Estimating and validating a dynamic risk factor model for pathogen transmission using community-level bird census data : avian influenza at the waterfowl/domestic bird interface in Zimbabwe

The ecology of hosts is crucial in understanding mechanisms of pathogens transmission and spread in complex multi-hosts systems. This paradigm is used to infer epidemiological interactions in the context of Avian Influenza Virus (AIV) maintenance and spread at the interface between wild and domestic birds in a southern African ecosystem. We counted and sampled waterfowl community every two months during two years in the Chivero-Manyame catchment, an important wetland of Zimbabwe. We used the overlap in space and time of bird communities combined with ecological dynamic and non-dynamic risk factors to evaluate a risk of AIV maintenance in waterfowl and a risk of transmission of AIV to the domestic populations (backyard and intensive poultry, farmed ostriches) through the identification of potential ''bridge species''. We tested the validity of this risk model using: 1) AIV prevalence data in the waterfowl community estimated in parallel with counting; 2) AIV prevalence data in bridge species identified. The waterfowl risk identified is related to the AIV prevalence but with time lag. Prevalence in the potential bridge species indicates that some of them can play a role as bridge species in the ecosystem. We believe this protocol is a) reproducible using available bird census data and useful to explore AIV risk and identify wild bird species potentially acting as reservoir or spreader of pathogens at a local scale; b) can be used as a management tool to improve surveillance at a local level. (Texte intégral