Protecting fishes in agricultural drains: evaluating Canadian policy using environmental DNA

Diverse aquatic communities inhabit agricultural drains, but maintenance of these drains can harm fishes and their habitats. In Canada, drains are classified based on their fish assemblages to protect sensitive species by managing drain-maintenance activities. Current Fisheries and Oceans Canada (DFO) fish sampling is challenged by environmental constraints and limited resources. To evaluate the effectiveness of current DFO methods to classify drains, we compared species detected by DFO methods (dip nets, electrofishing, seines) and by environmental DNA (eDNA) metabarcoding with two primers. eDNA detected more species ( n = 28) than the DFO methods ( n = 18) at 59% of drain segments, resulting in changes to 10 of 29 drain classifications. These results suggest that current DFO methods are insufficient to classify drains and, hence, adequately protect sensitive fishes and fish habitat within them. While we show the potential of eDNA to detect fish species in drains, both methods may be susceptible to imperfect detection resulting in incorrect drain classification. Therefore, studies should conduct occupancy modelling and cost analyses for both methods to determine the optimal sampling strategy for drain classification.