Evaluating environmental DNA-based quantification of ranavirus infection in wood frog populations

A variety of challenges arise when monitoring wildlife populations for disease. Sampling tissues can be invasive to hosts, and obtaining sufficient sample sizes can be expensive and time-consuming, particularly for rare species and when pathogen prevalence is low. Environmental DNA (eDNA)-based detection of pathogens is an alternative approach to surveillance for aquatic communities that circumvents many of these issues. Ranaviruses are emerging pathogens of ectothermic vertebrates linked to die-offs of amphibian populations. Detecting ranavirus infections is critical, but nonlethal methods have the above issues and are prone to false negatives. We report on the feasibility and effectiveness of eDNA-based ranavirus detection in the field. We compared ranavirus titres in eDNA samples collected from pond water to titres in wood frog (Lithobates sylvaticus; n = 5) tadpoles in sites dominated by this one species (n = 20 pond visits). We examined whether ranavirus DNA can be detected in eDNA from pond water when infections are present in the pond and if viral titres detected in eDNA samples correlate with the prevalence or intensity of ranavirus infections in tadpoles. With three 250 mL water samples, we were able to detect the virus in all visits with infected larvae (0.92 diagnostic sensitivity). Also, we found a strong relationship between the viral eDNA titres and titres in larval tissues. eDNA titres increased prior to observed die-offs and declined afterwards, and were two orders of magnitude higher in ponds with a die-off. Our results suggest that eDNA is useful for detecting ranavirus infections in wildlife and aquaculture.