The true picture of environmental DNA, a case study in harvested fishponds

The application of environmental DNA (eDNA) metabarcoding has revolutionised large scale biodiversity monitoring of aquatic ecosystems. Validation studies have been performed mainly in laboratories and mesocosm experiments, however large-scale field experiments are necessary to verify the robustness of eDNA based monitoring for more specific applications and different environmental conditions. Here, eDNA samples were collected from three fishponds with high fish density and broad species diversity during summer and autumn. This sampling design included a large number of spatial replicates evenly spaced across the pond surface and samples from the inflow, while pooled samples were used to test the effect of filtration volumes on detectability. Most common species were detected using eDNA, but rare species were often missed out under these high stocking densities. Average read counts and site occupancy positively correlated strongly with species abundance and biomass, with the exception of samples affected by PCR inhibition. Higher diversity detections were observed in autumn compared to summer samplings and in running compared to standing water. Fish communities detected in pooled samples reflect the overall community structure, and the species detectability increases with higher filtration volumes. This work highlights how eDNA based surveys can be optimised based on sampling conditions to achieve the highest overall detection, which has important implications for applying this method to aid management and policy initiatives.