A retrospective analysis of historical data of multi-environment trials for dry bean (Phaseolus vulgaris L.) in Michigan

Plant breeding programs use multi-environment trials (METs) to test multiple genotypes with the aim of selecting superior lines and increasing genetic gain. However, METs can be affected by the genotype x environment interactions (GEI) which hinders the ability to identify stable and high-yielding dry bean varieties. This study investigates the impact of GEI on grain yield (GY) in three dry bean market classes across multiple environments in Michigan. The goal is to identify genotypes with superior performance under GEI effects and describe the genetic correlation between locations and other agronomic traits. Furthermore, we evaluate genetic yield trends on three major dry bean market classes in Michigan. Using linear mixed models, stability analysis, and a multi-trait index, this study accounted for the effects of the environment on genotype performance to improve the prediction of GY. Our results highlight the importance of considering GEI in dry bean breeding programs to enhance the selection accuracy of GY. The genotype by environment interaction biplot analysis and the GEI model aided in identifying and classifying mega-environments, leading to specific strategies for testing commercial bean classes across target production environments. Multiple-trait performance selection showed positive results, although limited gains were observed for early maturing genotypes. Finally, the analysis of genetic trends reveals a positive trend in yield for black beans, while navy beans and small red beans show a slight decrease which can be attributed to unfavorable growing conditions and lack of genetic diversity in these market classes.