Characterizing and visualizing gene-flow in Coastal Douglas-Fir (Pseudotsuga menziesii (Mirb.) Franco var. menziesii): A landscape genetics approach

Background/Question/Methods

Genetic variation, structured spatially, within and between organisms is the result of interacting environment and genotype. Dispersal mechanism, landscape composition and connectivity contribute to population level genetic structure in plants. Landscape genetics is a broad framework that assesses how landscape composition, configuration and matrix quality impact an organism’s genetic variation, allowing correlative relationships to be identified. To improve our understanding of how landscape features influence gene-flow in a long-lived tree species, Coastal Douglas-Fir (Pseudotsuga menziesii (Mirb.) Franco var. menziesii) in the Pacific Northwest, we used genetic visualization, and landscape genetics analysis. Six microsatellite and 25 allozyme loci genotyped in 1363 P. menziesii samples were used to quantify levels of gene-flow. We assigned ancestry probabilities, generated in STRUCTURE, to individual samples and population genetic structure was inferred. F_stvalues were used to assess pairwise relatedness and a Mantel’s test was used to test Isolation by Distance (IBD). We tested the hypothesis that geographic features limit gene-flow across a large geographic area. We employed Euclidian distance analysis methods to test and visualize gene-flow across the study region using ArcGIS.

Results/Conclusions

11 subpopulations of P menziesii were identified across a geographically wide and diverse ecosystem. Isolation by Distance was identified as only a small factor (r = 0.05984 p < 0.001) in structuring population genetic variation. Barriers to gene-flow were identified as additional factors contributing to the observed genetic variation. Pairwise Euclidean kinship values suggest that gene-flow is restricted (F_st > 0.035) along the Coastal Mountain Range and across the high peaks of the Cascade Mountain Range. Additionally, gene-flow is restricted between the Siskiyou Mountains and the southern Cascades, which are geographically close. Moderate levels of genetic differentiation (F_st = 0.025-0.035) were observed across Puget Sound and latitudinally along the east side of the Coastal Mountain Range. Low levels of differentiation (F_st < 0.025) were observed across the Willamette Valley and along the Cascades. We found that gene-flow occurs at sufficiently high rates over most of the study area explaining the very low population genetic differentiation.