Global Avian Flu Risk: Waterbird Entropy for Targeted Surveillance

Avian influenza is a major zoonotic threat with pandemic potential 1. Waterbirds, natural reservoirs for most known subtypes2 of avian influenza viruses (AIVs), facilitate spillover into farm animals and humans through exposure and frequent virus reassortment 3,4. Here we propose waterbird activity entropy (WAE), a novel indicator quantifying waterbird activity intensity based on monthly distributions of 779 waterbird species worldwide. WAE demonstrates high predictive power (AUC = 0.86, s.d. 0.001) for global AIV cases, particularly for highly pathogenic H5N1. It identifies AIV exposure hotspots across 14% of global land area, with key regions—USA, EU, China, and India—housing 52% of exposed humans, 41% of cattle, and 51% of poultry. The USA and India face significant cattle exposure risks, with hotspot areas covering 226 Mha (68% of US hotspots) and 114 Mha (99% of India’s hotspots), respectively. Although sub-Saharan Africa accounts for less than 1% of reported AIV cases, it encompasses over 300 Mha of AIV exposure hotspots, representing approximately 15% of global hotspot areas and considerable surveillance gaps there. Waterfowl and seabirds contribute most to global exposure risk. This WAE-based framework, incorporating waterbird residency time, enhances AIV risk assessment, offering critical insights for anticipating AI emergence and improving surveillance.