Comparing methods for assessing habitat connectivity: a case study of guinas (Leopardus guigna) in a fragmented Chilean landscape

Abstract

Fragmentation creates a matrix which can facilitate or impede connections among patches of habitat. Least-cost path (LCP) and circuit theory (CT) are two methods commonly used to evaluate landscape-level connectivity among patches. Both methods use resistance surfaces that can be generated from Resource Selection Functions (RSF) or the Analytical Hierarchy Process (AHP). Despite the potential conservation implications of connectivity analyses, the methods are rarely compared. I quantified how RSF and AHP resistance surfaces affect estimates of connectivity among protected areas using a South American wild cat, güiña (<italic>Leopardus guigna</italic>), as a case study. I found that 1) path rankings and predicted locations of pinch points depended on the metric and resistance surface used, and 2) LCP is more sensitive to resistance surfaces than CT. These results confirm that connectivity analysis methods should be carefully considered and compared before they are used for conservation decisions.