Environmental DNA detects a rare large river crayfish but with little relation to local abundance

Environmental DNA (eDNA) is DNA extracted from environmental samples (e.g. water) that can be used to infer presence or abundance of species, sometimes with greater sensitivity than conventional sampling methods. Previous eDNA applications to lotic ecosystems have shown promise in accurately inferring species presence, although studies attempting to estimate species abundance have had mixed results. This may be because eDNA applications in lotic environments are challenged by directional streamflow, which has the potential to transport detectable eDNA downstream from its source. Our study sought to evaluate whether results from eDNA corresponded well with the presence and abundance of the narrowly endemic, large river specialist crayfish Faxonius eupunctus obtained through a rigorous, well-tested conventional sampling method, or instead, if downstream eDNA transport in this large river system might overwhelm the effect of local species abundance. We used a species-specific quantitative PCR (qPCR) assay to amplify F.eupunctus eDNA collected in surface water samples from streus eDams within the Eleven Point River drainage, Arkansas and Missouri, U.S.A. We estimated F.eupunctNA detection probabilities and examined relationships between eDNA detection probability and site-scale variables using a hierarchical occupancy and detection probability modelling framework. Results from eDNA sampling showed 90% agreement relative to our conventional sampling method in estimating F.eupunctus presence, although eDNA failed to detect F.eupunctus eDNA at two upstream sites where conventional sampling detected F.eupunctus individuals. We found a poor relationship between F.eupunctus eDNA detection probability and local F.eupunctus abundance, and a strong relationship between eDNA detection probability and upstream river distance, which we used as a proxy for the risk of downstream transport of eDNA from upstream F.eupunctus populations. Our results demonstrate eDNA is a largely reliable tool for estimating the presence of benthic organisms in large, freshwater rivers. However, the likelihood of detecting F.eupunctus eDNA presence in our study increased as we moved down the stream network, even though local species abundances were greatest at more upstream locations. Therefore, the ability of eDNA to accurately reflect species presence or abundance in some lotic environments may be hindered by the downstream transport of detectable eDNA.