A High-Resolution Geospatial Framework for Zoonotic Plague Risk Mapping in Madagascar

Objective: Plague ( Yersinia pestis ) remains endemic in Madagascar and causes recurrent outbreaks leading to substantial mortality. This study was aimed at developing a high-resolution spatial risk assessment integrating environmental suitability and population exposure. Methods: A total of 174 confirmed plague occurrence records (1939–2023) were analyzed with a Maxent ecological niche model to predict nationwide environmental suitability. Variable contributions and permutation importance were assessed. Model outputs were downscaled to 100-m resolution and integrated with LandScan ambient population data to generate a plague exposure risk index. Commune-level spatial clustering was evaluated through Getis-Ord G i * hotspot analysis. Results: Suitability was strongly associated with elevations of 1,000–1,750 m, moderate precipitation, and cooler temperatures in the Central Highlands. Elevation contributed 57.8% to model performance, whereas temperature seasonality showed high permutation importance (37.4%). Hotspot analysis identified 484 significant communes across 13 regions, and five highland regions accounted for 82.9% of hotspots. Approximately 14.8 million people (62.8%) reside in areas with some degree of risk, including 9.4 million people (40.1%) in high- to very-high-risk zones. Conclusion: Plague risk is concentrated in Madagascar’s highlands, where substantial populations remain exposed, thus supporting targeted surveillance and public health planning.