Genomic dissection of environmental responsiveness and cross-environment productivity in fenugreek

Dissecting the genetic basis of genotype × environment interaction (G × E) is important for breeding crops with stable or responsive performance across production environments. Using previously published seed-yield data from 107 Egyptian fenugreek accessions evaluated under two contrasting field environments, we investigated whether yield potential and environmental responsiveness share a common genetic basis. Seed yield per plant was measured in Giza (SYP_G; reference environment) and Wadi El-Natrun (SYP_W; contrasting environment). Nine derived indices were calculated to characterize environmental responsiveness, productivity, and stability: relative performance index (RPI), yield plasticity index (YPI), stress tolerance index (STI), log response ratio (LRR), mean productivity (MP), harmonic mean (HM), tolerance index (TOL), stress susceptibility index (SSI), and AMMI stability value (ASV). The near-zero correlation between SYP_G and SYP_W (r = 0.107) indicated substantial genotype reranking across environments, while several response indices were closely related to SYP_W and therefore predominantly reflected environment-dependent yield expression. Phenotypic analyses also identified contrasting performance strategies, ranging from highly responsive accessions, such as G15 (RPI = 3.629), to highly stable accessions, such as G41 (YPI = 0.001). These phenotypes were integrated with 38,142 SNPs using univariate association-mapping approaches. In addition, SYP_G and SYP_W were subjected to multivariate association-mapping. Association mapping identified one genome-wide-significant SNP for STI and several suggestive loci showed concordant associations across mapping methods. Within the current panel, marker coverage, and significance thresholds, most adaptation-index associations did not overlap previously reported yield loci; however, dDocent_Contig_39939_66, previously associated with seed yield in this panel, also showed cross-method suggestive associations with RPI and TOL. Comparative genomic analysis of three legume reference genomes supported two gene-linked candidate loci associated with STI: a cation/H⁺ antiporter and an importin subunit beta-1. The same analysis classified two of the strongest statistical signals, including the sole genome-wide-significant locus, as repetitive or transposable-element sequences rather than single-copy genes. These findings demonstrate that statistical association strength does not necessarily establish a coding-gene basis for loci identified using reference-free ddRAD markers. Within the current population size and marker coverage, environmental responsiveness and productivity appeared partially distinct but biologically connected components of environment-dependent yield performance. The results provide a genomic and phenotypic framework for prioritizing fenugreek germplasm for further evaluation while emphasizing the need for independent-population, multi-season, and functional validation of the identified candidate loci.