Abstract
Explanations of dispersal mechanisms in mammals that have stressed the importance of aggression by dominant (?) individuals as the immediate cause of the dispersal of less aggressive (more subordinate ?) individuals are insufficient for explaining recent data collected on a variety of mammals. In fact, avoidance of social interaction at the time of dispersal is more characteristic of some species in which individuals emigrate. Studies that have investigated genetic correlates of dispersal in rodent populations that undergo regular cycles are few and have not provided any causative explanations. In various canids and rodents, behavioral interactions at the time of dispersal do not appear to provide the necessary stimuli for dispersal. Knowledge of the behavioral interactions that occur before dispersal may provide a key to understanding both interspecific and intraspecific differences in social organization and dispersal patterns. Individuals who have the most difficulty interacting with littermates will not develop strong social ties with their siblings and will be the most likely individuals to disperse of their own accord. This hypothesis is testable by collecting data on the social interaction patterns of individual littermites throughout early development and during dispersal. In this way, the importance of a previously neglected factor in dispersal.sbd.namely, the relationship between the behavioral antecedents of dispersal and which individuals disperse at what age, and in what manner.sbd.can be assessed.