Abstract
Introduction : Detailed information is available on sleeping sickness and tsetse in the endemic foci in Côte d’Ivoire, with elimination of transmission on the horizon. Much less is known about the distribution of tsetse at national level and of African animal trypanosomosis (AAT) despite the heavy burden for smallholder farmers. As the country aims to improve livestock production, with a focus on the northern cotton belt, information on AAT risk is badly needed. To fill this gap, we predicted AAT and tsetse distribution at the national level.Methods : We developed species distribution models for the two main tsetse species (Glossina palpalis gambiensis and G. tachinoides), and AAT (Trypanosoma brucei brucei, T. congolense and T. vivax) at national level. We used presence only (Maximum Entropy; Maxent) and presence-absence models (Generalized Linear Model; GLM and Random Forest; RF) trained with data from a national atlas of tsetse and AAT currently under development (2012-22). We used readily available environmental variables on vegetation, temperature, rainfall, soil moisture, rivers and topography, human and livestock densities and distance to protected areas.Results : Our tsetse distribution models directly aided the Ivoirian Direction of Veterinary Services in designing targeted field sampling within the cotton belt, prioritizing areas critical for tsetse elimination efforts. Population genetics studies are underway to investigate potential isolation of tsetse populations, which may be linked to key environmental factors identified by our models, i.e. the distance to rivers and classified forests. This understanding, for example, helps in explaining tsetse absence in the South-West region. These combined insights are crucial to inform future control.Conclusion : Our tsetse and AAT distribution models inform policy decisions and help target surveys to fill data gaps as Côte d’Ivoire is in the process of developing its Progressive Control Pathway for AAT.