Fig. 2: The twelve constructed scenarios tested with Approximate Bayesian Computation. In these scenarios, t# represents the time scale in number of generations and N# represents the effective population size during the time period (e.g., 0–t1, t1–t2).
These 12 scenarios constructed by us are the ones among a multitude of pre-tested scenarios that most closely correspond to the possible migration dynamics.
For each scenario, 100,000 simulations were run. The models were compared using the logistic regression method, and the scenario with the highest posterior probability was determined to be the most realistic one.
To obtain the calibrated years before present (BP), the number of generations has to be multiplied by a given generation time. In the present study, the generation time of Larix is set to 25 years, since larch trees can start reproducing when they reach an age of 20–30 years (Semerikov et al. 2013).