Yield environment changes the ranking of soybean genotypes

Abstract

Context

Crossover interactions can hinder the identification of superior genotypes and the selection of evaluation sites. Identifying yield environments where frequent crossover interactions occur will help determine when narrowly or broadly adapted genotypes might excel or fail. This information can aid in targeted breeding and selection strategies.

Objective

This study aimed to characterize the genotypic variability in yield stability and its relation to yield, and to evaluate how changes in crossover interactions vary across yield environment among soybean maturity groups (MGs).

Methods

We studied 102 soybean genotypes spanning MGs 0.0–4.9 from 218 multi-environment trials during 2022 and 2023 in the United States (US). Genotypes were grouped in nine clusters based on MG adaptation zones for the US soybean production area. Yield stability was measured as the slope of regression in a reaction norm model. Genotype ranking consistency was estimated as the Spearman correlation between two consecutive yield environments.

Results

Yield stability ranged between 0.793 and 1.181 across the dataset. The relationship between genotype stability and yield was not significant across MG clusters. All clusters showed higher ranking consistency in the highest and lowest yield environment that they explored. However, the magnitude of crossover interactions and the specific yield environments where these interactions frequently occur varied across the MG clusters. Finally, genotype choice in low- and high-yielding environments can result in yield penalties up to 516 and 710 kg ha⁻¹, respectively. Genotype choice in yield environments with frequent crossover interactions resulted in the lowest yield penalty.

Conclusion

The absence of a trade-off between genotype stability and yield suggests that high-yielding soybean genotypes can be achieved through different breeding strategies. Crossover interactions were minimal in lower and higher-yielding environments among all MG clusters. The yield environment with frequent crossover and its magnitude varied among clusters. Genotype choice to minimize yield losses can vary across MG clusters and yield environments.

Implications

Identifying the yield environment with frequent crossovers can guide efficient resource allocation by selecting optimal test sites and prioritizing efforts in strategic breeding pipelines.