Spatial autocorrelation of the environment influences the patterns and genetics of local adaptation

Abstract

Environmental heterogeneity can lead to spatially varying selection, which can, in turn, lead to local adaptation. Long-held views in evolutionary biology predict that environmental heterogeneity that is spatially autocorrelated will lead to stronger local adaptation than if it is not. Here I test this explicitly, combining analysis of a dataset from a large-scale provenance trial conducted on lodgepole pine with simulations modelling local adaptation. I find evidence that environmental heterogeneity that is spatially autocorrelated leads to stronger local adaptation in natural populations. Using simulations, I examine how the genetics of local adaptation are influenced by the spatial pattern of environmental heterogeneity. The simulations confirm that local adaptation increases with the degree of spatial autocorrelation in the selective environment, but also show that highly heterogeneous environments are more likely to exhibit high variation in local adaptation, a result not previously described. Spatial autocorrelation in the environment influences the genetic architecture of local adaptation, with different patterns of allele frequency and effect size arising under different patterns of environmental variation. These differences will likely impact the ability to study the genetic basis of local adaptation. Overall, this work emphasizes the profound importance that the spatial pattern of selection can have on the evolution and how spatial autocorrelation should be considered in ecological and genetic studies of local adaptation.