Landscape genetics outperforms habitat suitability in predicting landscape resistance for congeneric cat species

Abstract

Aim The use of landscape resistance maps to model connectivity has become an indispensable tool for species conservation. However, different methods can be used to estimate landscape resistance, but there is no consensus on which is the most reliable one. Therefore, comparing the performance of those methods in predicting resistance can be quite useful to understand their limitations and conservation implications. Our goal was to evaluate the accuracy of two commonly used approaches, habitat suitability modelling and landscape genetics, in estimating landscape resistance to genetic connectivity of two species of Neotropical cats (Leopardus guttulus and L. geoffroyi) across their ranges. Location South America. Taxon Felidae—L. guttulus and L. geoffroyi. Methods For both species, we optimized a landscape genetics resistance surface using a restricted multivariate optimization approach and transformed a habitat suitability map into a resistance layer. We compared landscape resistance models created by these two approaches based on the models' Akaike information criterion scores and evaluated the similarities and differences in their predictions by calculating the correlation between the resistance layers and generating difference maps. Results The genetic approach greatly outperformed the habitat suitability approach in explaining movement driving gene flow for both species. For the studied species, habitat preference and genetic connectivity are influenced by different landscape features. Habitat alteration imposes great resistance for genetic connectivity, and the presence of natural vegetation remnants within altered environments is essential for their conservation. Main conclusions For the studied species, the transformation of habitat suitability models into resistance surfaces is a poor proxy for permeability to dispersal, and the use of genetic data is more reliable in modelling connectivity for species conservation. Habitat suitability and landscape resistance are not equivalent or even proportional for these species.