Spatial distribution and habitat suitability of tsetse (Glossina spp.) in Cote dIvoire: An ensemble modeling approach to support targeted disease control
Background Tsetse are vectors of trypanosomes responsible for African animal trypanosomosis (AAT) and human African trypanosomiasis (HAT). While Cote dIvoire has successfully eliminated HAT as a public health problem and approaches elimination of transmission, AAT remains a major obstacle to agriculture and livestock production. Understanding the spatial distribution of tsetse is essential for prioritizing and sustaining disease control and elimination efforts. Methodology/Principal Findings Using 1,702 occurrence records from the national tsetse atlas we modeled the habitat suitability of the nine tsetse species present in Cote dIvoire. We identified suitable habitats in unsampled areas and quantified environmental constraints on tsetse distribution. Resampling the data to a 1km x 1km grid produced spatially explicit outputs at a resolution more relevant for operational planning. An ensemble modeling approach was employed integrating four algorithms--Random Forest, XGBoost, Maximum Entropy (MaxEnt), and Generalized Additive Models (GAM)-- with satellite-derived environmental and anthropogenic predictors--which achieved high predictive accuracy, area under the curve and True Skill Statistics 0.80 and 0.83, respectively. Distance to waterbodies, soil moisture, distance to protected areas, maximum land surface temperature, and sheep density were key drivers of habitat suitability. Importantly, the models identified suitable habitats in 11 administrative regions not covered by the atlas, providing an improved national tsetse risk profile. Conclusions/Significance These results provide a detailed assessment of the ecological suitability of tsetse across Cote dIvoire and their persistence in agroecological mosaics with high human and livestock densities. We offer a high-resolution blueprint for vector and disease control, particularly in areas where field data are currently lacking. We provide a robust framework for evidence-based decision-making within the Progressive Control Pathway (PCP) for AAT by enabling the identification of priority areas and resource allocation optimization to improve livestock productivity through more effective AAT control and reduce the risk of resurgence of HAT.