The USGS National Water-Quality Assessment (NAWQA) Program monitors and addresses changes in the Nation’s water quality as a result of human activities and natural factors. Annual biological assessments of fish and invertebrate communities in surface waters provide important information about the relative stability and sensitivity of the biological integrity of these waters. However, traditional fish and invertebrate sampling protocols can be resource and labor intensive and cryptic and habitat specific species can be failed to be detected when present. Environmental DNA (eDNA) are DNA samples extracted from environmental samples, such as water. Using eDNA for metabarcoding surveys may be more cost effective, and potentially detect taxa not detected by traditional survey methods. The objective of our study was to compare the fish species detected by traditional survey methods versus eDNA metabarcoding and assess the relative sensitivity of each method. In the summer of 2015, 8-12 eDNA water samples per stream were collected in conjunction with traditional electrofishing surveys from five streams in Georgia, North Carolina, and South Carolina. Approximately 250-mL of water were filtered per sample, and extracted DNAs were screened at the cytochrome oxidase I (COI) marker, a mitochondrial gene commonly used for DNA metabarcoding. Amplified products were sequenced on an Illumina MiSeq, and sequence data were compared to a reference library with known DNA sequences to perform taxonomic assignments. Species detected with eDNA were then compared with the species detected during electrofishing surveys to assess the consistency of results between survey methods, determine which method was more sensitive, and determine if either method was associated with the detection of species was particular habitat tolerances
Results/Conclusions
The number of species detected from a combination of both methods ranged from 15-35. More than 50% of species detected at each site were detected by only one of the two methods. More species were detected with eDNA surveys than electrofishing at four of five sites, with the percent of species detected with eDNA only at these four sites ranging from 33-46% of the total number of species detected. In contrast, electrofishing only detected between 14-32% of the total species at a given site. While environmental DNA was effective at detecting some fish species in these streams, it did not detect all species detected by electrofishing, suggesting it may be a valuable additional tool for fish bioassessments versus a replacement of the traditional survey method for a comprehensive assessment of the fish communities.