Fig. 2: D. sylvestris population structure describes genetically and geographically distinct lineages.

A Map of sampled populations. Samples are represented as coloured shapes defined according to genetic cluster, as inferred via principal component analysis (PCA) of whole-genome sequences (PCA results for all sampled individuals in inset; eigenvalues in axes’ labels). Landscape colour reflects population effective diversity rates (q) calculated via EEMS. Thick black dotted lines represent major genetic boundaries as identified via Monmonier’s algorithm on populations’ F_ST. Numbers on line are (1) French Prealps-Maritime Alps boundary, (2) Po Plain, (3) Brenner zone, Puster and Gail valleys, and (4) Adriatic Sea. Plus, minus and asterisk symbols denote individuals and populations at contact zones used in chromosome painting analysis. B Admixture proportions of whole-genome sequences at K = 3; balanced dataset. Populations are ordered (from left to right): Apennine lineage (south to north-west), Alpine linage (south-west to north-east), and Balkan lineage (north to south). C A systematic pattern in residuals (blocks of red or blue; bottom) representing the difference between the observed admixture palettes for select individuals (top) and those reconstructed by chromosome painting supports a scenario of recent bottlenecks in the populations denoted with a minus superscript, over a scenario of between-lineage admixture in the populations denoted with a plus superscript; for the Apennine-Balkan (left) and Alpine-Balkan (right) clines (assessed separately at K = 2). Populations used for this analysis correspond to the populations marked with their respective symbol in A and surrounded by black borders in B.