Abstract
Although roads are widely seen as dispersal barriers, their genetic consequences for animals that experience large fluctuations in population density are poorly documented. We developed a spatially paired experimental design to assess the genetic impacts of roads on cyclic voles (Microtus arvalis) during a high-density phase in North-Western Spain. We compared genetic patterns from 15 paired plots bisected by three different barrier types, using linear mixed models and computing effect sizes to assess the importance of each type, and the influence of road features like width or the age of the infrastructure. Evidence of effects by roads on genetic diversity and differentiation were lacking. We speculate that the recurrent (each 3-5 generations) episodes of massive dispersal associated with population density peaks can homogenize populations and mitigate the possible genetic impact of landscape fragmentation by roads. This study highlights the importance of developing spatially replicated experimental designs that allow us to consider the large natural spatial variation in genetic parameters. More generally, these results contribute to our understanding of the not well explored effects of habitat fragmentation on dispersal in species showing "boom-bust" dynamics.