield in 8 of 10 comparisons. Th ese results suggest that seeding rates below those that are currently recommended could lower seed costs without reducing yield.

Seed number per unit area is important in determining yield in soybean [Glycine max (L.) Merr.]. Field and greenhouse experiments were conducted to evaluate the role of flowers per plant and flower and fruit abscission in determining pod and seed numbers. Two greenhouse experiments were conducted with the McCall cultivar [Maturity Group (MG) 00] and four cultivars [McCall, Hardin (MG I), Harper (MG III), Essex (MG V)] were used in field experiments in 1989 and 1990. The field experiments used either 0.38‐m (McCall and Hardin) or 0.76‐m rows (Harper and Essex) with a constant seeding rate of 26 seeds per meter. Shade cloth (63% in the greenhouse and 30 and 63% in the field) was placed over plants from R1 to maturity. Additional treatments in the greenhouse included shade from R1 to R4, R4 to R5, R5 to R6, and R4 to R8. Plants were harvested at maturity and flower and pod abscission was determined by counting both the scars marking the point of attachment of flowers and pods, and mature pods. Shade reduced pod numbers and the reductions were due to both increased flower and pod abscission and fewer flowers per plant. Both the number of nodes on the main stem and flowers per node were affected by the shade. Shade during any part of the R1 to R6 period reduced pod numbers in the greenhouse. These data indicate that environmentally induced variation in pod numbers in soybean is a result of both changes in flower and fruit abscission and the number of flowers produced.

The effects of short periods of water stress during seed‐filling on leaf senescence, seed‐fill duration, and yield of soybean [Glycine max (L.) Merrill] are not well understood. Short stress periods were investigated in two greenhouse experiments with cultivar Elgin 87 grown in soil‐filled pots. All pots received adequate water until the beginning of growth stage R6 when a continuous water‐stress treatment (pots received 40% of the water needed to bring controls to pot capacity) was initiated and maintained until maturity. Water stress was relieved in other pots (watered as the control) after 5 or 13 d in Exp. 1 and 3 or 6 d in Exp. 2. Each treatment was replicated six to eight times in a completely randomized design. The carbon exchange rate was rapidly reduced by continuous water stress resulting in earlier maturity, significantly lower yield (39%), and smaller seeds (25–33%). The carbon exchange rate rapidly increased to near control levels in the early stress‐relief treatment, but it was always less than the control for the rest of seed filling. These plants matured sooner and produced significantly lower yields (10–23%) and smaller seeds (9–17%) than control plants. Late stress relief also reduced yield and seed size relative to the control. Yield and seed size of both stress relief treatments, however, were greater than the continuous stress treatment. Water stress‐induced acceleration of senescence could not be stopped by eliminating the stress after a short period. Short periods of water stress during seed filling may, therefore, have larger than expected effects on yield.

The growth rate and duration of an individual seed are important parameters in the yield production process of a grain crop. Experiments were conducted in the CSIRO phytotron at Canberra, Australia to investigate the effect of temperature on the rate and duration of seed growth and associated plant characteristics in soybeans [Glycine max (L.) Merr.]. Soybeans (‘Fiskeby V’) were grown at 24/19 C until beginning seed growth and shifted to 18/13, 24/19, 27/22, 30/25, and 33/28 C (day/night temperature, 8‐hour day) until maturity. Seed growth rates (SGR) were estimated by harvesting seeds at 5‐day intervals, and from the growth rate of excised cotyledons cultured in nutrient media for 96 hours. The SGR of seed developing on the plant increased from 6.1 to 7.9 mg seed−1 day−1 as air temperatures increased from 18/13 to 27/22 C, but there was no further change as the temperature increased to 33/28 C. Excised cotyledons showed a similar growth rate response to temperature. Exposing the plants to high temperatures (33/28 C) during the period of flowering and pod set reduced SGR (36%) regardless of the temperature during the subsequent seed growth period, suggesting that seed growth is sensitive to temperature levels at these early stages. Seed respiration on a dry weight basis (mgCO2h−1 g−1) was not affected by temperature. The duration of seed growth was not affected by temperatures of 24/19 to 30/25 C but was reduced by 3 days at 33/28 C, and this was associated with accelerated leaf senescence as shown by leaf yellowing and reductions in CO2 exchange rate. Final seed size was reduced from 200 to 151 mg seed−1 at 18/13 and 33/28 C. The data suggest that SGR and duration are relatively insensitive to temperatures ranging from 24/19 to 30/25 when these are imposed after flowering and pod development. The reduction in the duration of seed growth at high temperatures may be one mechanism by which high temperatures reduce yield.

be less available soil moisture (Baldwin, 1974; Knapp et al., 1980). The yield loss from delayed planting cannot Late planting reduces soybean [Glycine max (L.) Merr.] yields inbe eliminated by irrigation (Boerma and Ashley, 1982; soybean-winter wheat (Triticum aestivum L.) double-cropping systems. We evaluated the hypothesis that the use of early-maturing Egli et al., 1987). Evaluation of late-planted systems soybean cultivars to shift reproductive growth into a more favorable with a crop simulation model (Egli and Bruening, 1992) environment would avoid some or all of this yield penalty. Soybean provided evidence that, in the absence of water stress, cultivars Hardin and Kasota [maturity group (MG) I], Burlison and lower levels of insolation during reproductive growth Elgin 87 (MG II), Pioneer 9392 and Probst (MG III), and Stressland were a major contributor to the yield loss, with temperaand Pennyrile (MG IV) planted in 38-cm rows were used in a 3-yr ture only becoming important for cultivars that matured irrigated experiment with two planting dates (early, mid-May; late, in late October or early November. late June) at Lexington, KY (38؇ N lat). Delayed planting reduced Combining early-maturing soybean cultivars with yield (7-36%) of all cultivars as a result of fewer seeds m Ϫ2 . Cultivars early planting to shift reproductive growth into a more from MG I and II did not produce higher yields in the late plantings.favorable environment (i.e., avoid drought) increased A combination of narrow rows (19 cm) and high seeding rates (105 seeds m Ϫ2 ) had no effect on yield of cultivars from MGs I and II in yields at several southern USA locations (Kane and either planting date. However, early maturity did provide an earlier Grabau, 1992;Bowers, 1995; Heatherly, 1999). It may be harvest date without significant yield loss. Seed number was signifipossible to use an analogous strategy with late plantings. cantly related to crop growth rate (CGR) during flowering and podEarly cultivars in late plantings could also shift reproset (r 2 ϭ 0.36) and to length of flowering and pod set (r 2 ϭ 0.56).ductive growth into a more favorable environment, al-Radiation use efficiency (g dry matter MJ Ϫ1 intercepted photosynthetthough in late plantings the primary objective would be ically active radiation) was generally reduced in the late plantings for an improvement in insolation (Egli and Bruening, 1992) MG III and IV cultivars but not for MG I and II. Early-maturinginstead of avoiding drought. Its possible that the earlier cultivars in an irrigated environment did not reduce the yield penalty occurrence of reproductive growth in early cultivars associated with late plantings.could reduce the yield penalty associated with late planting dates in double-cropping systems. The soil was a Lanton silt loam (fine-silty, mixed, thermic Cumulic Epiaquoll) in 1996, a Donerail silt loam (fine, mixed mesic Oxy-