Traditionally, ecological networks are designed adopting either a physical approach or an ecological approach based on Habitat Suitability Models (HSMs). However, the physical approach only considers structural connectivity, which does not guarantee the necessary functional connectivity. Conversely, if the actual suitability of each land-cover type for a given species is not known a priori, the HSMs cannot include fragmentation metrics describing the spatial arrangement of all the land-cover types equally suitable for the species, which should be merged together to obtain correct metrics. In addition, HSMs are not able to discern objectively if the species perceives a land-cover type as a node, a connectivity element or matrix. To overcome these issues, we proposed a new ecological approach, which compares different simulated species landscape perceptions (SSLPs) corresponding to every possible combinations of the land-cover types in the real landscape, alternatively assuming the role of a node, a connectivity element, or matrix. Basically, the method compares the ability of the fragmentation metrics calculated for each SSLP to explain the actual species distribution. The best performing SSLP will provide information about the land-cover types that should be used to design an effective ecological network for the target species, and their role.
Results/Conclusions
We applied the method to the case study of the Hazel Dormouse in an agricultural landscape in Northern Italy testing the role that hedgerows, poplar cultivations, biomasses and reforestations could play in an ecological network for the species. Starting from the best performing SSLP, we concluded that, for an effective conservation of the Hazel Dormouse, we should use hedgerows other than woodlands as nodes and poplar cultivations, biomasses and reforestations as connectivity elements. Specifically, the SSLPs in which hedgerows play the role of node performed better than the others in 63% of the cases, while the SSLPs in which poplar cultivations, biomasses and reforestations, considered separately or together, play the role of connectivity element performed better in the 100% of the cases. In conclusion, the application of the method to the case study demonstrated its ability in objectively differentiate the role of a node, a connectivity element or matrix for all land-cover types in a landscape. Overall, the proposed method can be applied to virtually every species sensitive to fragmentation of potentially every kind of habitat. Moreover, working on land-use polygons, it could be easily incorporated in landscape management plans, offering a great opportunity to integrate conservation and socio-economic interests.