Using environmental DNA (eDNA) to assess the presence of cavefish and cave crayfish populations in caves of the Ozark Highlands

Many cavefishes and cave crayfishes are considered of conservation concern; however, sampling these species is inherently difficult given their occupied environments. The goal of our project was to verify the presence of select karst organisms while developing the foundation for sampling approaches that might be useful to conservation and management agencies. Our project objectives were to develop assays to amplify deoxyribonucleic acid (DNA) from several species of Ozark cavefishes and cave crayfishes and complete an initial surveillance of locations across the Ozark Highlands using environmental DNA (eDNA). Using DNA either provided by agency cooperators or that we extracted from tissue samples, we PCR amplified and then sequenced the Cytochrome Oxidase 1 (CO1) gene for cave crayfishes and the NADH Dehydrogenase Subunit 2 (ND2) gene for cavefishes. We developed species-specific primers and probes for five cave crayfishes and two cavefishes. From February 2017 to May 2017, we sampled 1–5 sampling units from 42 caves, wells, and springs (i.e., sites) using eDNA and traditional visual surveys. We measured physicochemical parameters at each sampling unit to estimate detection probability associated with both techniques. We also calculated two occupancy covariates for each site using geospatial data. We successfully amplified Troglichthys rosae DNA from the environment and detected DNA representing this species at 24 of 40 sites. At 16 of the sites where we detected T. rosae DNA, we did not visually observe the species. Although our assay for Typlichthys eigenmanni successfully amplified the target DNA from the environment, it also resulted in false absences where the species was visually confirmed. Using eDNA to detect cave crayfishes was much more difficult. The assay for Cambarus subterraneus did not work for eDNA samples and we were unable to pick up DNA from the environment, even at locations where it was visually confirmed. Alternatively, the eDNA surveys worked well for C. tartarus and we were able to amplify DNA at every site where it was visually observed. Our assay for C. aculabrum was based on a single sample obtained from GenBank,and did not amplify eDNA from field samples. Lastly, our eDNA results from samples in the known range of Orconectes stygocaneyi suggested the species may be found at an additional cave. Detection using eDNA based on our O. stygocaneyi assay was likely low because it was designed from a pseudogene; however, positive eDNA samples were sequenced to confirm species-specific DNA. Detection probability of both cavefishes and cave crayfishes varied by survey technique and was influenced by water volume, water clarity, water velocity, and substrate. Detection of cavefishes and cave crayfishes via visual surveys decreased when water volume increased, whereas detection using eDNA increased with greater water volume. Detection between taxa using either sample method was highest in habitats classified by fine substrates, except for eDNA detection of crayfishes which was greatest in coarse substrates. Detection of cavefishes increased with water clarity, but detection of cave crayfishes increased with turbidity. Detection probability of both cavefishes and crayfishes using eDNA increased slightly with water velocity, but decreased with visual surveys as water velocity increased. Occupancy by both taxa was positively related to particular geologic series. Crayfish occupancy was negatively related to fine-scale anthropogenic disturbance (i.e., 500-m buffer around the site), whereas crayfish showed no relationship with disturbance. Our results suggest possible range extensions, provide insights to factors driving detection using both sample techniques, and suggest areas where recharge zones may be shared among caves. Future efforts focused on a comprehensive evaluation of genetic diversity among cave crayfishes to improve assay design could improve detection and the applicability of eDNA as a supplemental and non-invasive sampling approach.