In December 2015, a raccoon infected with the racoon rabies virus variant (RRV) was reported in Hamilton, Ontario. This report was the beginning of one of the largest outbreaks of raccoon rabies in Canada. As a result, the Ontario Ministry of Natural Resources and Forestry (MNRF) expanded existing vaccination campaigns to eliminate raccoon rabies from Ontario until at least 2025.
In this study, we asked the following: 1) what is the predicted geographic extent of uncontrolled rabies versus controlled rabies between 2015 and 2030, with controls continuing until 2025? And 2) when may the raccoon rabies outbreak in southern Ontario have actually started?
We used a spatially-explicit agent-based modelling approach to predict the number of new infections per year across southern Ontario according to different management scenarios. Modelling was done using SamPy, a new Python library, which simulates individual raccoons and their families, as well as raccoon and pathogen biology and intervention use. Using the mapping software QGIS and remote sensing data, we created a hexagonal grid of 10-km2 cells across southern Ontario, predicted the carrying capacity of each cell, and used georeferenced MNRF intervention data to predict current and future weekly rabies infections.
Results/Conclusions
Preliminary results show that, across all scenarios, no intervention use between 2015 and 2025 results in the furthest geographic spread of raccoon rabies by 2030, with infected raccoons spreading through Windsor, Ontario (~395 km from the first detected case) by 2025 and reaching the southern edge of Algonquin Park (~315 km from the first detected case) by 2028. The scenario with intervention use continuing until 2025 provides the highest likelihood of RRV elimination. Our calibration tests currently indicate that the outbreak, with earliest detections in the first week of December, may have actually started as early as the first week of October, 2015. In conclusion, our models account for the spatial variations in raccoon habitats and intervention strategies, the dynamics of host and pathogen biology, and the intensity of interventions in space and time. Our exploration of the possible geographic extent of the disease with and without control, and the potential start week of the outbreak, is only possible thanks to advances in agent-based modelling. Our findings will inform on-the-ground wildlife vaccine programs in southern Ontario as well as provide a starting point for a cost-benefit analysis of controlling raccoon rabies.
