Background/Question/Methods
Wild mallards are one of the primary reservoir species for avian influenza viruses. Therefore, understanding their virus shedding characteristics is essential for modeling transmission and spread of avian influenza viruses and for optimizing surveillance strategies to detect these viruses.
We experimentally infected mallards with low pathogenic avian influenza viruses (LPAIV) to investigate virus shedding characteristics. Our objectives were 1) to determine shedding duration and viral loads in oral-pharyngeal swabs, cloacal swabs, and feces, 2) to compare shedding characteristics in juvenile and mature mallards, and 3) to investigate the impact of coinfection with multiple sub-types on shedding rates.
In one experiment, we infected 23 mallards with a LPAIV, H4N6 (A/wildbird/PA/185996-06/07), isolated from wild bird feces. In a second experiment, we co-infected 18 mallards by inoculating them with one of three LPAIV subtypes and then exposing infected individuals to viruses shed by ducks infected with a different subtype. We collected cloacal swabs and oral-pharyngeal swabs from each duck as well as fresh feces from the floor of each pen daily for approximately nine days. We tested all samples with a real time RT-PCR assay developed to detect all avian influenza subtypes.
 Results/Conclusions
Each infected mallard shed virus and exhibited an antibody response to the infection. Viral loads varied significantly between oral-pharyngeal, cloacal, and fecal samples with viral loads significantly higher in fecal samples. On average, virus was detectable for two days longer for fecal versus oral-pharyngeal or cloacal samples. Based on logistic regression analysis, virus was 8% more likely to be detected in fecal samples compared to oral-pharyngeal swabs for samples collected during an eight day period post infection. Similarly, virus was 14% more likely to be detected in fecal samples compared with cloacal samples. Viral loads were significantly higher for older ducks compared to juveniles. For example, the mean total virus detected over eight days for cloacal swabs was 10^4.1 EID₅₀ for mature ducks and 10^1.5 EID₅₀ for juvenile ducks. Preliminary results indicate a synergistic impact on virus shedding when an individual is co-infected with multiple sub-types.
These results indicate that focusing surveillance efforts on fecal samples may improve the likelihood of detecting LPAIV in wild bird populations. In addition, differences in shedding characteristics between mature and juvenile mallards indicate that prevalence estimates may be biased if differential detectability between the two groups is not incorporated into the estimate of prevalence.
dicate estimates of prevalence need to adjust for differential detectability.