Enhancing Habitat Connectivity in Fragmented Landscapes: Spatial Modeling of Wildlife Crossing Structures in Transportation Networks

Roads fragment habitat and pose a variety of threats to wildlife by restricting their physical movements or causing direct mortality. Wildlife crossing structures can mitigate some of these effects by simultaneously connecting fragmented habitat and reducing collisions between animals and traffic. This research develops two classes of spatial models for optimally locating crossing structures in situations where the goal is to connect discrete habitat patches that are fragmented by roads. The first group of models uses variations of minimum spanning forest (MSF) problems to select crossing structure locations in scenarios where the goal is to connect all habitat patches in the landscape. This article explores three variations: distance-based, intersection-weighted, and component-based MSFs. The second set of models includes newly developed integer programs with objectives that maximize interpatch connectivity given a fixed number of crossing structures. One of these integer programs maximizes overall connectivity in a landscape, and the other maximizes the connectivity of patches to a core reserve. We illustrated the models using hypothetical landscapes and applied them to a case study where the goal is to connect vernal pool habitat for amphibians. Results of the applications guide a discussion of the relative merits of each approach for meeting various conservation goals related to improving habitat connectivity in fragmented landscapes. More broadly, this article develops an analytical framework for dealing with habitat connectivity issues due to road network fragmentation as well as a new neighborhood graph and several spatial models.