Ecological connectivity analysis to reduce the barrier effect of roads. An innovative graph-theory approach to define wildlife corridors with multiple paths and without bottlenecks

Ecological connectivity studies should be performed as baseline studies to prevent ecosystem fragmentation during the planning phase of a linear transport infrastructure. A landscape can be simplified as a graph network of habitat patches (nodes) and wildlife corridors (links) that connect them. Our analysis focused on roe deer (Capreolus capreolus L.), one of the large mammals most commonly hit by vehicles on the Spanish road network. We develop a network approach, implementing an iterative GIS methodology to obtain alternative corridors with comparable costs and without bottlenecks below a user-defined minimum width. This method enables the definition of the clearly delimited physical area of corridors according to a geometrical threshold width value, as well as multiple corridor connections for a pair of habitat patches. We compare the connectivity estimated with the least-cost path with our proposed methodology, observing even absence of significant differences at global scale, but not to local scale in our study area. Our results highlight the potential relative importance of each node habitat patch and corridor for the conservation of global connectivity. Finally, we discuss applications for locating habitat restoration as a compensatory measure and potential sites for wildlife crossings, creating new stepping stones and evaluating road layouts using the selected freeway as an example.