QTL dissection of yield-related traits in foxtail millet using cross-ecotype RIL

The grain yield of foxtail millet is strongly shaped by sowing-season environments and ecotype adaptation, however the genetic basis of yield components across contrasting regimes remains incompletely resolved. Here, we developed a recombinant inbred line (RIL) population derived from a hybrid between the spring-sowing cultivar Jingu 21 and the summer-sowing cultivar Chuang29, and evaluated eight yield-related traits across eight environments during 2022–2023. Composite interval mapping was performed separately in each environment and 112 QTL signals were identified, which were consolidated into 74 non-redundant QTLs. Among these, 24 were detected in at least two environments, and 47 showed no overlap with previously reported QTL intervals compiled from the literature. QTL clustering revealed multiple co-localized regions, including a prominent chromosome-9 hotspot affecting seven traits and detected 17 times across trait–environment combinations. Homology-informed mining within refined intervals prioritized 27 candidate genes. Haplotype analysis with an independent natural population ( n =1,844), identified significant associations for three genes— Seita.3G233000 (panicle length), Seita.4G281800 (thousand-grain weight) and Seita.5G068000 (panicle diameter). Analysis of additive effects found different contributions from the two parents, with Jingu21 contributing more TGW-increasing signals and Chuang29 contributing more PW/GWP-increasing signals across environments, which established the foundation pyramiding those super allele into an elite foxtail millet cultivar. Together, these results provide a cross-environment genetic framework for foxtail millet. They also offer actionable loci for marker-assisted improvement and gene discovery for yield-related traits.