Fig. 2
Demographic histories of Taxus baccata populations inferred with the Approximate Bayesian Computing approach based on coalescent simulations. Panel A—the results of the Euclidean distance-based UPGMA clustering of study populations based on the log-estimates of demographic parameters. Panel B—the relationship between the coalescent-ABC estimate of mutation-scaled current effective size (N0μ) and the estimate of effective population size obtained as a harmonic mean across linkage disequilibrium (LD) and sibship frequency-based (SF) estimators. Panel C—scale and time of demographic change inferred with the coalescent-ABC approach for the main UPGMA clusters. Scales of axes (effective size and number of generations) obtained after re-scaling the estimated coalescent parameters (N0μ, NAμ, and tμ) using the mutation rate estimated as a slope of regression line shown in the panel B