Fine‐scale environmental heterogeneity shapes fluvial fish communities as revealed by eDNA metabarcoding. Issue 4 (20th August 2020)

• Record Type: Journal Article
• Authors: Berger, Chloé Suzanne; Hernandez, Cecilia; Laporte, Martin; Côté, Guillaume; Paradis, Yves; Kameni T., Dominique W.; Normandeau, Eric; Bernatchez, Louis
• Is Part Of: Environmental DNA. Volume 2:Issue 4(2020:Oct.)
• Page Start: 647
• Page End: 666
• Publication Date: 2020-08-20
• Subjects: environmental DNA -- fish communities -- metabarcoding -- river; DNA -- Periodicals; Biology -- Periodicals; Microbial ecology -- Periodicals; Biology; DNA; Microbial ecology; Electronic journals; Periodicals; 572.86
• DOI: 10.1002/edn3.129
• Languages: English
• ISSNs: 2637-4943

Abstract

Abstract: Conservation of freshwater biodiversity requires being able to track the presence and abundance of entire fish communities. However, studying fish community composition within rivers remains a technical challenge because of high spatial and temporal physico‐chemical variability, anthropic activities and connections with other river catchments, which may all contribute to important variations in local ecology and communities. Here, we used environmental DNA metabarcoding to document spatial variation in fish communities at a small geographic scale in a large river system. The study was conducted in the Contrecoeur sector (5.5 km long and approximately 1–1.5 km wide) of the St. Lawrence River (Québec, Canada), where two water masses with different physico‐chemical properties, known as "brown waters" and "green waters, " flow in parallel with limited admixing. Water samples were collected during two consecutive days at 53 stations located in both water masses. Using universal PCR MiFish 12S primers, Illumina MiSeq sequencing, and the Barque (www.github.com/enormandeau/barque ) eDNA analysis software developed by our group, a total of 67 fish species were detected. PERMANOVA and redundancy analyses (RDA) performed on relative read abundance revealed that each water mass comprised distinct communities that depended on turbidity, depth, and to a lesser extent on the upstream versus downstream position along the study area. eDNA metabarcoding results were compared with those of traditional surveys conducted previously in the sector and up to 40 km upstream of it. As previously reported, higher species diversity was detected by eDNA and with substantially lower sampling effort. Our results represent one of the few studies documenting the potential of eDNA metabarcoding to investigate small‐scale variation in 2D spatial patterns of distribution of whole fish communities associated with habitat characteristics within a lotic system. Abstract : We conducted a eDNA metabarcoding study in the Contrecoeur sector of the St. Lawrence River (Québec, Canada), which is only 5.5 km long and is characterized by two water masses with very different physico‐chemical properties that flow in parallel with limited admixing. We collected water samples during two consecutive days at 53 stations localized in both water masses. Using universal PCR MiFish 12S primers and Illumina MiSeq sequencing, we found variation in fish community composition within each water mass that was significantly associated with turbidity, depth, and to a lesser extent with the upstream‐downstream position along the study area. Our results demonstrate that it is possible to detect lateral (from shore toward center) variation in eDNA, and consequently in fish community composition at a small geographic scale in rivers, and that eDNA metabarcoding can efficiently complement traditional capture methods in order to more accurately document the role of physico‐chemical factors in shaping local fish community composition in large fluvial systems.