Traditional biological surveys for the detection and quantification of fish assemblages are invasive, time consuming and costly. The recent further development of environmental DNA (eDNA) sequencing has become a promising modern alternative to the traditional techniques. eDNA has the opportunity to streamline fish conservation and monitoring in a variety of habitats. Succeeding in both marine and freshwater ecosystems this method is still considered untried in many subsystems, including turbid neotropical systems.
In this study, we compare the use of traditional fish survey methods (fyke netting, electro-fishing, and box trapping) to modern fish survey methods (eDNA collection) relative to species detection and abundance estimates in a turbid river system in South America.
eDNA yielded higher detection rates than traditional capture methods over two survey years (2018 and 2019), though in 2019 could only account for approximately 60% of detected species. Ordination clearly separated fish assemblages (presence/absence) detected by traditional and modern techniques, indicating that neither technique alone provides sufficient representation of fish communities present. Review of capture data indicated that electro-fishing coupled with eDNA can provide the most efficient and representative fish survey methods in turbid neotropical systems. Together these techniques accounted for approximately 96 and 97% of all detected fish species in 2018 and 2019, respectively.
A correlation between eDNA read number and capture abundances was found concerning species that dominated capture abundances. We suggest that eDNA metabarcoding could be used as a rudimentary tool to provide information on the relative rank abundance of fish.
Our study demonstrated that eDNA is another fish monitoring technique that should be used in combination with other traditional fishing methods to appropriately characterise fish assemblages in turbid neotropical systems. Specifically, eDNA in combination with electro-fishing can provide the most enhanced detection rates and therefore increase confidence in monitoring results.