Geostatistical mapping of transboundary cattle disease risks in Ethiopia

Abstract

Ethiopia experiences devastating economic losses from an ongoing endemic burden of trans-boundary animal diseases (TADs). TADs are highly transmissible infectious diseases of animals, often able to spread rapidly with significant economic and public health consequences. Contagious bovine pleuropneumonia (CBPP), foot-and-mouth disease (FMD), and lumpy skin disease (LSD) are among the global priority TADs for cattle. In this study, we used responses from a survey about cattle disease delivered to livestock keepers across Ethiopia. We used generalized additive mixed models applying neighborhood cross-validation method which accounts for spatial dependence in the data to investigate the spatially variable relationships between bioclimatic variables and distribution of CBPP, FMD, and LSD in Ethiopia. We also developed model-based risk maps of these diseases using a geostatistical kriging method to guide knowledge-based decision making. The results show the risks of CBPP vary with altitude and relative humidity, risks of FMD with temperature and relative humidity, and of LSD with temperature and precipitation. The gaussian spatial smooth terms are all significant. The maps are produced using rigorous statistical analysis with very low prediction errors and can thus be considered reliable. Our results have implications for the impacts of climate change, and the vulnerability of communities in high-risk areas. The risk maps illustrate how such maps contribute to climate-informed disease early warning systems.