Figure 3
The four population divergence scenarios tested by the Approximate Bayesian Computation (ABC) approach using each target population (Canada, United States, Kenya, Indonesia, Lebanon, Japan and Australia). Since an invasive population generally starts with a few individuals, an initial size reduction (noted here ‘N pop X b’ for each target population and ‘N EU b’ for population Europe) was added. Times are not to scale, 0 indicates the present day and t on top of the bar indicates the oldest split, bd illustrates the duration of the bottleneck period. A ghost population is a possible unsampled source population.
