Ecological dynamics of emerging bat virus spillover

CIDD researchers Raina Plowright and Peter Hudson have constructed a conceptual model to describe patterns of Hendra virus emergence in Australia. The work, published this year in the Proceedings of the Royal Society B, used Hendra virus, as a model system for understanding broader patterns of zoonotic disease spillovers that emerge from bats. These broader patterns are of particular interest in light of the devastating consequences such spillovers have had on countries where Ebola is epidemic. Hendra virus is an appropriate model system as it spills over from bats into horses, with occasional onward transmission from horses into humans.

In Hendra virus emergence and in other viral diseases with bat reservoirs, bat viral shedding appears to be pulsed in space and time. Currently, two different hypotheses are used to account for those spatiotemporal pulses. The first hypothesis is that spatiotemporal patterns of bat viral shedding are a function of episodic shedding from chronically infected bats. If infected bats shed only some of the time, and if viral shedding depends on nutrition, lactation, or some other environmentally or temporally linked predictor, then spillover risk should scale with those predictors. By contrast, if spatiotemporal patterns in viral shedding depend instead on transient waves of infection permeating the bat community, then risk at a given point in space and time depends on risk in surrounding areas during the preceding time period.

The researchers point out that these two hypotheses point to fundamentally different processes: a within-host immune function process, in the case of the episodic shedding hypothesis, and a metapopulation-level host contact process, in the case of the transient waves hypothesis. Furthermore, each hypothesis currently rests on a very limited empirical basis. The episodic shedding hypothesis needs to be tested in controlled settings where within-host markers can be tracked longitudinally and environmental drivers can be directly controlled. The transient wave hypothesis requires detailed knowledge of basic disease transition rates, many of which are not well known, along with better knowledge of bat density and movements in the wild.

On the recipient host side, the authors note that historically only a small proportion of horses in a given paddock generally become exposed. Horses that feed in the “drip zone” surrounding bat colony trees are subject to the highest risk of pathogen encounter. The authors note that Hendra spillover peaks during the period of lowest pasture productivity. They point out that hungry horses may be more likely to eat partially consumed fruits under bat colony trees during that season.

Finally, the researchers point out that the true horse exposure rate is unknown. It is possible that many horses are exposed, but only a few develop disease. This could be due to heterogeneity in horses’ innate immune function and individual behavior patterns. Currently, very little data is available to explore horse heterogeneity.

Applying an understanding of Hendra virus emergence, the researchers identified six requirements as essential for viral emergence. These factors included the presence of infected bats, the shedding of viral particles that can survive in the environment and host requirements such as novel hosts needing to be exposed and susceptible to the virus.

Taken together, the synthesis in the recent paper underscores the importance of scale in understanding disease emergence. The authors point out that management actions can be designed to target each specific condition required for disease emergence, and that some scales may be more amenable for management than others. Plowright and colleagues promote a cross-scale approach ranging from cellular to landscape levels when tackling the management challenges of bat virus emergence events.

 Synopsis written by Kezia Manlove.