Abstract
Pinus radiata D.Don is one of the most widely planted exotic conifers. It is also a threatened species because native populations are small, disjunct and challenged by pests and pathogens, deforestation and climate maladaptation. Genomic tools can both enhance genetic improvement in operational breeding programmes, and support conservation efforts. Using PacBio long-read sequencing we assembled 20.6 Gbp of the large and complex P. radiata genome into 305,330 scaffolds, achieving a scaffold N50 of 196.22 kbp, which corresponds to 89% of its estimated genome size. Gene annotation, based on transcriptome data with a 97.9% BUSCO score, yielded 86,039 gene models. Linkage maps were used to anchor 7,952 contigs totalling 1.79 Gbp (approximately 9% of the assembly) across 12 pseudomolecules, which included c. 26% of the predicted genes. Genome re-sequencing (5.2x) of 40 trees, from four native populations and a major Australasian breeding population, uncovered c. 608.3 M SNPs which were used for population genomic analyses. A key finding of these analyses was the slower-than-expected decay of linkage disequilibrium (i.e., r2 > 0.2 up to 30 kb for SNPs with minor allele frequencies ≥ 0.10), suggesting recent drastic reductions of effective population size. Our findings indicate that genomic prediction could use fewer markers than the ca. 30k that are currently employed. Additionally, this study highlights the potential for Southern Hemisphere breeding programmes as ex situ conservation resources and established a foundation for functional characterisation of the P. radiata genome.