Cumulative Response to Various Recurrent Selection Schemes in Soybean: Oil Quality and Correlated Agronomic Traits1

Abstract

practiced in a soybean [Glycine max (L.) Merr.] population to indirectly decrease linolenic acid percentage in the seed oil. Three selection schemes were employed during specific stages of the program: Individual or mass (Cycles 1 to 3), mass plus within half‐sib family (Cycles 4 to 7), and S1‐progeny selection (Cycle 8). In addition, reverse selection for low oleic acid (LO) was initiated in Cycle 6 and continued simultaneously with HO selection. Objectives of this study were to evaluate cumulative selection response for fatty acid traits and other correlated traits, and to evaluate the potential for further selection progress. To achieve those objectives, selected lines from each cycle were composited in bulk populations and tested in eight environments. Regression analysis showed that the average rate of change in oleic acid percentage increased from 1.15 f 0.17% per cycle (HO mass selection) to 2.64 ± 0.24% per cycle (HO mass plus within half‐sib family selection). The difference in observed genetic gains was consistent with theoretical expectation. Rate of change in linolenic acid percentage, however, did not increase in magnitude, averaging −0.34 ± 0.03% (HO mass selection) and −0.24 ± 0.05% per cycle (HO mass plus within half‐sib family selection). Divergent responses to HO and LO selection for oleic and linolenic acid percentages indicated that genes governing those traits were not fixed after six cycles. Seed weight and days to maturity were correlated with selection for oleic acid, but no consistent trends throughout the selection program were noted for seed yield or oil and protein percentages. Although further progress from HO selection can be expected, a change in selection criterion to low linolenic acid percentage may be justified to minimize bias caused by genotype ✕ environment interaction.