Environmental transmission of low pathogenicity avian influenza viruses and its implications for pathogen invasion

Abstract

Understanding the transmission dynamics and persistence of avian influenza viruses (AIVs) in the wild is an important scientific and public health challenge because this system represents both a reservoir for recombination and a source of novel, potentially human-pathogenic strains. The current paradigm locates all important transmission events on the nearly direct fecal/oral bird-to-bird pathway. In this article, on the basis of overlooked evidence, we propose that an environmental virus reservoir gives rise to indirect transmission. This transmission mode could play an important epidemiological role. Using a stochastic model, we demonstrate how neglecting environmentally generated transmission chains could underestimate the explosiveness and duration of AIV epidemics. We show the important pathogen invasion implications of this phenomenon: the nonnegligible probability of outbreak even when direct transmission is absent, the long-term infectivity of locations of prior outbreaks, and the role of environmental heterogeneity in risk.

Fig. 1.

Effects of pH and temperature on the environmental persistence time of H2N2 isolates, as measured by R_t (11). The quantity R_t denotes the number of days required for viral abundance to decline by one log₁₀ unit (12) and is directly related to our model parameter η in Eq. 1: η = log_e10/R_t. The surface represents a regression model fit to the data.

Fig. 2.

Illustration of an epidemic, assuming an entirely susceptible population (S₀ = 10⁴), with a single infected bird and 10² virus in the environment (measured in EID₅₀ per liter). Inset plots show details of the initial (Upper) and final (Lower) phases of the epidemic (note logarithmic scale on y axis). Epidemic takeoff and tail are shown to be determined by environmental transmission (green dash-dot line). Model parameters were κ = 10², ω = 10¹² per year, 1/η = 30 days, ρ/L = 1 per year, β = 0.006 per year per individual, 1/γ = 7 days. These parameters result in R₀^direct = 1.15 and R₀^env = 1.6 × 10⁷. Detailed descriptions of model parameters and sources for their numerical values are presented in Table 1. The figure was generated by integrating the mean field equations, described in SI Appendix.

Fig. 3.

AIV invasion success as a function of the direct transmission rate (β) and the rate of water consumption (ρ). (Top) Contours represent the probability of observing 20 cumulative infections, starting from S(0) = 10⁴, I(0) = 1, and V(0) = 10⁴. The very high frequencies in the contours were smoothed by using a convolution kernel. The black line demarcates the region R₀^direct = 1. Outbreaks to the left of this line, therefore, are mostly sparked by environmental transmission. (Bottom) Quantification of the mean number of environmental and direct transmission events per year. To illustrate direct transmission frequency when R₀^direct < 1, contours in the Bottom Right frame are plotted on a log₁₀ scale. For each combination of parameter values, 1,000 stochastic realizations were generated. Other model parameters as in Fig. 2, with L = 10⁷ and V₀ = 10². The sensitivity of these findings to changes in model parameter values are discussed in detail in the SI Appendix.

Fig. 4.

Invasion implications of environmental transmission. In A, we assume 10⁴ susceptible migrants arriving at a lake of volume L (liters), containing V₀ virus (measured in units of EID₅₀ per liter). The color surface shows the cumulative fraction of birds infected, averaged over 1,000 stochastic realizations. The transparent mesh presents the outcome in the worst-case scenario (top 1% simulations). In B, we present a contour plot of the probability of an outbreak resulting in > 100 infections. As the contours demonstrate, the outcome is largely independent of initial virus concentration, suggesting virus in the environment could represent a long-term source of infections. Insets demonstrate, however, that changes in V₀ affect the distribution of outbreak sizes. Model parameters were κ = 10, ω = 10¹² per year, 1/η = 1 month, ρ = 10⁴ liters per year, β = 7.8 × 10⁻³ per year per individual, 1/γ = 7 days. These parameters result in R₀^direct = 1.5. The sensitivity of these findings to changes in model parameter values are discussed in detail in the SI Appendix.