Individual‐based models allow accurate prediction of introduced large herbivore populations in rewilded landscapes

Abstract Trophic rewilding via the (re)introduction of keystone species, such as large herbivores, is increasingly being considered in Europe to support nature recovery and improve the resilience of ecosystems in the face of rapidly changing environmental conditions. Large herbivore presence can both benefit and disbenefit local communities, making it important to predict likely expansion patterns and identify, among other things, possible zones of human–wildlife conflicts. We built a predictive, spatially explicit, individual‐based model (IBM) to examine reintroduced Corsican red deer ( Cervus elaphus corsicanus ) population expansion in Corsica using the recently developed integrated Step Selection Function approach. We used GPS data collected during a 5‐year intensive field study of reintroduced red deer to develop an SSF describing habitat selection. We then combined the outputs of this SSF with information on deer life histories in an IBM to predict deer expansion on the island in the coming years. Our model accurately recreates the observed recolonisation patterns to date in the three monitored reintroduction sites, adequately predicting home ranges, mother‐offspring home range centroid distances and habitat use. We therefore used this model to predict deer distribution expansion in the next 5 years, using information from all known reintroduced populations on the island. Under this model, we predict deer range expansion rate to vary between ca. 130 and 166 km 2 per annum. Furthermore, we identify potential zones of future human–deer conflict, with the greatest potential conflict identified for the southern populations. Synthesis and applications . As the number of trophic rewilding projects increases in Europe, there is a real need to anticipate the ecological and societal consequences of species (re)introductions to ensure their long‐term success. Predictive approaches that integrate locally calibrated information on movement and life histories provide a unique opportunity to increase the cost‐effectiveness of such projects, enabling the identification of potential human–wildlife conflict zones before conflict occurs. This is especially important for island fauna such as the Corsican red deer, which are known to be more vulnerable to extinction and for which reintroduction outcomes tend to be less studied.