A simulation-based approach to evaluating population structure in non-equilibrial populations

Résumé

The standard null model of panmixia used to test for population subdivision is based on a set of assumptions that can be violated given recent events likely to result in non-equilibrial genetic composition coupled with the complex life histories of many species. Bowhead whales (Balaena mysticetus) represent such a species. Bowhead whales also have a well-documented history of severe commercial harvest in the recent past which would be expected to leave a population out of genetic equilibrium. They also have a very long life span, overlapping generations, and age and sex-structured migrations. In addition, samples come from whales killed in a hunt known to be non-random with respect to size at different whaling villages. Sampling of such a population could lead to erroneous conclusions regarding population structure, which could have real consequences for aboriginal whaling. To better interpret the results of standard population genetic analyses, an individual-based model of bowhead whale population dynamics and genetics was created using the R package rmetasim. The model re-created as closely as possible all aspects of the demography, genetics, and whaling history of bowhead whales. Simulated datasets were generated by sampling from the simulated population in a way that matched the age, sex and geographic distribution of empirically collected samples. The empirical bowhead datasets were compared to null distributions generated from the simulated datasets for a variety of genetic analyses. The analysis indicates that the empirical genetic data sampled from the Bering-ChukchiBeaufort (BCB) stock of bowhead whales are more consistent with the model of a population with the same whaling history and demographic composition as BCB whales than they are with a single, randomly-mating population in genetic equilibrium under a standard Wright-Fisher model. Additionally, it was demonstrated that by failing to account for the unique features of the population dynamics of the species, standard tests of genetic differentiation based on panmixia may produce misleading results. The approach outlined will likely prove useful for evaluating population structure in other species likely to be out of equilibrium.