LAUSR

Abstract Rabies virus (RABV) is a deadly zoonosis that circulates in wild carnivore populations in North America. Intensive management within the USA and Canada has been conducted to control the spread of the raccoon (Procyon lotor) variant of RABV and work towards elimination. We examined RABV occurrence across the northeastern USA and southeastern Québec, Canada during 2008–2018 using a multi-method, dynamic occupancy model. Using a 10 km × 10 km grid overlaid on the landscape, we examined the probability that a grid cell was occupied with RABV and relationships with management activities (oral rabies vaccination (ORV) and trap-vaccinate-release efforts), habitat, neighbour effects and temporal trends. We compared raccoon RABV detection probabilities between different surveillance samples (e.g. animals that are strange acting, road-kill, public health samples). The management of RABV through ORV was found to be the greatest driver in reducing the occurrence of rabies on the landscape. Additionally, RABV occupancy declined further with increasing duration of ORV baiting programmes. Grid cells north of ORV management were at or near elimination ($\hat{\psi }_{{\rm north}}$ = 0.00, s.e. = 0.15), managed areas had low RABV occupancy ($\hat{\psi }_{{\rm managed}}$ = 0.20, s.e. = 0.29) and enzootic areas had the highest level of RABV occupancy ($\hat{\psi }_{{\rm south}}$ = 0.83, s.e. = 0.06). These results provide evidence that past management actions have been being successful at the goals of reducing and controlling the raccoon variant of RABV. At a finer scale we also found that vaccine bait type and bait density impacted RABV occupancy. Detection probabilities varied; samples from strange acting animals and public health had the highest detection rates. Our results support the movement of the ORV zone south within the USA due to high elimination probabilities along the US border with Québec. Additional enhanced rabies surveillance is still needed to ensure elimination is maintained.