Evaluating the effectiveness of local- and regional-scale wildlife corridors using quantitative metrics of functional connectivity

While corridors in conservation have a long history of use, evaluations of proposed or existing corridors in conservation landscapes are important to avoid the same fate as poorly-functioning "paper parks". We used resistance surface modeling and circuit theory to evaluate a number of corridors developed at regional and at local scales that aim to improve connectivity for large wildlife in the central part of the Kavango-Zambezi transfrontier conservation area. We used hourly GPS data from 16 collared African elephants (Loxodonta afriGana), and associated environmental data at used versus available movement paths, to develop a hierarchical Bayesian path selection function model. We used the resulting resistance surface across the study area as an input into circuit theory modeling to assess how well connectivity levels were captured by both types of corridors relative to several alternative scenarios. We found that the majority of regional-scale corridors performed relatively well at capturing elevated levels of connectivity relative to non-corridor comparisons, with 7 of 9 corridors rated as good or better in terms of how they captured electrical current levels (a proxy for connectivity). In contrast, only 14 of 33 smaller-scale, local corridors captured significantly higher levels of connectivity than adjacent non-corridor areas. Our results have practical implications for the design and implementation of wildlife connectivity conservation efforts in the world's largest transfrontier conservation landscape. Modern connectivity science approaches can help evaluate which proposed corridors are likely to function as intended, and which may need further refinement.