Assessing the breeding phenology of a threatened frog species using eDNA and automatic acoustic monitoring

Abstract

Background Climate change has driven shifts in breeding phenology of many amphibians, causing phenological mismatches (e.g., predator-prey interactions), and potentially population declines. Collecting data with high spatiotemporal sensitivity on hibernation emergence and breeding times can inform conservation best practices. However, monitoring the phenology of amphibians can be challenging because of their cryptic nature over much of their life cycle. Moreover, most salamanders and caecilians do not produce conspicuous breeding calls like frogs and toads do, presenting additional monitoring challenges. Methods In this study, we designed and evaluated the performance of an environmental DNA (eDNA) droplet digital PCR (ddPCR) assay as a non-invasive tool to assess the breeding phenology of a Western Chorus Frog population (Pseudacris maculata mitotype) in Eastern Ontario and compared eDNA detection patterns to hourly automatic acoustic monitoring. For two eDNA samples with strong PCR inhibition, we tested three methods to diminish the effect of inhibitors: diluting eDNA samples, adding bovine serum albumin to PCR reactions, and purifying eDNA using a commercial clean-up kit. Results We recorded the first male calling when the focal marsh was still largely frozen. Chorus frog eDNA was detected on April 6th, 6 days after acoustic monitoring revealed this first calling male, but only 2 days after males attained higher chorus activity. eDNA signals were detected at more sampling locales within the marsh and eDNA concentrations increased as more males participated in the chorus, suggesting that eDNA may be a reasonable proxy for calling assemblage size. Internal positive control revealed strong inhibition in some samples, limiting detection probability and quantification accuracy in ddPCR. We found diluting samples was the most effective in reducing inhibition and improving eDNA quantification. Conclusions Altogether, our results showed that eDNA ddPCR signals lagged behind male chorusing by a few days; thus, acoustic monitoring is preferable if the desire is to document the onset of male chorusing. However, eDNA may be an effective, non-invasive monitoring tool for amphibians that do not call and may provide a useful complement to automated acoustic recording. We found inhibition patterns were heterogeneous across time and space and we demonstrate that an internal positive control should always be included to assess inhibition for eDNA ddPCR signal interpretations.