Résumé
Multi-locus genetic processes in subdivided populations can be complex and
difficult to interpret using theoretical population genetics models.
Genetic simulators offer a valid alternative to study multi-locus genetic
processes in arbitrarily complex scenarios. However, the use of
forward-in-time simulators in realistic scenarios involving high numbers
of individuals distributed in multiple local populations is limited by
computation time and memory requirements. These limitations increase with
the number of simulated individuals. We developed a genetic simulator,
MetaPopGen 2.0, to model multi-locus population genetic processes in
subdivided populations of arbitrarily large size. It allows for spatial
and temporal variation in demographic parameters, age structure, adult and
propagule dispersal, variable mutation rates and selection on survival and
fecundity. We developed MetaPopGen 2.0 in the R environment to facilitate
its use by non-modeler ecologists and evolutionary biologists. We
illustrate the capabilities of MetaPopGen 2.0 for studying adaptation to
water salinity in the striped red mullet Mullus surmuletus.