Figure 1 Supporting information. Since we did not have information about the local ancestry patterns for all 146 genotypes with phenotypic and gene expression data (in RASPberry we only included 36 P. trichocarpa admixed individuals), we used the phased introgressed region of chromosome 15 from all P. trichocarpa genotypes (146) as well as from the reference panel of individuals and implemented a NJ tree analysis - 1000 bootstrap replicates in MEGA (Tamura et al. 2007). These NJ trees, based on either the entire 880-kb chromosome 15 region B or a smaller 300-kb portion at the start of the B region, revealed a well-defined cluster (bootstrap values: 80 and 99 based on 880-kb and 300 kb regions respectively) where haplotypes from pure P. balsamifera individuals grouped with haplotypes from admixed P. trichocarpa individuals. Based on these trees we identified eight genotypes homozygous for P. balsamifera haplotypes (bb: both haplotypes of the individual were found inside the P. balsamifera cluster), 32 heterozygous genotypes (bt: only one haplotype was found inside the P. balsamifera cluster) and 104 genotypes homozygous for P. trichocarpa haplotypes (tt: none of the haplotypes were inside the P. balsamifera cluster). Figure 5 Supporting information. Neighbor-Joining (NJ) tree of parental linages based on the first 880 kb of chromosome 15. 25 P. trichocarpa accessions are shown in red and 25 P. balsamifera are shown in blue and purple. Branch lengths represent genetic distances using the p-distance method. For P. balsamifera, blue represents haplotypes from populations in the northwestern parts of the range and purple represents haplotypes from central populations. The NJ analysis was conducted in MEGA6.