New Approach Achieves 98% Accuracy in Detecting and Counting Fish in Hydropower Fish Passages

Researchers at Oak Ridge National Laboratory (ORNL) improved the resolution and accuracy of a noninvasive approach for detecting fish near hydropower facilities. This approach involves collecting and analyzing RNA and DNA (genetic material present in nearly all living organisms and released into the environment) from water samples collected in fish passage structures. 

Fish passages have many designs, such as ladders and lifts, and are designed to help fish move through hydropower facilities. By comparing the genetic signatures of fish found in water samples to visual counts of fish at fish passage viewing windows, the team was able to identify species and estimate the number of fish moving through a fish passage with 98% accuracy. These findings show that environmental RNA (eRNA) can be just as effective as manual methods of fish counting, thereby saving time, decreasing costs, and ensuring safety of the personnel tasked with manual collecting and counting while also supporting efforts to protect fish and their habitats. 

Counting fish is an important and often legal requirement that helps monitor and determine the environmental impacts of hydropower facilities on local ecosystems. Conventional biomonitoring often involves a time-consuming and expensive process of manually catching and identifying fish or passive monitoring through viewing windows that may be hindered by water clarity, daylight hours, and other conditions. 

To secure DNA and RNA samples, ORNL's researchers instead used a proprietary sampler to noninvasively filter water and collect environmental DNA (eDNA) and eRNA shed into the water from fish waste, reproduction, and death. Once the eDNA and eRNA is extracted, the genetic code is sequenced and compared to publicly available DNA and RNA sequences of known species to enable identification. 

ORNL conducted comprehensive sampling at five hydropower dams in Michigan over three months in coordination with the Michigan Department of Natural Resources. Researchers compared eDNA and eRNA data to video recordings to determine which fish species and how many were moving through fish passages. The results showed that fish estimates from eDNA samples were not well correlated with visual counts, but estimates from eRNA samples had a near perfect match. Since eRNA is only produced from living organisms and degrades within hours, it provides an accurate snapshot of which species and how many individuals were at a given location recently.

In the past year, ORNL presented findings on its work at the Sustainability in Hydropower Conference in Norway and the American Fisheries Society's annual meeting in North America. The team continues to build partnerships with state and federal agencies, such as the Texas Parks and Wildlife Department and the U.S. Fish and Wildlife Service, to expand their research to other hydropower facilities and aquatic systems. Currently, the team is conducting a study in coordination with the Tennessee Valley Authority, Electric Power Research Institute, and Stantec to look at broader aquatic community composition at Hiwassee Dam in North Carolina and to expand this monitoring approach to other animals and plants, including freshwater mussels, salamanders, and aquatic plants.