The 82nd Annual Meeting of the American Association of Physical Anthropologists (2013)

The interplay between behavior and disease: investigating pathogen transmission dynamics in wild chimpanzees with social network models


1Odum School of Ecology, University of Georgia, 2The Dian Fossey Gorilla Fund International, 3Section of Integrative Biology, University of Texas at Austin, 4Department of Anthropology, University of Illinois at Urbana-Champaign

Thursday All day, Clinch Concourse Add to calendar

Great apes demonstrate tremendous variation in social contacts, which epidemiological studies show can be crucially important for predicting pathogen spread. Wild apes are also vulnerable to several directly-transmitted respiratory pathogens that have caused major population declines. Network-based epidemiological models provide an effective approach for mathematically formalizing heterogeneous transmission pathways and can be a powerful tool for predicting outcomes of future pathogen introductions and disease control efforts. Using such models, we investigated how behavioral association patterns may affect pathogen transmission dynamics within the wild Kanyawara chimpanzee community (N=37) in Kibale Forest, Uganda. Specifically, we used association data from field observations to construct nine monthly contact networks on which we stochastically simulated epidemics to explore how outbreak sizes depend on i) pathogen infectiousness, ii) network position of the index case, and iii) the month of the outbreak. We then used generalized linear models to identify individual characteristics that influence outbreak size. We found that outbreak sizes were largest in the presence of estrous females due to an increase in mean network connectedness. Outbreak sizes were also greatest when the index case was a high-ranking male or a chimpanzee with a large family unit, as these individuals tended to be the most highly connected. Follow-up simulations of vaccination strategies revealed that for a fixed level of coverage, epidemic risk could be greatly reduced by targeting these individuals relative to random vaccination. Overall, our work demonstrates how combining behavioral data with network models can offer practical applications for pathogen control of endangered wildlife species.

This study was funded by the US Fish and Wildlife Service, the ARCS Foundation, Morris Animal Foundation, the Margot Marsh Biodiversity Founation, and Fulbright.

comments powered by Disqus