Assessing the Performance of Environmental DNA Metabarcoding in Fish Diversity Surveys of the Texas Gulf Coast

Abstract As natural and anthropogenic disturbances to coastal ecosystems increase in frequency and intensity, the need for efficient and cost-effective biodiversity survey methods also rises. We evaluated the efficacy of environmental DNA (eDNA) metabarcoding for profiling coastal fish diversity in Texas, USA. Our objectives were to (1) characterize coastal fish richness using metabarcoding at 18 sites across 580 km (360 miles) of Texas coastline and (2) compare eDNA results with traditional survey data. Simultaneously, we developed a novel filtration system that applied pre-filtration to increase processed water volume, and we explored relationships between filtered water volume, local environmental characteristics, and eDNA results. Based on Illumina MiSeq sequencing with MiFish Universal primers, we identified 61 fish species. When we compared eDNA detections (i.e., sequencing reads) with traditional sampling results based on Texas Parks and Wildlife Department Houston Advanced Research Center (HARC) surveys in 2008, we found no relationship (Spearman’s rank correlation p = 0.484). Rather, eDNA and traditional survey results shared 41 detections in common (species × site combinations); there were 59 species × site combinations for which traditional survey methods detected a species that eDNA did not, and 45 species × site combinations existed where eDNA detected a species that traditional surveys did not. Overall, our Texas case study suggests that eDNA metabarcoding provides a useful addition to the coastal biodiversity assessment toolbox, but rather than replacing traditional survey methods, a combination of eDNA and traditional approaches could serve to maximize the comprehensiveness of coastal fish surveys.