Influenza A viruses (IAVs) are endemic in wild birds, but can spillover into poultry and cause serious economic harm. Quantifying infection kinetics is critical to developing predictive disease models aimed at understanding pathogen spread in an effort to prevent spillover. In this study, we evaluated whether IAV exposure dose mediates infection dynamics in mallards, a common IAV reservoir host. We experimentally inoculated 3 groups of 10 mallards with 103, 104, or 105 EID50 of an H6N2 IAV collected from North American waterfowl during surveillance operations. Each inoculated mallard was housed with 3 naïve contacts. We collected fine-scale viral RNA shedding information throughout the infection in a scheme designed to capture the eclipse, exponential growth, and waning phases of infection. On the day of inoculation, we collected oral and cloacal swabs from all individuals every 2 hours post exposure, then every 4 hours the next day, every 8 hours for the following two days, and daily through day 10. All samples were tested by qPCR. We compared viral RNA output curves by assessing viral RNA peak load, total (cumulative) load, peak day, and shedding period for each dosage group. We modeled log-transformed cumulative viral loads using an exponential asymptote function.
Results/Conclusions
In general, viral RNA shedding patterns varied across each of the metrics evaluated with significant individual heterogeneity evident across individuals. Specifically, for directly inoculated mallards 1) peak viral RNA loads increased with exposure dose (103.57 EID50 equivalents/mL for 103, 105.42 for 104, and 105.53 for 105), 2) total load increased with exposure dose (103.72 EID50 equivalents/mL for 103, 105.64 for 104, and 105.85 105), 3) peak day decreased with exposure dose (day 6 for 103, day 3 for 104, and day 2 for 105), and 4) the shedding period increased with exposure dose (4.54 days for 103, 6.23 days for 104, and 6.71 days for 105). Contact mallards showed similar patterns, but differences were more muted and infection characteristics were more variable. On average, the infection curves for mallards inoculated at 104 and 105 EID50 were more similar to each other than the infection curves of the birds infected at 103, suggesting a possible saturation effect at higher exposure doses. These results provide important information for the development of SIR (Susceptible-Infected-Recovered) models for predicting epizootic characteristics, not only by providing detailed data for parameterization of such models, but also by highlighting variation in infection characteristics across exposure conditions.