Modelling functional landscape connectivity from genetic population structure: a new spatially explicit approach

Functional connectivity between spatially disjoint habitat patches is a key factor for the persistence of species in fragmented landscapes. Modelling landscape connectivity to identify potential dispersal corridors requires information about those landscape features affecting dispersal. Here we present a new approach using spatial and genetic data of a highly fragmented population of capercaillie (Tetrao urogallus) in the Black Forest, Germany, to investigate effects of landscape structure on gene flow and to parameterize a spatially explicit corridor model for conservation purposes. Mantel tests and multiple regressions on distance matrices were employed to detect and quantify the effect of different landscape features on relatedness among individuals, while controlling for the effect of geographic distance. We extrapolated the results to an area-wide landscape permeability map and developed a new corridor model that incorporates stochasticity in simulating animal movement. The model was evaluated using both a partition of the data previously set apart and independent observation data of dispersing birds. Most land cover variables (such as coniferous forest, forest edges, agricultural land, roads, settlements) and one topographic variable (topographic exposure) were significantly correlated with gene flow. Although inter-individual relatedness inherently varies greatly and the variance explained by geographic distance and landscape structure was low, the permeability map and the corridor model significantly explained relatedness in the validation data and the spatial distribution of dispersing birds. Thus, landscape structure measurably affected within-population gene flow in the study area. By converting these effects into spatially explicit information our model enables localizing priority areas for the preservation or restoration of metapopulation connectivity.