Integrated multi-trait selection framework for Norway spruce: Phenotypic spatial correction and multi-trait GWAS

Norway spruce is an important economic and ecological species in northern forests. However, early genetic evaluation is influenced by strong spatial heterogeneity, neighbour competition, and the polygenic, small-effect nature of key traits, together obscuring genetic signal and reducing selection efficiency. Here, we evaluated eight spatial models for growth and phenological traits across 819 individuals from 45 half-sib families. Subsequently, the corrected phenotypes were combined with multi-method selection and multi-trait genome-wide association study (GWAS) to enhance early genetic evaluation and candidate gene discovery. Tree height, diameter at breast height, new branch height (NH), and crown width (CW) were best fitted by the AR1 ×AR1 model, whereas bud burst stage and stem straightness were fitted by a block-based AR1 ×AR1 model. Under the best models, narrow-sense heritability ( h 2 ) was 0.18–0.41, with the largest increases for NH (65%) and CW (45%). The Met-Ipsq2 strategy (intersection of principal component analysis, selection index, and Qi) balanced genetic gain, and diversity at 5% intensity, and selected 41 superior individuals. Multi-trait GWAS identified 21 SNPs and annotated 13 candidate genes, covering GTL1 (a trihelix transcription factor) and pathways related to sugar metabolism and vesicle transport. Finally, we developed a seven-marker Kompetitive Allele-Specific PCR (KASP) panel that reliably separated superior from inferior trees and demonstrated a significant enrichment of favourable alleles in elite individuals. This panel serves as an efficient, low-cost screening tool for early seedling selection and provides robust support for operational genomic selection, thereby enabling a scalable breeding workflow and advancing the molecular-assisted deployment of Norway spruce.