Sex-Specific Dispersal Patterns in Two Human Populations of Highland New Guinea

Abstract
We present a model of the factors underlying male-female differences in natal dispersal, along with maximum-likelihood procedures for estimation and testing. Special attention is given the effects of geographic distance, population density, and endemicity (philopatry) on dispersal, although the model can be extended to include any number of additional predictor variables. The model is applied to data collected among the Gainj and Kalam, 2 tribal populations of highland Papua, New Guinea. The analyses suggest that males in these populations are less likely to disperse than females, but more likely to move to more-distant locations when they do disperse. The dispersal of both sexes is negatively density dependent (i.e., individuals move away from areas of higher density before they reproduce), with density affecting both sexes to about the same degree. Extension of the model to include linguistic differences among locations shows that male dispersal is more constrained by language than is female dispersal.