Ebola Virus
Ape Population Impact
In 2002, while analyzing wildlife survey data for a World Bank Funded National Park planning process in Gabon, I found a strong correlation between distance from human Ebola outbreak sites and non-human ape (gorilla and chimpanzee) density. After compiling additional survey data from colleagues and correcting for factors such as hunting intensity, further analyses indicated that Ebola had caused a roughly one third decline in the world gorilla population, with a smaller but still substantive impact on chimpanzees. Subsequent analyses based on data collected by Magdalena Bermejo and colleagues showed that more than 5,000 gorillas died from Ebola just in the region surrounding Lossi Sanctuary of northwest Congo. I then led a report that resulted in the up-listing of western gorillas (Gorilla gorilla) to Critically Endangered, one of the first cases in which emergent disease, rather than hunting or habitat loss was the major driver of up-listing.
Wildlife Transmission Dynamics
When I first published work on Ebola reservoir dynamics in 2005, it was not well-received. The consensus view was long that Ebola emergences were effectively unpredictable, with climate human habitat disturbance often cited as triggers. However, that conclusion was not supported by my analyses of the spatio-temporal pattern of outbreaks and the spatial structure of the Ebola genome. Rather, Ebola appeared to spread in consistently moving, highly predictable epizootic waves, presumably through the putative reservoir host, fruit bats. Phylogenetic analyses by coauthor Roman Biek supported this view, suggesting that all known Ebola isolates were descendant from the same common ancestor circulating just previous to the first documented Ebola outbreak in 1976.
Fifteen years later, the reservoir dynamics of Ebola remain unresolved. On one hand, subsequent patterns of emergence within gorilla range have been extraordinarily well predicted by the original wave spread model. On the other hand, phylogenetic analyses have estimated subsequent Ebola isolates sampled far apart in space to be closely related. This has been interpreted as indicating rapid long distance spread of the virus, presumably through bat migration. I’ll soon publish results showing that it actually reflects catastrophic failure of phylogenetic model assumptions.
Potential for Control
These results had serious conservation implications in that they implied that gorilla vaccination with newly developed Ebola vaccines might be efficiently targeted just ahead of advancing epizootic waves. Data from Bermejo and colleagues also indicated extended, localized waves of transmission through gorilla populations, contradicting the assertion that most gorillas were being infected by “massive spillover” directly from bats. Wave spread opened the possibility of barrier vaccination as used subsequently, for example, in rabies outbreaks amongst Ethiopian wolves. Simulation modeling and analyses of observational data collected by Thomas Breuer and other researchers in Congo also suggested that gorilla societies had the kind of long-tailed contact network structure that might allow efficient disease control through the preferential vaccination of “superspreaders”. This epidemiological evidence that Ebola impact on gorillas might be efficiently controlled spurred me onto a campaign to get Park managers in Africa to vaccinate wild gorillas. They wouldn’t act, so, I led two vaccine trials on captive chimpanzees and one on wild gorillas to demonstrate that vaccination could be safe, immunogenic, and cost effective. These vaccine trials dropped me into two controversies involving both antivaxxers and animal rights activists.