The genetic basis of dispersal in a vertebrate metapopulation

Dispersal affects evolutionary processes by changing population sizes and their genetic composition, influencing the viability and persistence of populations. Investigating which mechanisms underlie variation in dispersal phenotypes and whether populations harbor adaptive potential for dispersal is crucial to understanding the eco-evolutionary dynamics of this important trait. Here, we investigate the genetic architecture of dispersal in an insular metapopulation of house sparrows. We use an extensive long-term individual-based ecological dataset and high-density single nucleotide polymorphism (SNP) genotypes for over 2500 individuals. We conducted a genome-wide association study (GWAS), finding a relationship between dispersal probability and a SNP located near genes known to regulate circadian rhythmic, glycogenesis and exercise performance, among other functions. However, this SNP only explained 3.8% of variance, suggesting that dispersal is a polygenic trait. We then used an animal model to estimate heritable genetic variation (Va), which composes 10% of the total overall variation in dispersal probability. Finally, we investigated differences in Va across populations occupying ecologically relevant habitat types (farm vs. non-farm) using a genetic-groups animal model. We found higher mean breeding value, Va, and heritability for the farm habitat, suggesting different adaptive potentials across habitats. Moreover, dispersal phenotypes may depend on genotype-by-environment interactions. Our results suggest a complex genetic architecture of dispersal, and demonstrate that adaptive potential may be environment-dependent in key eco-evolutionary traits. The eco-evolutionary implications of such environment-dependence and consequent spatial variation are likely to be ever more important with the increased fragmentation and loss of suitable habitats for many natural populations.