Bridging the data gap: Integration of spatial modelling in wildlife disease surveillance

Abstract

Although wildlife disease surveillance is central to the One Health framework, data on hosts, vectors, and pathogens are frequently incomplete, biased, and heterogeneous. This poses major challenges for inferring spatial disease patterns. Spatial modelling offers a powerful means to extract epidemiological insight from such data and to support disease management and conservation decisions. This review evaluates three complementary spatial modelling frameworks, i.e., data-driven, process-driven (mechanistic), and decision-driven, to demonstrate how each can extract epidemiological insights from imperfect wildlife data. For each framework, (1) relevant data sources and their inherent challenges are summarized; (2) spatial models commonly used in wildlife epidemiology are described; (3) their advantages, limitations, and practical strategies for bias mitigation such as restricted background sampling, effort covariates, and integrated distribution models are discussed; and (4) their applications are illustrated through case studies. Overall, this review aims to position spatial modelling as a proactive tool for guiding surveillance design and management decisions, in addition to serving the retrospective analytical approach.