19-EEID US-UK Anticipating dynamic responses to disease control interventions in reservoirs: the science of vampire bat rabies management

Wild animals host a wide variety of pathogens that can spread between host species. Such diseases, including Ebola and COVID-19, significantly affect human health, agriculture and wildlife conservation. Historically, disease mitigation methods (e.g. vaccination, therapeutics) have focused on humans or livestock rather than the wild animals that perpetuate the pathogens in nature. For example, we treat humans after they become ill from a disease that originated in other species, rather than limiting the spread of the disease within its animal source. Controlling diseases within wildlife could be more effective in controlling disease but is currently limited by three factors. First, many diseases are maintained in cycles that spread across landscapes, but wildlife are notoriously difficult to observe at these large spatial scales, making responses to interventions unpredictable. Second, tools like vaccines have been difficult to administer to sufficient numbers of individuals to actually reduce disease transmission in the wild. Third, interventions are usually bounded by societal constraints, whether financial (e.g., limited funds to invest) or sociological (e.g., conflicting desires among multiple stakeholders). This project uses a highly tractable disease that has major health and agricultural impacts as a model system to determine how each limiting factor for disease control in wildlife can be overcome.

The project will conduct field and laboratory research to test specific hypotheses about vampire bat-transmitted rabies, a viral infection that has major human health and agricultural impacts across Latin America. The study will take advantage of two areas where novel tools and strategies enable unprecedented insights into disease spread and management: the miniaturization of animal-borne tracking systems and the development of vaccines that can spread among wildlife in the wild. Field experiments using animal-borne GPS tags, along with landscape-scale data on bat presence from questionnaires and historical rabies outbreaks in cattle, will be used to generate a spatially-explicit dynamic model of bat abundance and dispersal. Studies on captive and wild vampire bats will resolve key uncertainties related to the implementation of self-spreading therapies that target bats. Finally, stochastic epidemiological models will use parameters estimated from fieldwork and captive studies and the bat population dynamic model to identify optimal intervention strategies for localized control and regional elimination of vampire bat rabies that preserve diverse stakeholder requirements (e.g. wildlife conservation goals as well as improved human and livestock health).

Broader impacts: This research constitutes a step towards preventative management of an important bat virus while strengthening international research capacity. In doing so, it will create generalizable quantitative frameworks that will be useful to inform how potentially revolutionary technologies could be applied to other wildlife diseases.

The major goal of this research is to assess ecological mechanisms underlying the success or failure of current and proposed disease control interventions in wildlife reservoirs. We focus on vampire bat (Desmodus rotundus) transmitted rabies, a highly tractable model that also carries substantial applied importance for human and livestock health in Latin America. As with many wildlife zoonoses, the primary limits to evidence-based management of vampire bat rabies are both epidemiological and sociological. We need better understanding of how rabies spreads across heterogeneous landscapes and how interventions can be designed to take advantage of these dynamics. Yet, theoretically optimal solutions must be robust to real world constraints such as limited funding and conflicting needs among stakeholders. In our study, field experiments using animal-borne GPS tags, along with landscape-scale data on bat presence from questionnaires and historical rabies outbreaks in cattle will allow us to generate a spatially explicit dynamic model of bat abundance and dispersal. Studies on captive and wild vampire bats will resolve key uncertainties related to the implementation of self-spreading therapies that target bats. Stochastic epidemiological models will use parameters estimated from fieldwork and captive studies and the bat population dynamic model to identify optimal intervention strategies for localized control and regional elimination that accommodate diverse stakeholder requirements. This study will constitute a major advance towards evidence-based management of an economically damaging zoonosis affecting livestock-reliant countries across Latin America. More broadly, it will forge new theoretical ground to illustrate how the next generation of self-disseminating technologies for wildlife disease control and the next generation of technologies in animal spatial ecology can be quantitatively integrated to improve human and animal health.