ORIGINAL RESEARCH article

Front. Vet. Sci.

Sec. Veterinary Epidemiology and Economics

Volume 12 - 2025 | doi: 10.3389/fvets.2025.1541019

Investigating the role of environmental factors in the French highly pathogenic avian influenza epizootic in 2022-2023

Provisionally accepted
  • 1Laboratoire de Ploufragan-Plouzané, Agence Nationale de Sécurité Sanitaire de l’Alimentation, de l’Environnement et du Travail (ANSES), Ploufragan, France
  • 2Agence Nationale de Sécurité Sanitaire de l’Alimentation, de l’Environnement et du Travail (ANSES), Maisons-Alfort, France

The final, formatted version of the article will be published soon.

The recurring epizootics of highly pathogenic avian influenza (HPAI) in France have been associated with changes in the epidemiological landscape, such as higher frequency of detections in wild birds and introductions into backyard farms. This highlights the need for a deeper understanding of the factors that drive the spread of HPAI, particularly environmental ones, which, unlike other factors, are still understudied. In this study, we examined various farm and environmental variables around the 2022-2023 outbreak sites in France to unravel potential common traits among detected outbreaks. From August 2022 to March 2023, 397 poultry farms were infected, including different species and production types. For each outbreak, the farm characteristics and variables related with their direct environment within a 2km radius were collected. Based on the Gower distance, accounting for qualitative and quantitative variable, clusters were identified using k-medoid partitioning algorithm. A random forest analysis was further used to hierarchize the relative role of each variable in the clustering process, to assess the importance of the farm structural and environmental conditions on the outbreak occurrence.To disentangle the impact of environmental factors from intrinsic herd characteristics, the method was applied twice: first, using the whole dataset including the farm characteristics and environmental variables (first scenario); second, accounting exclusively for the environmental variables (second scenario). Overall, farm variables such as farm type were crucial in the clustering process, overpassing most of the environmental factors, although the distance from "particular risk zones" and the coastline were also important. However, the clusters obtained with the second scenario that counts only for the environmental variables, remained consistent

Keywords: Poultry, Epidemiology, clustering, random forest, machine learning, Classification

Received: 06 Dec 2024; Accepted: 30 Apr 2025.

Copyright: © 2025 BEN SALEM, Andraud, Bougeard, Allain, Salines, Thomas, Schmitz, Saint-Cyr, Fiore, Le Bouquin and Scoizec. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Maryem BEN SALEM, Laboratoire de Ploufragan-Plouzané, Agence Nationale de Sécurité Sanitaire de l’Alimentation, de l’Environnement et du Travail (ANSES), Ploufragan, France

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.