Genetic architecture of dual-purpose maize integrating grain yield and stover quality across environments via multi-trait GWAS and genomic prediction

Abstract

Dual-purpose maize breeding aims to concurrently enhance grain productivity and stover value for fodder or bioenergy, a challenge due to the complex and potentially correlated genetic control of these traits. We conducted an integrated genetic analysis using multi-trait genome-wide association studies (GWAS), a selection index, and genomic prediction in 300 diverse maize inbred lines evaluated across multiple environments within a single year. Substantial phenotypic variation was observed for grain yield, stover yield, and stover quality traits. Broad-sense heritability was moderate for yield traits (0.4-0.5) and high for stover fiber composition and digestibility (0.6-0.8), indicating strong genetic determinism. Grain and stover yields exhibited a favorable positive genetic correlation, while correlations between grain yield and stover quality were weak or non-significant, suggesting minimal biological trade-offs. Multi-trait GWAS identified several significant QTL, including pleiotropic loci influencing both grain and stover performance, and candidate genes associated with lignin biosynthesis affecting fiber digestibility. Genomic prediction using GBLUP achieved moderate accuracy, with modest improvements from multi-trait models. A Dual-Purpose Index effectively identified superior genotypes, with top-ranked lines showing simultaneous gains in grain and digestible stover yield.