Evaluating least-cost model predictions with empirical dispersal data:: A case-study using radiotracking data of hedgehogs (Erinaceus europaeus)

Habitat fragmentation and habitat loss are widely recognized as major threats to biodiversity on a regional as well as on a global scale. To restrict its effects, ecological networks such as the trans-European network NATURA2000 are being developed based on the assumption that structural connections between habitat fragments lead to increased exchange through dispersal and a higher viability of (meta)populations. However, there is a great need for techniques that translate these networks and/or structural characteristics of landscapes into functional connectivity for specific organisms. Least-cost analysis has the capacities to fulfill these needs, but has never been validated against actual observations of dispersal paths. Here we present a method to validate the results of a least-cost analysis by comparing realized movement paths of hedgehogs in unfamiliar areas, obtained by radiotracking, with statistics on landscape-wide distribution of cost values. The degree of correspondence between empirical dispersal paths and the output of a least-cost analysis can be visualized and quantified, and least-cost scenarios can be statistically compared. We show that hedgehogs moved along paths with significantly lower cost values than the average landscape, implying that they took better than random routes, but performance was relatively poor. We attribute this to the relatively generalistic habitat use of the model species and the rather homogeneous landscapes. We conclude that this approach can be useful for further validation of the least-cost model and allows a direct comparison of model performance among different taxa and/or landscapes. (C) 2007 Elsevier B.V. All rights reserved.