H5N1 Detected in Dead Waterfowl at a Major São Paulo Park
Dead white-faced whistling ducks in Ibirapuera Central Park yielded a divergent highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b virus in São Paulo, placing a busy urban park squarely inside the current panzootic. The finding matters because the site is not an isolated wetland or remote bird sanctuary. It is a dense human–animal interface, visited by thousands of people each day, where resident waterfowl, migratory birds, synanthropic mammals, domestic animals, and people share the same ecological space.
The detection came after unusual avian deaths were reported in the park in 2025. Laboratory testing of cloacal and tracheal swabs from the two dead ducks confirmed influenza A and H5N1 by reverse transcription real-time PCR. Whole-genome sequencing was then performed directly from the original samples, allowing investigators to move beyond a simple positive result and reconstruct the virus’s evolutionary path.
A Reassortant Virus Built From Regional and Imported Lineages
Genomic characterization showed a reassortant virus assembled from multiple influenza gene pools. The hemagglutinin, neuraminidase, polymerase acidic, and matrix protein segments belonged to Eurasian lineages, while the nonstructural protein segment fell within a North American lineage. The remaining internal genes, including polymerase basic 1, polymerase basic 2, and nucleoprotein, required closer phylogenetic resolution because they could not be cleanly assigned at first pass.
That deeper analysis tied the polymerase basic 2 and nucleoprotein segments to lineages from Colombia and the Brazilian Pantanal, with more than 94% nucleotide identity and tight clustering on phylogenetic trees. In practical terms, the virus was not a single imported entity frozen in time. It had mixed with locally circulating low pathogenicity avian influenza viruses already present in South American bird populations.
The biology here is familiar to influenza surveillance teams but still unsettling at the bedside of ecology: segmented viruses swap genome pieces when multiple strains infect the same host. Reassortment can alter host range, transmissibility, and fitness without requiring a wholly new viral family. In avian influenza, that process gives the virus a fast evolutionary shortcut.
One sentence captures the concern plainly: this was not a static lineage, but a moving genetic target.
Evidence Points to a Second North American Introduction
The hemagglutinin sequence added another layer. It shared 97.9% nucleotide identity with a 2023 Brazilian reference isolate and more than 98.6% identity with recent North American strains, suggesting divergence from the first lineage established in Brazil. Maximum-likelihood trees of the hemagglutinin and neuraminidase genes placed the São Paulo isolates in a monophyletic cluster among North American lineages from 2022 to 2025.
That pattern supports a second introduction of HPAI H5N1 into South America rather than simple local persistence of the earlier Brazilian strain. Bayesian temporal analysis estimated the time to the most recent common ancestor with North American H5 lineages at about mid-2023, with a mean tMRCA in May 2023 and a 95% highest posterior density interval from March 2023 to July 2023. The temporal signal aligns with genomic evidence: an introduced virus appears to have circulated for months after entering South America, reassorted with local avian influenza pools, and then reached urban wildlife in São Paulo.
The park detections therefore sit at the end of a longer ecological chain. Migratory birds likely moved the virus across geographic boundaries, while regional bird populations provided the setting for further mixing and diversification. The result is a virus that carries both imported and locally generated genetic features.
Urban Wildlife Surveillance Moves From Theory to Necessity
Ibirapuera Park is more than a city green space. It is a large recreational and cultural site that also functions as a shared habitat for multiple animal species. Once H5N1 was confirmed, park authorities implemented prevention, management, and isolation measures to reduce transmission risk. Those steps are immediate public health responses, but the case also exposes a broader surveillance problem: urban wildlife is often monitored less systematically than commercial poultry or well-defined wildlife reserves.
The finding strengthens the case for One Health surveillance that includes urban ecosystems, not only farms and remote conservation areas. In cities, spillover risk is shaped by close contact among wild birds, people, companion animals, and scavengers. A park lake becomes more than habitat; it becomes a mixing zone where infected carcasses, bird droppings, and shared water can facilitate onward spread.
This is where genomic surveillance pays off. Without sequencing, the event might have been logged only as another positive H5N1 detection. With sequencing, the virus reveals its ancestry, its reassortment history, and the timing of its arrival. Those details help public health and veterinary teams decide whether they are dealing with local persistence, a fresh introduction, or both.
Reassortment Signals Ongoing Viral Exchange Across South America
The São Paulo virus was not an isolated genetic oddity. Similar exchanges have already been described in other South American H5N1 genomes, and the present case adds another example of ongoing genetic exchange between introduced HPAI viruses and regional avian influenza viruses. That continuing exchange is why South America has become a critical region for surveillance.
The risk extends beyond bird mortality. Cross-species transmission remains a constant possibility wherever infected wild birds overlap with domestic animals and human activity. Highly pathogenic avian influenza can move unpredictably across species boundaries, and urban parks may serve as overlooked waystations in that process. The detection in white-faced whistling ducks confirms that even well-trafficked city environments can host viruses with complex evolutionary histories.
For clinicians and public health specialists, the practical consequence is straightforward: surveillance systems must be able to recognize genomic divergence quickly, not just detect a known subtype. The faster a reassortant virus is identified, the faster authorities can intensify wildlife monitoring, evaluate local animal health risks, and assess whether additional spillover investigations are needed.
The unanswered question is not whether the virus can keep evolving. It already has. The remaining challenge is how quickly surveillance can keep pace when that evolution happens in the middle of a city park.