Selection, patches and genetic variation: A cellular automaton modellingDrosophila populations

Summary

Many models have been proposed in which environmental heterogeneity promotes genetic diversity. Such models describe the situation where different phenotypes have different fitness values in different types of patch and are the genetic equivalent of the traditional resource partitioning models in ecology which allow the coexistence of species. Here we construct a different type of cellular model in which polymorphisms in populations ofDrosophila can be maintained without traditional resource partitioning. Parameter values taken from laboratory and field observations represent fungal breedingDrosophila. Some stochasticity is used in the description of the migration between patches. In the model space is divided into a uniform matrix of cells each of which has the potential to contain an ephemeral resource item (fungal fruiting body). Square arenas of up to 400 cells were used. Genotypes arrive at a fresh site, breed (Hardy-Weinberg equilibrium) and lay eggs. The eggs hatch and the larvae compete using the Hassell-Comins competition equations, as if they were three different species. Adult emergents all migrate to an adjacent cell. The aggregation patterns observed in nature are produced using an ‘attraction probability’ where each fly has a chance of moving to the currently most densely populated adjacent patch. This ‘black box’ description of migration produces distribution patterns which are indistinguishable from those seen in wild populations of fungal breedingDrosophila. Results show that the ‘attraction probability’ is the key factor in the maintenance of polymorphism and that even when the competitive advantage of the superior genotype is very great, polymorphisms can be maintained.