Genome-wide association and genomic prediction reveal the genetic architecture of dry rubber yield and stem girth in rubber tree (Hevea brasiliensis)
Abstract
The rubber tree ( Hevea brasiliensis Müll. Arg.) is the world’s primary source of natural rubber, yet genetic improvement remains challenging because of its long breeding cycle and the complex genetic architecture of economically important traits. Here, we integrated quantitative genetic analysis, genome-wide association studies (GWAS), and genomic prediction to investigate the genetic basis of dry rubber yield (DRY) and stem girth in three Thai rubber breeding populations evaluated over 6–14 years. Following quality control, 346 genotypes and 24,242 high-quality SNP markers were retained for genomic analyses. Mixed linear models revealed high entry-mean broad-sense heritability for DRY ( H 2 = 0.84–0.97) and stem girth ( H 2 = 0.85–0.97). DRY exhibited greater genetic advance (25.47–60.88% of the population mean) than stem girth (5.41–17.81%), indicating greater selection potential despite stronger temporal genotype × year interactions. Genome-wide linkage disequilibrium decayed to r 2 = 0.10 within approximately 151–217 kb. Multi-model GWAS identified five genome-wide significant loci for DRY on chromosomes 3, 5, 10, 12, and 18 and four genome-wide significant loci for stem girth on chromosomes 4, 9, 11, and 12. The strongest association was detected for stem girth on chromosome 12 at 81.0 Mb ( P = 9.68 × 10⁻ 11 ). Candidate-gene analysis and functional annotation of associated regions containing annotated genes highlighted biological processes related to signal transduction, primary metabolism, transcriptional regulation, hormone signaling, and cell wall-related processes. Genomic prediction showed trait-dependent performance, with kernel- and relationship-based models achieving the highest predictive ability for stem girth (approximately 0.60–0.65), whereas Random Forest performed best for DRY (approximately 0.55–0.60). These findings provide genomic resources and predictive approaches that may facilitate genomic-assisted selection for improved rubber yield and growth in H. brasiliensis .