chickpea (Cicer arietinum L.) cultivation from spring to autumn is a promising strategy 7 to escape terminal drought and maximize yield potential in semi-arid regions; however, success 8 depends on identifying genotypes with suitable phenology and yield performance. This study 9 evaluated 332 desi chickpea genotypes (330 landraces, two checks) over two seasons (2021–2022) 10 under autumn-sown rainfed conditions in western Iran, using Linear Mixed Models (LMM), 11 Genotype × Yield × Trait (GYT) biplots, Structural Equation Modeling (SEM), and the Multi-Trait 12 Genotype-Ideotype Distance Index (MGIDI). Results revealed trait-dependent genetic variation, 13 with heritability highest for hundred-seed weight (H² = 94.9%) and moderate for plant height (H²= 14 40.6%) and days to flowering (H²= 35.1%), while seed yield showed no detectable genotypic 15 variance under the single-replicate design. No genotype excelled across all GYT combinations, 16 though G313 ranked highest for yield × hundred-seed weight in both years, a promising parental 17 candidate. SEM identified plant height as the dominant trait underlying agronomic performance in 18 both seasons, with a direct positive association with seed yield emerging in 2022, while the top 19 genotype for the yield–height combination shifted between years (G33 in 2021, G49 in 2022). 20 Integrating both seasons, MGIDI ranked 114 landraces above the better commercial check, with 21 G5 first overall, confirming its identification as a GYT winner. Plant height should be one 22 component of a broader selection strategy rather than a standalone target, given genotype × year 23 variability and lodging risk. The identified superior landraces (G5, G313) offer valuable genetic 24 resources for breeding climate-resilient chickpea cultivars.
