Dispersal Behavior and Evolutionary Metapopulation Dynamics Michael c. Singer and ilkka hanski

Abstract “Dispersal is the glue that keeps local populations together in a metapopulation” (Hansson 1991). Seen from the perspective of population biology, dispersal is also another kind of glue, a behavioral trait that couples ecological metapopulation dynamics with evolutionary dynamics. Indeed, we might describe our work on this topic as a study in either behavioral or evolutionary ecology, were it not that “behavioral ecology” is not regarded as proper ecology by ecologists, and “evolutionary ecology” is not really regarded as study of evolution by evolutionary biologists. Sadly, these researchers are usually correct; work that claims to be interdisciplinary at the interfaces between behavior, ecology, and evolution rarely achieves its aim. Yet metapopulation biology absolutely requires a truly behavioral understanding of the decision making that underlies dispersal, a truly ecological understanding of the interplay between dispersal and population dynamics, and a truly evolutionary understanding of the ways that dispersal responds to spatially and temporally varying environmental conditions. Studies of checkerspots have made substantial progress toward achieving this desirable combination of knowledge, and reciprocal interactions between the behavioral, ecological, and evolutionary facets of dispersal are becoming clear. For example, the influence of host plant preference (behavior) on dispersal drives population turnover (ecology) via biased immigration and colonization rates (Hanski and Singer 2001), while biased colonization rate, in turn, contributes to the evolution of host plant preference (Hanski and Heino 2003), bringing us back to the influence of preference on dispersal. Before reaching this level of complexity, however, we begin with a discussion of the classic approaches to the study of insect dispersal and how they have been applied in studies of butterflies in general and checkerspots in particular.