Data from: Using environmental DNA metabarcoding to monitor fish communities in small rivers and large brooks: Insights on the spatial scale of information

Monitoring fish communities is central to the evaluation of ecological health of rivers. Not only presence/absence of species is important to assess, but also the species composition of local fish assemblages is a crucial parameter. Lotic fish communities are traditionally monitored via electrofishing, characterized by a known limited efficiency and high survey costs. The use of environmental DNA-based analyses could serve as a non-destructive alternative, but this approach requires further insights in practical sampling schemes incorporating transport and dilution of the eDNA fragments; as well as optimization of molecular detection in terms of predictive power and quality assurance. By introducing fifteen species known to occur in Belgian waters via a controlled cage experiment, we aim to extend the knowledge on streamreach of eDNA in small rivers and large brooks, as laid out in the European Water Framework Directive's water typology. Introducing fish communities in two transects of a species poor river characterized by contrasting river discharge rates, we found strong and significant correlations between the eDNA relative abundances and the relative biomass per species in the cage community. Despite a decreasing correlation over distance, the underlying community composition remained stable over a distance of 300 m up to 1 km downstream of the cages, depending on the river discharge rate. Such decrease in similarity between relative source biomass and the corresponding eDNA-based community profile with increasing distance downstream from the source, can partly be attributed to variation in species-specific eDNA persistence. Our findings offer novel insights on eDNA behaviour and characterization of riverine fish communities. We conclude that water sampled from a relatively small river offers an adequate snapshot of the total fish community composition occurring within an upstream perimeter ranging between 300 and 1000 meters. The potential application for other river systems is discussed in this study.Data can best be accessed using Microsoft Excel and R. Funding provided by: Fonds Wetenschappelijk OnderzoekCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100003130Award Number: 1S23822NFunding provided by: Fonds Wetenschappelijk OnderzoekCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100003130Award Number: 1S01822NA controlled cage experiment was performed. Fifteen fish species were held in keepnets (~cages) in a small river in Belgium. At varying distances downstream from the nets, water samples were taken. Environmental DNA was extracted and samples were analysed via both droplet digital PCR (see Van Driessche et al., 2022) as well as via eDNA metabarcoding. This dataset includes the data of the eDNA metabarcoding. The entire set-up was repeated in two river transects with contrasting river discharge rates. Two levels of source fish biomass were used. Raw data was deposited on the NCBI's Sequence Read Archive (SRA) under BioProject number PRJNA904931. The bioinformatical pipeline as used on these raw read counts is available on Zenodo (https://zenodo.org/record/3731310#.Y8pdbXbMI2w). The OBITools software was used for further processing of the generated sequence data. The resulting count table as available here was used for further quality screening and cleaning, as well as for statistical analyses