Figure 5. A. Barplots represent the proportion of observed jSFS, at the corresponding average number of SNPs in the jSFS, that are classified as a given model, which is indicated by the color of the barplot. Solid barplots (left side) represent Tsuga heterophyllapredictions and diagonal stripped barplots (right side) represent predictions for Thuja plicata. B. Table indicating the average number of model votes for the most selected model, “AV + sc + bot/exp”, for both species, along with the average estimated posterior probability (PP) for the same model. C. Corresponding observed jSFS at each SNP count (10000, 3000, and 1000) for Tsuga heterophylla(top row) and Thuja plicata (bottom row).
Table 1. The average number of unlinked SNPs used in the 100 empirical datasets, with the corresponding number of samples from the coastal and inland populations, where each sample represents an allele for an individual, most often both alleles are included, but in some cases only one allele from an individual is included in the construction of the jSFS. The error rate for all models represents the average error rate for all model classifications for the classifier constructed with the corresponding data size. The error rate with RD collapsed corresponds to the overall error rate for the classifier when the four Recent Dispersal models are collapsed into a single model, RD.