Mapping the ecological suitability of highly pathogenic avian influenza: what our new study suggests

Published on January 28, 2026, by Marie-Cécile Dupas

Our new study, just published in eLife, uses ecological niche modeling to explore how the environmental suitability of highly pathogenic avian influenza (HPAI) H5N1 and H5Nx may have shifted since 2020. Read the whole study here.

What we did: we built ecological niche models comparing two periods: 2015-2020 and 2020-2022, using over 25,000 outbreak records from the EMPRES-i database. We examined how environmental variables, such as poultry densities, land cover, proximity to water, correlated with outbreak locations in both wild and domestic birds.

What we observed: a shift in predictors for domestic bird outbreaks. We found that the relative influence of intensive chicken density and cultivated and managed vegetation increased notably post-2020, while eco-climatic factors such as land surface temperature and precipitation showed only moderate influence. This could suggest a shift in transmission dynamics; perhaps more farm-to-farm spread and fewer introductions from wild ducks. However, we cannot distinguish between index cases and secondary transmission in the EMPRES-i database, so this could not be confirmed quantitatively. It is also worth noting that intensive poultry density may partly reflect where surveillance is concentrated.

Expanded areas of ecological suitability: our risk maps identify regions in Europe, Asia, and the Americas as potentially suitable for HPAI circulation, with some expansion post-2020. Some of these predictions — particularly in regions with few reported outbreaks — represent spatial extrapolations and should be treated as hypotheses rather than confirmed risk zones. Notably, Australia appears suitable in our models yet has seen no H5N1 incursions despite extensive surveillance and millions of migratory birds arriving from affected regions.

A greater diversity of infected wild bird species: our diversity indices suggest more wild bird species — particularly seabirds like gulls, terns, and gannets — have been detected with HPAI post-2020. However, surveillance efforts in Europe increased substantially after 2020, which could influence these metrics.

What we concluded: the patterns we observe could help inform where surveillance and control measures might be prioritised. As H5N1 continues to affect new host species and regions, updating our understanding of its ecological niche remains important.

Figure 3: areas ecologically suitable for local H5N1 or H5Nx circulation leading to infection cases in domestic bird populations. We estimated the ecological suitability for two different time periods (2015–2020 and 2020–2022) and for both wild and domestic bird populations. Dynamic visualisations of the results are available here: https://mood-platform.avia-gis.com/core.

Reference: Dupas MC, Vincenti-González MF, Dhingra M, Guinat C, Vergne T, Wint W, Hendrickx G, Marsboom C, Gilbert M, Dellicour S (2025). Global risk mapping of highly pathogenic avian influenza H5N1 and H5Nx in the light of epidemic episodes occurring from 2020 onward. eLife 15: RP104748