Simple and complex dynamics in the model of evolution of two populations coupled by migration with non-overlapping generations

Purpose is to study the mechanisms leading to genetic divergence (stable genetic differences between two adjacent populations). We considered the following classical model situation. Populations are panmictic with Mendelian rules of inheritance. The action of natural selection (differences in fitness) on each of population is the same and is determined by the genotypes of only one diallel locus. We assume that adjacent generations do not overlap and genetic transformations can be described by a discrete time model. This model describes the change in the concentration of one of the alleles in each population and the ratio (weight) of first population to the total size. Methods. We used the analogue of saddle charts to construct parametric portraits showing the domains of qualitatively different dynamic modes. The study is supplemented with phase portraits, basins of attraction and bifurcation diagrams. Results. We found that the model dynamic regimes qualitatively coincide with the regimes of a similar model with continuous time, but only for a weak migration. With a strong coupling, fluctuations of the phase variables are possible. We showed that the genetic divergence is possible only with reduced fitness of heterozygotes and is the result of a series of bifurcations: pitchfork bifurcation, period doubling, or saddle-node bifurcation. After these qualitative changes, the dynamics become bi- or quadstable. In the first case, the solutions corresponding to the genetic divergence are unstable and are just a part of the transient process to monomorphic state. In the second case, the divergence is stable and appears as 2-cycle for a strong migration coupling. Conclusion. In neighboring populations, movement towards an asymptotic genetic structure (monomorphism, polymorphism or divergence) can be strictly monotonous or in the form of damped unstable or undamped stable fluctuations with a period of 2 for biologically significant parameters. For insignificant parameters, we found a complex dynamics (chaos) that consist of divergent fluctuations around fixed points and quasi-random transitions between them.