Premium
Direct and Indirect Recurrent Selection for Protein in Soybeans 1
Author(s) -
Miller J. E.,
Fehr W. R.
Publication year - 1979
Publication title -
crop science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.76
H-Index - 147
eISSN - 1435-0653
pISSN - 0011-183X
DOI - 10.2135/cropsci1979.0011183x001900010025x
Subject(s) - biology , population , selection (genetic algorithm) , inbreeding , zoology , microbiology and biotechnology , demography , artificial intelligence , sociology , computer science
Our objectives were to evaluate the effectiveness of population improvement by recurrent selection for protein percentage in soybean [ Glycine max (L.) Merr.] seed by direct selection for high protein and indirect selection based on low oil, and to investigate methods of preventing change in maturity of populations during recurrent selection. A population was developed by crossing 12 high protein lines with 12 high yielding lines and intermating their progeny three times. A random sample of 100 S 1 lines from the cycle 0 population was evaluated in two replications at two locations in Iowa in 1975. The 10 lines with highest protein and 10 with the lowest oil were used to develop two cycle 1 populations. Performance of 100 S 1 lines from each population was evaluated in two replications at two locations in Iowa in 1976. Protein increased from 43.1% in the cycle 0 population to 44.6% in the high protein and 43.9% in the low oil populations. Direct selection caused a significant decrease in both oil and carbohydrate, and indirect selection significantly reduced oil but did not change carbohydrate percentage. Selection for low oil cannot be considered an efficient method for protein improvement when cost and time for protein analysis are not limiting factors. Maturity was 5.3 days later in the high protein population and 9.3 days later in the low oil population than in the cycle 0 population. The change in maturity was unacceptable, and three selection methods were evaluated, with and without restrictions on inbreeding, for avoiding similar shifts in the selection of lines for development of cycle 2 populations. Maturity class selection, linear regression adjustment, and a selection index were all successful in selecting lines with an average maturity similar to the cycle 1 populations. Restrictions on amount of inbreeding generally reduced expected gain from selection for seed composition.