Landscape Permeability Improves Climate-Based Predictions of Butterfly Species Persistence

Abstract

Habitat modification alters species' capacities to track shifting climatic conditions. Broad-scale analyses that explore demographical responses to on-going climate change tend to neglect the influence of the underlying landscape pattern. However, many landscapes are fragmented by human activities, which might make dispersal for many species more challenging. Determining the extent to which landscape factors affect broad-scale distributional patterns has implications for our ability to predict realistic climate change impacts on species. Here, we constructed species-specific measurements of landscape permeability for 96 butterfly species in southern Ontario to test whether this landscape characteristic affected species' distributions at macroecological scales. We used multiple logistic regression models to test for the effects of permeability and its interaction with temperature on butterfly species presence/absence. We found that 48% of butterfly species responded to landscape permeability alone or in interaction with temperature. In general, the effect was positive (87%) and species were more likely to be present with increasing landscape permeability. For 61% of the species that responded to broad-scale landscape permeability, the interaction of temperature with permeability was statistically significant. In warm areas, species were more likely to be present if landscape permeability was high. Landscape permeability explained 3-43% of residual variability in species' presences after accounting for temperature. Finally, we show how fine-scale permeability measurements can be combined with large-scale patterns of diversity to inform conservation efforts. Landscape permeability can affect species' distributions at broad-scales and understanding factors that potentially influence species' dispersal can improve predictions for how species respond to changing climatic conditions.