Fate and Transport of Vertebrate DNA in Surface Water Environments: Developing a Basis for Quantification through Environmental DNA Monitoring

Abstract

Many environmental engineering applications require sampling of DNA to be effective. Organisms shed DNA into the environment, and that environmental DNA (eDNA) can be collected, extracted, amplified, and quantified to provide information about the shedding organisms. Regardless of the target organism, shed eDNA is subject to environmental degradation and partitioning into various compartments. Fate and transport of eDNA in aquatic systems was examined using two vertebrate organisms (invasive bigheaded carps, Hypophthalmichthys spp., and endangered Topeka shiners, Notropis topeka). Target eDNA concentrations were higher in sediment than the overlying water column and were correlated with biomass density of the target fish. Target eDNA in sediments was detectable and quantifiable at least 132 days after removal of the fish, suggesting eDNA persistence in aquatic systems may be significantly longer than previously reported. Degradation rates were highly correlated with initial eDNA concentrations in ponds with bioturbation and reduced macrophytes, suggesting degradation may be dominated by enzymatic hydrolysis in those systems. N. topeka DNA from museum specimens was extracted and sequenced for cytochrome oxidase 1 (COI), cytochrome oxidase b (CytB), NAD dehydrogenase 2 (ND2), and the control region (D-loop). Such data are rare, and these tissue extractions and sequences represent important contributions to preservation of an endangered species. Both detection probability and concentration of N. topeka eDNA in water samples increased at biomass densities two to three times larger than naturally occurring schools (80 versus 20 to 30 fish). After an initial spike, fish eDNA in water dropped below detection limits within 7 days, regardless of stocking density. However, at 14 days detection in high density tanks increased to initial spike levels and remained so at 26 days after stocking. Fish eDNA was detected in water 27 days after fish removal, regardless of density, though at lower detection probability than during fish presence. Despite consistent detection, water column concentrations of fish eDNA did not exceed 20% of the initial spike over the course of 335 days. eDNA monitoring must account for partitioning and differential fate in the environment. Therefore, future studies should focus on models incorporating resuspension, ecological condition, and the mechanisms of degradation.