Factors affecting the accuracy of genomic selection for growth and wood quality traits in an advanced-breeding population of black spruce ( Picea mariana )

摘要: Genomic selection (GS) uses information from genomic signatures consisting of thousands genetic markers to predict complex traits. As such, GS represents a promising approach accelerate tree breeding, which is especially relevant for the improvement boreal conifers characterized by long breeding cycles. In present study, we tested in an advanced-breeding population black spruce (Picea mariana [Mill.] BSP) growth and wood quality traits, concurrently examined factors affecting model accuracy. The study relied on 734 25-year-old trees belonging 34 full-sib families derived 27 parents that were established two contrasting sites. profiles obtained 4993 Single Nucleotide Polymorphisms (SNPs) representative as many gene loci distributed among 12 linkage groups common spruce. models four Validation using independent sets showed accuracy was high, related trait heritability equivalent conventional pedigree-based models. forward selection, gains per unit time three times higher with than selection. addition, also accurate across sites, indicating little genotype-by-environment interaction area investigated. Using half-sibs instead full-sibs led significant reduction accuracy, inclusion relatedness contributed its accuracies. About 500 1000 sufficient obtain almost all markers, whether they well spread genome or single group, further confirming implication potential long-range disequilibrium (LD) high estimates obtained. Only slightly when marker subsets identified carry large effects, minor role short-range LD this population. This supports integration advanced-generation programs, given prediction relatively small number due family structure programs similar ones cycles, much larger gain can be at early age approach. thus appears highly profitable, context species are amenable mass vegetative propagation selected stock, such spruces.