Context‐Specific Marker‐Assisted Selection for Improved Grain Yield in Elite Soybean Populations

Abstract

Despite the importance of grain yield potential to plant breeders and society in general, it has been difficult to identify grain yield quantitative trait loci (QTL) effective for marker‐assisted selection (MAS) across a wide range of genetic and/or environmental contexts. However, as genotyping becomes more cost effective, it might be feasible to use preliminary yield trials to model a target genotype within each context and immediately select the progeny that approach that target genotype in real time. In the present study, elite soybean cultivars with residual heterogeneity were leveraged as populations (the genetic context) to detect yield QTL within a limited set of environments (the environmental context), to model a target genotype, and to select subline haplotypes that comprised the target genotype. The yield potential of the selected subline haplotypes were then compared to their respective mother lines in highly replicated yield trials across multiple environments and years. Statistically significant yield gains of up to 5.8% were confirmed in some of the selected sublines, and two of the improved sublines were released as improved cultivars. This context‐specific MAS (CSM) approach might also be applicable to the more typical biparental and backcross populations commonly used in plant breeding programs. Factors that can affect the efficiency and applicability of CSM are discussed.