Long-Distance Migration of Drosophila. 3. Dispersal of D. melanogaster Alleles from a Maryland Orchard

We released approximately a million marked Drosophila melanogaster individuals to measure the amount of long-distance movement of genes in a temperate area. To model natural movement as realistically as possible, we marked the flies by heterozygosis for two linked recessive alleles, released the flies as pupae and not as adults, and released them over a 2-mo period in an isolated Maryland orchard. Three months after beginning the release, we found marked flies in appreciable numbers up to 6 km from the release site, with some released alleles traveling as far as 10 km. This movement is similar to that found in our previous studies in Death Valley, California. Long-distance movement may be a general feature of D. melanogaster and not merely an artifact of handling adult flies, marking them with fluorescent dust, or releasing them in the desert. After one winter, the marker alleles could not be recovered at any location except a year-round produce stand 8.7 km from the release site. Combined with seasonal observations of the appearance and disappearance of flies at the release site, this observation implies that outdoor populations cannot survive northern winters but must be reestablished by migrants from warmer habitats. We suggest that D. melanogaster in northern North America consists of demes that are regularly connected by migration but become extinct in winter. This population structure causes considerable gene flow, which may homogenize neutral genetic variation among populations, and reduces the possibility that the well-known morphological, inversion, and allozyme clines are caused by historical accidents or genetic drift.