COS 208-3 - Breaking barriers: Making wildlife disease models accessible for decision makers

Abstract: Ecological models are powerful tools for elucidating the impacts of wildlife disease in individuals, populations, or ecosystems. While ushering in valuable knowledge, new complex models require specialized knowledge to derive, program, run, troubleshoot, and interpret results. Although these complexities are surmountable to academics trained in model theory, implementation, or programming, they render these models largely inaccessible to broader audiences, including wildlife managers, policymakers, stakeholders, and scientists in other disciplines. The Wildlife Health Lab at Cornell University recognized this limitation and responded by developing novel “model translation” pipelines within the Surveillance Optimization Project for Chronic Wasting Disease (SOP4CWD). Specifically, our team dedicated time and effort to programming complicated models into automated scripts, minimizing the need for wildlife agencies to employ “in-house” biometricians to adapt models for use. Initial pipeline prototypes were provided through interactive R Shiny applications, allowing the user to load data, run the model, and view the results. Secondary evolutions include the fully automated “CWD Data Warehouse”, which intakes wildlife disease surveillance data, demographic parameters, and risk factors, meshes them with disease models, and provides, at the click of a button, resulting graphics. Utilization of the CWD Data Warehouse removes barriers to agencies lacking technical expertise and allows application of complicated models for the betterment of our natural resources. The techniques we are developing can be applied to other wildlife diseases and ecological models in general to make them more accessible.