Abstract
Finger millet productivity is strongly influenced by genotype × environment interaction (GEI), which complicates the identification of high-yielding and stable genotypes. This study evaluated 35 genetically diverse finger millet genotypes across three agro-ecological zones viz., Odisha (E1), Jharkhand (E2), and Bihar (E3) during two rabi seasons (2023-24 and 2024-25). A randomized block design with three replications was implemented and key quantitative traits i.e. grain yield per plant, 1000-grain weight, and number of fingers on the main ear were recorded. AMMI and GGE biplot analyses were applied to assess GEI, stability, and adaptability. Genotype G18 (VR-1176) consistently emerged as the most stable and high-yielding across environments, followed by G13 (VL-Mandia-352), G28 (Bada Mandia), G3 (PR-1639), G25 (Bada Kumnda), G26 (Badatara), G11 (VR-1223), G15 (VR-12-38), G14 (OEB-610), and G33 (FEZN-84). AMMI 1 and AMMI 2 biplots confirmed these findings, highlighting G18 and G15 as superior performers. Among sites, Jharkhand (E2) was identified as the most favourable environment. Additionally, molecular profiling using UGEP markers 46, 66, and 68 revealed polymorphic banding in high-yielding genotypes, which validates phenotypic observations. The integration of phenotypic and molecular analyses provides a robust framework for identifying finger millet genotypes with both high productivity and yield stability, supporting their recommendation for breeding programs and wider cultivation.