Table S1. Population codes (Site #), number ofindividuals sampled (N), elevation, geographic coordinates,geographic distance to the nearest heterospecific population, andposterior probability (P) of being the ancestral locationfrom the BEAST analyses for samples included in this study.

Table S2. Summary statistics for 11 sequencedloci: the number of sequences obtained (N), length offragment (L), segregating sites (S), polymorphismsunique to C. xantiana ssp. xantiana(Px), polymorphisms unique to C. xantianassp. parviflora (Pp), shared polymorphisms(Ps), average posterior probability of phasedhaplotypes (Phase) and nucleotide substitution model(NST).

Table S3. Summary statistics for fourmicrosatellite loci: percent successful amplification, number ofalleles, allele size ranges, observed heterozygosity(Ho), and inbreeding coefficients(Fis).

Table S4. Uniform priors scaled by locus lengthfor Approximate Bayesian analyses. θxan andθpar describe Watterson’s θfor Clarkia xantiana ssp. xantiana andssp. parviflora respectively. τM isthe time that migration starts or stops, τDis the divergence time, and the asymmetric migration rates Mxan→par andMpar→xan between taxa.

Table S5. Results of cline model selection fromCfit7 for hybrid indices from four different molecular datasets.Negative centers indicate a shift from the location of contactbetween the two taxa in the direction of where xantiana populations are found. Best fitting models are in bold.

mec5715_sm_TableS1-S5.docx128KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.