Seasonal dynamics and potential drivers of ranavirus epidemics in wood frog populations.

Epidemics in wildlife populations often display a striking seasonality. Ranaviruses can cause rapid, synchronous mass mortality events in populations of wood frog (Rana sylvatica) larvae in the summer. While there are several possible explanations for this pattern-from seasonal introductions of the virus to environmental stressors to windows of susceptibility to mortality from infection during development-most studies have focused on single factors in laboratory settings. We characterized the time course of ranavirus epidemics in eight ephemeral ponds in Connecticut, USA, measuring the prevalence and intensity of infections in wood frog larvae and Ranavirus DNA in water samples using environmental DNA methods. We found little evidence that the timing of pathogen introduction affected the timing of epidemics (rising prevalence) or the resulting die-offs. Instead, we observed a pulse in transmission asynchronous with die-offs; prevalence reached high levels (≥ 50%) up to 6 weeks before mortality was observed, suggesting that die-offs may be uncoupled from this pulse in transmission. Rather, mortality occurred when larvae reached later stages of development (hind limb formation) and coinciding water temperatures rose (≥ 15 °C), both of which independently increase pathogenicity (i.e., probability of host mortality) of infections in laboratory experiments. In summary, the strong seasonality of die-offs appears to be driven by development- and/or temperature-dependent changes in pathogenicity rather than occurring chronologically with pathogen introduction, after a pulse in transmission, or when susceptible host densities are greatest. Furthermore, our study illustrates the potential for eDNA methods to provide valuable insight in aquatic host-pathogen systems.