Maladaptation in a Marginal Population of the Mosquito Fish,gambusia Affinis

Field biologists commonly assume that the organisms with which they deal are well adapted-even optimally adaptedto local circumstances. Both Ehrlich and Raven (1969) and Endler (1977) have de-emphasized the role of gene flow in preventing large scale geographic differentiation and local adaptation. This paper documents a case in which gene flow may have prevented small scale local adaptation in one population of mosquito fish, Gambusia affinis. It carries two messages: field-workers should check the assumption that their study organisms are adapted to the local environment because that assumption does not always hold, and there are limiting cases involving high dispersal rates over short distances in which gene flow can overwhelm local selection pressures. Evidence that local populations are being maintained away from the locally optimum phenotype is rare. Camin and Ehrlich (1958) analyzed the banding patterns of water snakes (Natrix sipedon) on islands in Lake Erie. They showed that on the islands, where snakes were probably confined to a band of sandy substrate near shore by lack of water inland, patternless, light-colored snakes were most common. In contrast, on the mainland where water snakes inhabit wooded streams with darker, more complex substrates, richly patterned, dark-colored snakes predominated. However, a small proportion of the island snakes were patterned, and therefore maladapted to the local environment, an observation that