Shedding light on avian influenza H4N6 infection in mallards: modes of transmission and implications for surveillance

Abstract

Background: Wild mallards (Anas platyrhychos) are considered one of the primary reservoir species for avian influenza viruses (AIV). Because AIV circulating in wild birds pose an indirect threat to agriculture and human health, understanding the ecology of AIV and developing risk assessments and surveillance systems for prevention of disease is critical.

Methodology/principal findings: In this study, mallards were experimentally infected with an H4N6 subtype of AIV by oral inoculation or contact with an H4N6 contaminated water source. Cloacal swabs, oropharyngeal swabs, fecal samples, and water samples were collected daily and tested by real-time RT-PCR (RRT-PCR) for estimation of viral shedding. Fecal samples had significantly higher virus concentrations than oropharyngeal or cloacal swabs and 6 month old ducks shed significantly more viral RNA than 3 month old ducks regardless of sample type. Use of a water source contaminated by AIV infected mallards, was sufficient to transmit virus to naïve mallards, which shed AIV at higher or similar levels as orally-inoculated ducks.

Conclusions: Bodies of water could serve as a transmission pathway for AIV in waterfowl. For AIV surveillance purposes, water samples and fecal samples appear to be excellent alternatives or additions to cloacal and oropharyngeal swabbing. Furthermore, duck age (even within hatch-year birds) may be important when interpreting viral shedding results from experimental infections or surveillance. Differential shedding among hatch-year mallards could affect prevalence estimates, modeling of AIV spread, and subsequent risk assessments.

Figure 1. Mean shedding of H4N6 AIV RNA by mallards infected via oral inoculation or water transmission.

(A) Cloacal shedding, (B) Oropharyngeal shedding. RRT-PCR was used to detect and quantify viral RNA in cloacal and oropharyngeal swabs. Values were extrapolated from a standard curve and presented as PCR EID₅₀ equivalents/mL on a log scale. Each point represents the arithmetic mean of 6 mo mallards (n = 11) or 3 mo mallards (n = 12). On average, the shedding curves differed between 6 mo and 3 mo mallards. The dashed line represents our RRT-PCR threshold of detection as 10⁰ PCR EID₅₀ equivalents/mL.

Figure 2. AIV RNA detected in water tanks.

AIV RNA was detected and quantified by RRT-PCR in the water tanks housing orally-inoculated mallards (represented by solid markers). Quantities were extrapolated from a standard curve and presented as PCR EID₅₀ equivalents/mL on a log scale. On 5 dpi, mallards were removed from two of the pens (69 and 70). These pens were disinfected, leaving only the AIV contaminated water in water tanks. When naive ducks were exposed to the AIV contaminated water, they became infected and maintained the virus concentration (represented by open markers) for the duration of the experiment. The dashed line represents our RRT-PCR threshold of detection as 10⁰ PCR EID₅₀ equivalents/mL.

Figure 3. Mean H4N6 AIV RNA detected using 3 different sampling methods.

AIV RNA quantities detected in cloacal swabs, oropharyngeal swabs, and fecal samples were compared. Because fecal samples were originally weighed (not swabs), we determined that an average fecal swab contained 0.0929±0.0089 g of feces and present these data as ‘fecal swab equivalents.’ Estimated virus concentrations in fecal swab equivalents were higher than the concentrations detected in cloacal and oropharyngeal swabs regardless of the age of ducks or route of infection. Error bars represent standard errors.

Figure 4. Comparison of probability of detecting H4N6 AIV RNA using 3 different sampling methods.

Fecal samples show a significantly higher probability of detection 4–8 dpi. Overall, the predicted probability of detecting AIV for a single swab is 38% for a cloacal swab, 44% for an oropharyngeal swab, and 52% for a fecal swab equivalents.