A continuous phylogeographic approach to explore environmental and ecological drivers of Highly Pathogenic Avian Influenza outbreaks in Italy

International audience

anglais. Since 2016, north-eastern Italy has faced recurring outbreaks of highly pathogenic avian influenza (HPAI) viruses in both poultry and wild bird populations, with changing dynamics and unpredictable patterns. This variability prompted us to explore which drivers might have influenced the viral lineage dispersal history and transmission using a Bayesian phylodynamic framework.<p>We analysed Italian HPAI H5N8 and H5N1 genomic sequences from the majority of notified infected wild birds and poultry farms, alongside environmental and ecological data, including raster layers of population density, elevation, vegetation indices, and land cover types. Continuous phylogeographic analyses using a relaxed random walk diffusion model implemented in BEAST 1.10, and the "seraphim" R package were employed to estimate viral lineage dispersal velocity and location and assess the impact of drivers.</p><p>While none of the variables were strongly linked with dispersal velocity, population diversity indices, land cover types related to agricultural areas and wetlands, elevation and proximity to waterbodies were associated with dispersal locations, with varying statistical support across epidemic waves.</p><p>Our findings highlight key factors influencing HPAI dispersal, although potential drivers not yet considered may still play a role, such as wild bird movements, migration patterns, and the potential spread of HPAI through biological carriers like insects or rodents (biotic means) or through contaminated water, soil, and surfaces (abiotic means). Our study underscores the value of integrating genetic and epidemiological data to achieve a comprehensive understanding of HPAI ecology, which is essential to refine prevention and response strategies.</p>