Co-occurrence of invasive and native carnivorans affects occupancy patterns across environmental gradients

Abstract

Understanding species interactions and their effects on distributions is crucial for assessing the impacts of global change, particularly for invasive species. Co-occurrence models can help investigate these effects when interactions are likely given shared traits. For such an assemblage of invasive and native carnivorans, we examined how patterns of co-occurrence change across space and environmental gradients using a static multispecies occupancy model that accounts for imperfect detectability and models co-occurrence as a function of environmental variables, and also extended it to be temporally dynamic. We focused on invasive raccoons, which pose threats to humans and wildlife globally. In Japan, raccoons prey on many native taxa, but little is known about interactions with sympatric carnivorans. We searched for signals of competitive exclusion of native raccoon dogs (tanuki) and invasive masked palm civets by applying the model to detection data from a broad-scale trapping effort over 6 years. Forest cover was the strongest predictor of occupancy for individual species and raccoon co-occurrences, and raccoon occupancy probability increased with forest cover conditionally depending on the co-occurring carnivoran: only tanuki absence or civet presence had positive responses. However, tanuki occupancy probability increased with forest cover despite any co-occurrence. Thus, we found no evidence of competitive exclusion by raccoons, contrary to our expectations. As parts of the world with invasive raccoons can also have invasive tanuki, our findings may have broad management implications. The model we present should be useful for inferring signals of biotic interactions between species with low detectability over multi-year time frames.