Synthesising ecology and epidemiology for better One Health surveillance: Japanese encephalitis virus

Too often we do not consider the kinds of interactions that are central to pathogen transmission in maintenance host communities. Too often we do not even consider maintenance host communities. Our team seeks a deeper understanding of pathogen transmission within these communities to better inform spillover processes at the wildlife-livestock-human interface. Japanese encephalitis virus (JEV) serves as a useful example of this work. In 2022, JEV outbreaks were reported in domesticated pigs and humans across a very wide geographic range of eastern and southern Australia. Prior to these outbreaks, JEV incidence in Australia had been limited to a small number of sporadic cases in Far North Queensland and the Torres Strait Islands. It was speculated that the recent widespread emergence was instigated by sustained anomalously high precipitation in eastern Australia following multiple years of the La Niña phase of the El Niño southern oscillation. Our initial investigation found that, although high precipitation was associated with the occurrence of these outbreaks, the structure of the landscape was even more important. This was largely because landscape structure directs how water moves through and occupies space, but also because landscape structure determines how vectors and hosts move through and occupy space. In other words, landscape is an important driver of ecological interactions. The key aspects of landscape that impacted the JEV outbreaks are shown in Figure 2 below and you can find more details on this work here, if interested. However, these findings were only preliminary and don’t tell us about the nature of the abiotic and biotic interactions among maintenance communities and vectors, especially where these occur at points of wildlife-livestock-human interface. Which leads us to the next step…

To go a bit deeper, what more can we learn about hosts (or possible hosts), the communities in which they’re structured, and how these community structures couple with landscape structure? And how might such structures relate to the spillover of JEV to pigs and humans? As a next step to building an ecology-based approach to JEV surveillance, we explored how pools of individual species of bird hosts would be likely to assemble in local communities given the particular kinds of landscape composition and configuration associated with JEV outbreaks noted above. We were able to see what species communities might look like given particular configurations of landscape structure (Figure 3).

We also looked to see how individual species traits characterise the pools of species that reflect the landscape structure associated with JEV spillover risk (Figure 4). Interestingly, one primary trait stood out both in Australia and across the broader central Indo-Pacific biogeographical region: hand-wing index, which is a robust indicator of dispersal capacity in birds. Birds with higher hand-wing index tend to have higher dispersal capacity. So, if species pools with greater dispersal capacity structure well with high-risk JEV landscapes, this could have important implications for the mobility of the virus through broad geographic space.

While helpful, these findings so far do not provide us with the critical information we need to draw more definitive (and practical) conclusions about the ecology and epidemiology of JEV as it circulates in communities in Australia, which we ultimately need if we want to prevent spillover. This is because we still lack key information on the occupancy and abundance of species in local, realised bird communities and mosquito communities, and the interactions among the species within those communities. These ecological interactions are essential to what drives transmission of virus in time and space and so must go hand in hand with the measurement of virus itself in mosquitoes and hosts. This is exactly what we are doing. Using the information provided in the previous steps to guide where and how our sampling efforts might be best directed, we have developed a deeper approach to JEV surveillance that is grounded in the actual maintenance communities and attempts to delineate the community structure of mosquitoes and birds, the interactions among them, and the presence of virus, rather than simply the presence of virus alone as is the common practice under conventional pathogen surveillance. For example, we are simultaneously deploying mosquito traps and camera traps at key points of interface to get a more accurate picture of the composition of mosquito and bird communities in locations where JEV (as well as other arboviruses) does and does not circulate. Additionally, by analysing the blood meals of fed mosquitoes, we can identify which hosts the vectors select for feeding. We can then compare host species feeding preferences of mosquito vectors to the actual composition of maintenance host communities, thus providing a way to demarcate specific pathways of ecological interaction that can promote ongoing transmission of virus.