Abstract

Native and introduced species can coexist in rich ecosystems where competition for resources is less intense; however, native species can be excluded from suitable habitats hosting a high abundance of introduced species. To test these hypotheses, we estimated the probability of introduced bees (buff-tailed bumblebees and honeybees) to co-occur and dominate giant bumblebees in forested landscapes of northern Patagonia. We evaluated the effects of 29 remote-sensing variables of forest ecosystems on bee abundance using data from 56 landscapes. Then, we used a probabilistic Bayesian model to estimate the effects of the most important predictors of bee abundance on their co-occurrence and dominance probabilities. Introduced and native bees exhibited heterogeneous and scale-dependent responses to environmental conditions, with only six variables affecting their cooccurrence and dominance probabilities. A high maximum temperature was associated with both a decreased co-occurrence and an increased dominance of introduced bumblebees, and the opposed pattern was found for vegetation continuous fields, an index of canopy continuity. Similar responses were found for the co-occurrence probability of honeybees and giant bumblebees. Thus, species dominance and co-occurrence changed differently along environmental gradients, with co-occurrence increasing as native and introduced species become more abundant. A high dominance of introduced bees in unsuitable habitats suggests interspecific competition is strengthened as resources become scarcer, thus reinforcing the idea that abiotic and biotic filters interact with each other.