DISPERSAL AND GENETIC STRUCTURE IN KANGAROO RATS.

We used spatial autocorrelation of allele frequencies to examine local structure in a population of bannertailed kangaroo rats for which Wright's isolation-by-distance model seems applicable, and for which we can estimate neighborhood size based on 10 years of data on demography and dispersal. The uniform dispersion and strong philopatric tendencies of this species provide a test case for the idea that restricted dispersal can lead to local genetic structure in small mammals. Whether we considered such complications as nonnormal dispersal distances, variation in lifetime reproductive success, fluctuating population density, and adult as well as juvenile dispersal, our estimate of effective population size was fewer than 15 animals. Nevertheless, data from four polymorphic allozyme loci analyzed over a range of separations between 50 m (approximately one home range diameter) and 1,000 m detected no evidence for spatial clustering of alleles. One resolution of this apparent paradox is that "gamete dispersal," caused by the movements of males away from their residences during the breeding season, may be a significant (and unmeasured) component of gene dispersal. Our analyses also demonstrate that a decline in population density may actually increase neighborhood size. A more general implication is that even extremely philopatric mammals have effective population sizes large enough to prevent the development of local genetic structure.