Differentiation of Island Populations: A Laboratory Model

Publisher Website
Google Scholar
Abstract
The experimental model, using flour beetles (Tribolium), was designed to mimic in the laboratory the genetic processes of differentiation of island populations. Eight "archipelagos," each composed of 18 "islands," were maintained for 11 generations. Each island was founded by one pair of adults such that F1 offspring were all heterozygotes at a marker locus. A randomly selected pair of sibs continued each island population from one generation to the next, allowing no interisland migrations in six archipelagos. In two archipelagos, interisland migration was enforced (OV). Replicate archipelagos were held in two environments: a constant, optimal one (three replicates, labeled IHC) and a randomly variable one (three replicates, labeled IHV). The two OV replicates were maintained in the randomly variable environment. The genetic processes of island differentiation were studied by partitioning genetic diversity into its components: within islands, among islands within archipelagos, and among archipelagos. Replicate archipelagos (founded by identical combinations of males and females on each island) were similar in response: diversity among archipelagos was rarely larger than 1%. Diversity among islands within archipelagos rapidly increased in IHC and IHV and reached about 60% of total diversity in about five generations. When, at this point, OV archipelagos branched off from IHC by imposed interisland migration, diversity among islands rapidly decreased to about half its previous value. Similar calculations were made for eight sets of published data about real-island populations of Drosophila, lizards, and mammals. The populations of the experimental "islands" reached higher levels of differentiation than six of the real-island populations. The two sets of data for the Galapagos lizards turned out to be more highly differentiated, possibly because some island populations of these lizards are taxonomically considered differnet species or subspecies (J. Wright 1984). From this study, it seems that the mating system and genetic bottlenecks were most important to the differentiation of island populations and that environmental variation had little effect. With refinement, a model of this sort may be used to investigate processes of population differentiation on islands, which cannot be observed directly in nature.