Avoiding inbreeding is considered a key driver of dispersal evolution, and dispersal distances should be especially important in mediating inbreeding risk because the likelihood of mating with relatives decreases with dispersal distance. However, a lack of direct data on dispersal distances has limited empirical tests of this prediction, particularly in the context of the multiple selective forces that can influence dispersal. Using a headwater salamander system, we tested whether spatial variation in environmental conditions leads to differences in dispersal distances, resulting in spatial variation in the effect of dispersal on inbreeding risk. Using capture-recapture and population genomic data from 5 streams, we found that dispersal distances were greater in downstream reaches than upstream reaches. Inbreeding risk was lower for dispersers than non-dispersers in downstream reaches, but not in upstream reaches. Furthermore, stream reaches did not differ in spatial patterns of individual relatedness, indicating that variation in inbreeding risk was in fact due to differences in dispersal distances. These results demonstrate that environmentally associated variation in dispersal distances can cause the inbreeding consequences of dispersal to vary at fine spatial scales. They also show that selective pressures other than inbreeding avoidance maintain phenotypic variation in dispersal, underscoring the importance of addressing alternative hypotheses in dispersal research.

Dispersal data: collected using capture-mark-recapture methods over 4 summer field seasons in 5 headwater streams; xy-coordinate data: x values are 0, y-values represent longitudinal length along the stream; genotypic data: generated from ddRADSeq approach, described in manuscript methods.