Determining priorities for restoring connectivity in a Mediterranean landscape fragmented by roads

European landscapes have been intensively affected by human activity during centuries, producing natural habitat loss and fragmentation. Roads are a disrupting element in landscapes, acting as barriers to wildlife movement and increasing mortality due to road kills. It is indispensable for wildlife conservation to maintain ecological connections among the remaining natural areas in the territory. Understanding connectivity patterns is crucial for this purpose, and localizing and prioritizing critical points for connectivity is essential to create a functional ecological network (Gurrutxaga & Saura, 2014; Santini et al., 2016; Terrones & Bonet, 2016).

The aim of this work is to contribute to the improvement of the methodological framework analyzing the effect of roads in the functional connectivity of a Mediterranean landscape, and prioritizing critical points for restoring connectivity. The study was conducted in the Valencian Community, in eastern Spain. This territory is characterized by a high contrast between the plain coastal zones, with a high density of developed urban areas, crops and roads, and the steepest inland zones, less fragmented and covered by forests, shrublands and oldfields.

We have developed connectivity least-cost models. We used as objective species a functional group, consisting of carnivorous mammals present in the study area: red fox (Vulpes vulpes), stone marten (Martes foina), common genet (Genetta genetta) and European wildcat (Felis silvestris). To evaluate the effect of fragmentation by roads, we worked with two scenarios: the first one considering the present land cover, and the second one, removing the fragmentation effect of roads (equivalent to a total permeability of land affected by roads). We have created two resistance maps, one for each scenario, using three layers of geographic information: land cover map, hydrography, and distance to urban areas. We have defined core areas as forest zones included in areas of Natura 2000 network (n=148). We have calculated least-cost paths between each pair of core areas with Pathmatrix (Ray, 2005) (n=10,443 connections). We have analyzed least-cost paths using Conefor (Saura & Torné, 2009), for two median dispersal distances (2 and 6 km). We have calculated the differences in the Probability of Connectivity index (PC) for each link between the two scenarios.

Results showed an important effect of roads on global values of connectivity. PC index decreased 37.76 % between the two scenarios for a median dispersion distance of 2 km, and 26.83 % for 6 km. We have established a prioritisation of the least-cost paths with a higher decrease in PC index, which are mainly located in the inland zones. The results contain detailed cartographic information for the detection and selection of the most important spots for restoring connectivity and defragmenting the road network, and hence, for improving functional connectivity for carnivores in the Valencian Community.