Skeleton-guided 3D digitization standardizes complex trait phenotyping and supports reproducible locus discovery in...

Accurate and standardized phenotyping of complex, environmentally sensitive quantitative traits remains a major bottleneck for reliable locus discovery and breeding applications. Here, we established a skeleton-guided 3D digital phenotyping framework that generates standardized digital replicas and enables precise quantification of fruit and plant architecture traits in cucumber. The workflow was applied to a permanent recombinant inbred line (RIL) population (n = 211) evaluated across two seasons (2023-2024), from which nine traits were extracted from 3D models. All 211 RILs were whole-genome resequenced to generate genome-wide SNPs, enabling construction of a high-density linkage map and subsequent QTL mapping, complemented by GWAS for physical anchoring of association signals. Using this integrated design, we identified 29 QTLs across the nine traits and resolved cross-season major-effect loci with consistent genetic signals. Notably, two cross-season loci were detected as novel: FL4.1/FSL4.1 affecting fruit length and fruit stalk length, and NLB1.1/LLB1.1 affecting branching. GWAS further anchored lead variants to physical coordinates and supported cross-season associations. Together, these results demonstrate that standardized 3D phenotyping provides a reproducible and interoperable trait definition framework that supports cross-season locus discovery and downstream marker development for quantitative genetic dissection in cucumber.