Infectious H5N1 viruses detected in the air and wastewater of dairy farms, along with evidence of covert infection of cattle, suggest that the outbreak may be spreading in more ways than previously recognized.
Research: Surveillance of California dairy farms reveals multiple possible sources of H5N1 influenza virus infection. Image credit: Parilov/Shutterstock.com
In a recent study published in PLOS Biologyresearchers conducted a comprehensive surveillance study of California dairy farms to assess the current status of highly pathogenic avian influenza (HPAI) H5N1 in California. The study tested multiple transmission routes (air, wastewater, etc.)., and cow breath) at 14 H5N1-positive dairy farms, revealing a previously overlooked route of virus transmission.
Expansion of dairy infections in California raises new concerns about infection
Highly pathogenic avian influenza (HPAI) H5N1 (clade 2.3.4.4b, genotype B3.13) was first detected in U.S. dairy cows in March 2024 and has since spread to 16 states. California, one of the nation’s leading dairy producing states, is one of the most affected states, with 771 herds reported testing positive for the virus by September 2025.
The initial consensus among experts was that the virus was spread primarily through direct contact with unpasteurized milk, a hypothesis supported by the observation of high viral loads in milk samples tested. However, subsequent reports of dairy workers contracting the virus after indirect contact and the detection of viral RNA on surfaces called into question earlier ideas, suggesting that the environment itself may be contaminated.
Furthermore, it remained unclear whether the H5N1 virus could remain infectious when aerosolized during milking procedures or whether reclaimed water used for pen cleaning and irrigation could act as a viable environmental reservoir.
Researchers track H5N1 at 14 dairy farms
This study aimed to address these knowledge gaps by conducting an extensive environmental and biological sampling program on 14 H5N1-positive farms in California. Air sample data for this study was obtained through three different air sampling techniques: (1) Open-face polytetrafluoroethylene (PTFE) filter cassette for personal exposure modeling at a flow rate of 5 L/min. (2) Handheld MD8 Airport sampler with gelatin filter at 50 L/min. (3) 200 L/min stationary AirPrep Cub 210.
At the same time, environmental water samples were collected from four locations along the reclaimed water flow: milk line cleaning, sump pump, fertilizer lagoon, and irrigated fields. The air and water samples thus obtained were genetically classified and quantified using real-time quantitative polymerase chain reaction (RT-qPCR) and digital droplet PCR (ddPCR).
Additionally, this study assessed within-animal virus distribution through a 7-day longitudinal study in 14 cows. Milk was collected separately from individual udder compartments. Finally, whole genome sequencing (WGS) analysis was used to identify and flag emerging virus variants in the air and water.
Infectious H5N1 detected in milking parlor air
The study’s surveillance efforts successfully detected H5N1 viral RNA in 21 of the 35 air samples analyzed. Alarmingly, air testing revealed that the milking parlor concentration (Cair) reached 104 genome copies per liter (gc/L) and four of the positive air samples contained infectious virus. In this study, low virus concentrations ranging from 4 to 41 gc/L were detected in breath samples taken from cattle rows.
The results of the wastewater analysis were also alarming, with all sampled locations (including the fertilizer lagoon) containing detectable viral contamination. Samples with virus concentrations above 650 gc/mL were selected for infectivity testing, and two samples were found to contain infectious virus.
Genome sequence analysis identified that air samples taken from farm “EC” carried the N189D mutation (H3 number 193) in the hemagglutinin (HA) protein. This is a site that previous studies have previously linked to altered receptor binding specificity and improved binding to human-like 2,6-linked sialic acids. However, the authors emphasized that it remains unclear whether this particular mutation increases human tropism or zoonotic risk.
Additionally, the study’s separate antibody analysis on farm FB revealed that 6 out of 10 cows that appeared perfectly healthy expressed H5-specific antibodies in their milk (evidence of past viral exposure and immune response), and did not exhibit mastitis (inflammation of the mammary glands) or significant reduction in milk production, traits typically used to diagnose infectious diseases.
Finally, the infection pattern in cow udders was surprisingly heterogeneous, rather than being limited to a consistent quarterly pattern. Rather than just a quarter of the cows becoming infected, as is often the case with bacterial mastitis, multiple infected cows often remained positive for the virus throughout the seven-day test period. This observation suggests that shared milking equipment may not be the only mechanism by which the virus moves from cow to cow.
Multiple transmission routes complicate outbreak containment
This study’s identification of infectious aerosols in milking parlors and persistent viral loads in wastewater lagoons indicates that these locations are important locations for biosafety interventions. The discovery of the N189D mutation further highlights the possibility that viruses can acquire adaptations related to mammalian or human receptor binding, although the functional significance of this mutation remains unclear.
The authors recommend that future efforts focus on rapid diagnosis at the cow level and the introduction of respiratory protection for workers in enclosed parlor spaces to reduce the risk of inhalation.
Click here to download your PDF copy.
