Canid guild changes in Europe: evidence for a continental-scale mesopredator release of golden jackal (Canis aureus)

The golden jackal’s Canis aureus range in Europe is expanding rapidly and populations have increased significantly during the last decades. The presence of this new carnivore could impact existing animal communities and is already receiving high interest among wildlife managers. The decrease and fragmentation of previously dense grey wolf Canis lupus populations by humans have been suggested as potential trigger for such pattern. Historical trends in the populations of both canids and local evidence of golden jackal disappearance in areas recently recolonized by wolves tend to support this hypothesis. In this study, we developed species distribution models (SDMs) at a continental scale to investigate the golden jackal environmental niche and identify the potential influence of wolf presence on jackal habitat suitability. Since jackals are highly mobile and opportunistic animals, dispersers can temporarily move through nearly any habitat type. To prevent overestimation of the species’ environmental niche, we restricted our study to established territorial jackal groups sampled by means of acoustic stimulation. Nine di!erent SDM algorithms were calibrated and evaluated within the core distribution range of the species. We used environmental variables relevant to the species ecology: annual snow cover duration, land-cover and wolf presence; all uncorrelated and mapped at a 5 km resolution. Results/Conclusions We surveyed a total of 2,497 distinct locations across 11 European countries and detected 820 territorial jackal groups. GBM and Maxent algorithms performed best (average AUC = 0.91). Snow cover duration accounted for the highest variable contribution (37.2%), followed by wolf presence (20.8%). Forest and agricultural land prevalence, as well as distance from settlements and hydrological features were also selected in the best models. Jackal habitat suitability was highest in areas with short snow cover duration and heterogeneous land cover. Average jackal probability of presence ranged from 0.21 in areas of permanent wolf presence to 0.73 in areas of wolf absence. Although snow cover duration was the most influential variable to predict jackal distribution, the model predictive ability was significantly improved by including the wolf presence covariate; implying that grey wolf presence also significantly a!ects golden jackal habitat suitability. More generally, this analysis supports the hypothesis that jackal expansion in Europe was triggered by a large-scale mesopredator release following wolf persecution in 19th and 20th century