Document Type

Dissertation - Open Access

Award Date


Degree Name

Doctor of Philosophy (PhD)


Plant Science

First Advisor

C. Dean Dybing


Field studies were conducted in 1982, 1983 and 1984 in ten soybean (Glycine max (L.) Merrill) genotypes of maturity group 00 to III, to evaluate the effects of genotype and year on flower abortion (FA), flowers/raceme (FPR), raceme length (RL), pod types (PT), seeds/raceme (SPR), seed weight (SW), and yield/raceme (YPR). Reproductive efficiency index (REI) was calculated for parental genotypes and their segregating populations. The effects of main stem node position and raceme flower position on FA and PT were also studied. Genetic studies were on F2's of 13 crosses plus F3 and F4 generations of 'IX93-100' x 'Amsoy. Effects of selection for flower abortion and its impact on REI, seed yield and yield components were studied in the F4 of IX93-100 x Amsoy. Significant genotypic differences were found for all characteristics measured. FA ranged from 33% for Amsoy to 78% for 'Chusei'. Genotypes with longer racemes produced and aborted many more flowers than those with fewer flowers and shorter racemes. Exceptions to these were observed in the HE-78 lines which suggested that unit RL and FPR are not the sole determinants of FA. Magnitude of FA varied with years but the ranking of genotypes did not change. Percent PT (aborted or set into 1-,2-,or 3-seeded pods) of a genotype was not affected by main stem node position. Results suggested that one or two nodes/plant immediately above the top-most branch can represent a genotype for FA. Although genotypic differences were evident for the fates of the first four flowers on raceme, the true picture of genotype with respect to FA was not represented by these positions. Therefore, study of flowers beyond the fourth floral position on raceme would be more effective in identifying superior genotype for increasing seed yield. The fate of flowers produced at different sections of raceme depended both on floral position and genotype. Crosses differed for average FA and other characters with segregation occurring for all traits studied in all crosses. The ranges of F2 and F3 distributions indicated transgressive segregation for all characters. A continuous distribution suggested quantitative inheritance on-normality of distribution in F2 and F3 plus higher narrow sense heritabilities indicated that parents differed for few major genes with large additive effects for FA, FPR, RL and PT. Presence of partial to no dominance on different crosses depended on genotypes of the parents. The possibility that flower, pod and seed abortion are the same events was suggested by the association of lowest percent 3-seeded pods with lowest REI, association of highest percent 3-seeded pods with highest RE I, and near perfect regression line between REI and FA in parents and segregating populations. They might share the same genetic system and the only difference might be in timing of the event. FA, FPR and RL were positively associated with each other and negatively associated with PT. SPR showed negative relation with FA in all populations except in high x low FA crosses. Results of yield components associations in these crosses suggested possibility of developing genotypes that produce many SPR but still have relatively higher FA than low aborting parent. Selection for FA in the F3 of IX93-100 x Amsoy was effective in changing abortion to levels theoretically expected. There were corresponding changes in other characters. Selection for low FA increased percent 3-seeded pods, REI, SPR, SW, hastened maturity, decreased internode length, and consequently increased seed yield/unit area when compared to the parents and population selected for high FA. Although results indicated that early generation selection for FA was effective in improving REI and seed yield, it is suggested that direct selection for high REI would be more effective for seed yield improvement in soybean. Wider ranges of segregation coupled with high narrow sense heritability for flower abortion and pod type, the variation in magnitude and sign of character associations plus the results of the selection experiment suggested that selection is possible for longer racemes with lower level of flower abortion, higher REI with higher seed weight to improve seed yield in soybean.

Library of Congress Subject Headings

Soybean -- Genetics
Soybean -- Seeds




South Dakota State University