We developed stage of development descriptions which we believe apply to all soybean (Glycine max (L.) Merr.) genotypes grown in any environment. The descriptions apply to single plants or a community of plants and are precise and objective.Vegetative and reproductive development are described separately. Vegetative stages are determined by counting the number of nodes on the main stem, beginning with the unifoliolate node, that have or have had a completely unrolled leaf. Reproductive stages Rl and R2 are based on flowering, R3 and R4 on pod development, R5 and R6 on seed development, and R7 and R8 on maturation.The stage descriptions should enhance soybean research by standardizing descriptions of soybean plant development. The system also will be used by the soybean hail insurance industry for stage determination in adjustment of losses.

The estimation of yield loss from defoliation and stem cut‐off is based on the assumptions that soybean (Glycine max (L.) Merr.) cultivars do not differ appreciably in their ability to recover from plant injury, recovery is similar across locations, and percentage yield loss at all reproductive stages can be approximated from research involving only three of the stages. The objective of our study was to test the validity of the assumptions by comparing the response of indeterminate and determinate soybean cultivars to 100% defoliation and 100% half‐plant cut‐off at six reproductive stages from R2 to R7. Two indeterminate cultivars, ‘Hark’ and ‘Beeson’, were studied at Ames, Iowa, and Lafayette, Ind., and two determinate cuitivars, ‘Hill’ and ‘Lee’ were evaluated at Stuttgart, Ark.The determinate cultivars had significantly greater yield reduction from 100% defoliation than did the indeterminate cultivars at all reproductive stages, except R7. Average yield reduction from defoliation for all stages was 59% for the determinate cultivars, compared with 39% for the indeterminate cultivars. Maximum yield loss from 100% defoliation occurred at R4 (86%) and R5 (88%) for determinate cultivars and at R5 (82%) for indeterminate cultivars. Average yield loss from half‐plant cut‐off was similar for indeterminate (34%) and determinate (33%) cultivars, but there was a significant interaction with stages. Yield loss for indeterminate cultivars increased from R2 to R5, then remained constant from R5 to R7. The yield loss from half‐plant cut‐off with determinate cultivars increased progressively rom R2 to R7. The determinateness of a cultivar should be considered when assessing yield reductidn from defoliation and half‐plant cut‐off during reproductive development. Damage from hail, insects, and other factors will be most accurately estimated by establishing separate loss values for indeterminate and determinate types.There was no significant difference between Iowa and Indiana in the percentage of yield loss from 100% defoliation. For the cultivars and locations studied, location effects were not important in the estimation of yield loss from plant injury.Yield loss from 100% defoliation at the six reproductive stages did not follow the relationship predicted from previous studies. Our data indicated that maximum loss for indeterminate cultivars occurred at R5, not R4 as has been assumed. Some yield loss occurred at R7 indicating that the description used to identify plants at physiological maturity was incorrect. A new description based on pod or seed maturation is needed to identify plants that have reached physiological maturity.

Previous experiments have shown that some cross‐pollination (usually <1%) occurs in soybean [Glycine max (L.) Merr.] under field conditions. Information is needed on a broader spectrum of germplasm to determine if genotypes differ significantly for this trait and should be managed differently to produce genetically pure seed in small plots bordered by other genotypes, Twelve soybean cultivars differing in maturity were evaluated during two growing seasons to determine the extent of cross‐pollination under field conditions where both honey bees (Apis mellifera L.) and indigenous insect populations were present. Flower color, pubescence color, and quintafoliolate leaves were used as genetic markers to measure extent of cross‐pollination when plants were grown in adjacent rows 102 cm apart. In maturity group (MG) IV, cultivars did not differ significantly (P > 0.05) for percent cross‐pollination in 1990, but were different in 1991 and when the 2 yr were combined. ‘Competitor’ was significantly higher for percent cross‐pollination (1.22%) than the other cultivars within this MG. Cultivars within MG V and V1 differed significantly in 1990 but were not different in 1991. Based on a 2‐yr average, cross‐pollination varied from a low of 0.09% for ‘Walters’ to a high of 1.63% for ‘Brim’. Results show that cultivars differ significantly (P > 0.05) in the extent of cross‐pollination and as much as 2.5% outcrossing may occur in MG IV, V, and VI cultivars in some environments where adequate pollinators are present and other conditions are favorable. Based on these results, soybean breeders should use greater isolation of initial progeny rows and breeder seed of a new cultivar than has commonly been used in the past because of the increased potential for outcrossing of some cultivars.

Rather wide year‐to‐year fluctuations in natural crosspollination were found to occur when the Lee and Jackson varieties were planted in adjacent rows in Arkansas during a 3‐year period. Natural cross‐pollination was 0.44% in 1961 and 0.03% in 1962 and 1963 when the varieties were grown in adjacent rows. With distances of more than 15 feet from the pollen source, natural crosspollination was rare and did not vary greatly during the 3‐year period in these experiments.

Extensive research has been conducted on effects of moisture stress on growth and development of soybean [Glycine max (L.) Merr.] but information is limited on the response of determinate cultivars to moisture stress initiated at different growth stages. Moisture is a major factor limiting yield in most areas where soybeans are grown. Moisture stress experiments were conducted under field conditions using four determinate soybean cultivars of Group VI maturity, ‘Davis’, ‘Lee 74’, ‘Sohoma’, and ‘Centennial’. The experiments were conducted in 1980 and 1981 on a Captina silt loam (fine‐silty mixed, mesic Typic Fragiudult) at Fayetteville, AR. These experiments consisted of two moisture stress treatments which were initiated at the R2 and R4 growth stages and a control which was irrigated as needed throughout the season. Measurements were made on both roots and aboveground plant parts. When a moisture deficit was initiated by placing black plastic over the plots at either R2 or R4, seed yield was significantly reduced both years. A moisture deficit following application of plastic at R2 reduced yield more than at R4 in 1981, but the reverse was found in 1980, an extremely dry year. A moisture deficit initiated at R4 significantly reduced seed size and seed number both years whereas seed size was not significantly reduced either year by deficits initiated at R2. Sohoma produced significantly larger seeds than the other three cultivars when plants were moisture stressed at R4. No significant cultivar X moisture stress interactions involving seed yield were observed in either year; however, there was a definite trend for seed yields of Sohoma to be reduced less by moisture deficits initiated at stage R4 in both 1980 and 1981. Percent flower and pod shedding did not appear to be affected to any great extent by the moisture stress treatments in either year. In general, as the total number of flowers and pods increased, the number of flowers and pods shed increased proportionally. There were about four times as many roots under the row as were present in the furrow middles in both years. Also, a moisture deficit restricted root growth under the row in the upper portion of the profile, but increased it in the lower portion in the dry year of 1980.