Supplementary data
Figure S1: The ML-graph inferred by TreeMix with 16 migration
events based on the COMBINED dataset. The arrow indicates the two
significant migration events with the highest migration weight. Drift
parameter is shown on the x-axis.
Figure S2: ML tree with rootstrap support values for
root inference. Values were estimated with IQ-TREE 2 on consensus
sequences per population (2867 bp), using a non-reversible model (model
12.12, all sites, 10000 bootstrap replicates).
Figure S3. Functional enrichment of GOs for proteins with
outliers, according to A) GBS and B) RADSeq dataset.
Figure S4 . PCA of the 18 outliers detected in the COMBINED
dataset. All but one indicated a clear clustering by archipelago. Black
dots correspond to the most frequent allele, while Red dots indicate the
heterozygotic genotype.
Table S1: Individual sample metadata.
Table S2: GBS read summary.
Table S3: Genetic diversity estimates for individual datasets
(GBS and RADSeq).
Table S4: List of outliers identified by both Bayescan and
PCADAPT (Fst>0.8, q<0.01,) for the three
datasets: COMBINED (N=18), GBS (222) and RADSEQ (190) and corresponding
annotation.
Table S5: Archipelago-segregating SNPs (fixed alleles).