Across the Pan-Tibetan Plateau, highly pathogenic avian influenza does not follow the same geography in wild birds and domestic poultry. A new 20-year analysis shows that infection risk in migratory birds is shaped mainly by water, terrain and climate, whereas poultry outbreaks are much more closely associated with economic activity, transport networks and production systems. The findings suggest that treating both host populations as a single epidemiological system can hide important areas of wildlife risk.
The study analysed highly pathogenic avian influenza (HPAI) outbreaks recorded between 2004 and 2024 across Xizang, Qinghai, Gansu, Sichuan, Yunnan and Xinjiang.
This vast region is particularly important for avian influenza ecology because three major global migratory routes cross it: the West Asian–East African, Central Asian and East Asian–Australasian flyways. The Pan-Tibetan Plateau has also remained an important centre of HPAI activity since the major wild-bird mortality event at Qinghai Lake in 2005.
According to the researchers, approximately 30% of China's reported HPAI outbreaks in wild birds and 10% of poultry outbreaks have occurred within the region. Since 2020, more than half of China's reported wild-bird outbreaks have originated there.
The researchers assembled 137 confirmed outbreaks, including 26 in wild birds and 111 in domestic poultry, and combined them with 34 climatic, environmental and socioeconomic variables at a spatial resolution of 2 × 2 km.
Wild birds followed the landscape
More than 93% of predicted high-risk areas for wild birds were concentrated in Xizang, Qinghai, Gansu and Xinjiang. Risk was driven mainly by natural features. The strongest variables included terrain slope, distance to surface water, winter precipitation, temperature during the driest period and vegetation cover.
Infection probability declined rapidly on slopes steeper than about 16 degrees and fell sharply as birds moved farther from water, approaching very low levels beyond approximately 45 km from aquatic habitats.
These patterns fit the ecology of migratory waterbirds, which concentrate around lakes, wetlands and stopover areas where viral transmission can be amplified. Wild-bird outbreaks also showed a distinctive seasonal pattern, corresponding broadly to spring migration and breeding activity.
Poultry outbreaks followed roads and economic activity
Around 70% of high-risk poultry areas were located in Yunnan, Sichuan and Gansu and around provincial capitals. Here, the dominant predictors were socioeconomic rather than ecological.
Gross domestic product and proximity to railway and road networks were particularly influential, suggesting that poultry density, trade, transport and movement between farms play a major role in shaping transmission risk. The model identified elevated poultry risk within approximately 81 km of railways, 38 km of national highways, 54 km of provincial highways and 33 km of county roads.
The temporal pattern was also different. Poultry outbreaks showed one increase around June and July and another broader peak extending through late autumn and winter.
Strikingly, the historical pattern changed after 2020. Before then, poultry outbreaks dominated regional records, but no domestic poultry outbreaks were recorded in the study area between 2021 and 2024, while wild-bird infections continued to be reported.
A combined model missed part of the wildlife risk
Because poultry outbreaks outnumbered wild-bird records by more than four to one, the mixed model increasingly resembled the poultry model and underestimated infection risk in wild birds. Separate models for wild birds and poultry both performed strongly, with more than 95% of validation points falling within predicted high-risk areas. This suggests that combining host species simply to increase the number of observations can produce a misleading risk map when the underlying transmission processes are different.
Where the two systems overlap may matter most
Some high-risk zones nevertheless overlapped, particularly in Gansu, Xinjiang, Xizang and Qinghai. The authors highlight these areas because expanding poultry production and fragmented wetlands may increase opportunities for contact between domestic birds and migratory wildlife.
These interfaces could be particularly important for surveillance because they combine two distinct transmission systems: one driven mainly by bird ecology and another by poultry production and connectivity.
The study therefore argues for host-specific HPAI surveillance rather than a single regional model. For wild birds, surveillance should concentrate on wetlands, migration corridors and seasonal aggregation sites. For poultry, the priorities shift toward production density, trade routes, transport hubs and areas of intense economic activity.
Commentaires
No comments yet.
Sign in to comment