Insights into the representativeness of biodiversity assessment in large reservoir through eDNA metabarcoding

technology and methods have evolved over time from identifying individual organisms based on specific DNA barcodes to the current capability to analyze hundreds of thousands of DNA barcodes obtained from environmental samples (e.g. water, soil, air), allowing the description of entire communities in terms of species composition and dynamics [4]. This high-through- put approach, known as ’DNA metabarcoding,’ is especially potent at uncovering biodiversity [e.g. 5–8]. Monitoring biodiversity through environmental DNA (eDNA) poses enormous challenges, especially in aquatic systems, due to the vast scales involved, and the combination of environ- mental conditions such as water current speed and density, wind, and tides, which influence particle transport [9]. Hydrodynamic modeling is an alternative approach for investigating the environmental influence on the transport of molecules. Such modeling approaches offer physi- cal insights into the water body by considering outflow volume, water renewal rate, current velocity, and flow direction. This information together can enhance our understanding of the fundamental processes affecting the spatial and temporal (including seasonal) distribution of eDNA [10–12]. In recent years, eDNA has gained acceptance due to its non-intrusive nature and cost-effec- tiveness [13, 14]. It has proven highly effective because eDNA metabarcoding studies involve multiple sampling sites with several replicates at each location, and, based on the assumption that eDNA in aquatic systems is dispersed by water flow and begins to degrade over time,

increasing the number of sampling sites enhances the detection of a broader range of species [15, 16]. This approach addresses many concerns scientists and regulators have regarding the effectiveness of traditional environmental programs, which may struggle with issues related to longevity, continuity, scale, or statistical significance [17, 18]. Several approaches can help assess the optimal experimental design for a given study area. For instance, rarefaction curves can reveal the correlation between sampling size and the number of sequences per taxon (or Molecular Operational Taxonomic Unit, MOTUs), demonstrating the effectiveness of both the number of samples collected and the molecular approach [19]. Although this analysis is often employed in eDNA studies to determine whether the number of sampling sites or the molecular protocol sufficed to capture diversity [20–22], it has rarely been used to assess the optimal number of samples in large reservoirs with hundreds of square kilometers [23, 24]. In this context, multivariate analyses, such as Principal Component Analysis (PCA) and Discrim- inant Analysis of Principal Components (DAPC), can offer insights into the distribution of variation across all sites, indicating which samples contribute most to the overall variability [19]. Finally, diversity analyses, including richness, Shannon’s H index, and species frequency computation, can complement these findings by providing visualizations that enable the rec- ognition of community patterns and differences. These combined methods can inform how

comprehensive the experimental design and the applied molecular approach were. The aim of this study was to build a large-scale biodiversity inventory of a HPP reservoir utilizing eDNA and compare its findings to the previous reports either published in the scien- tific literature or filed over decades with regulators. The robustness of the inventory, as well as the most cost-effective experimental design and optimal number of samples, were investigated. Toward this goal, hydrodynamic modeling was initially conducted at an 800 km2 HPP reser- voir, Três Irmãos (TI, São Paulo, Brazil), coupled with pilot sampling at a single site to gain insight into local diversity using eDNA metabarcoding and three distinct markers (COI, 16SrRNA and 12SrRNA). Subsequently, multiple stations were sampled across the reservoir to evaluate large-scale biodiversity. Multivariate analyses, rarefaction curves, and community comparisons were employed to assess the effectiveness of the sampling design in compiling a comprehensive biodiversity inventory of the reservoir. PLOS ONE eDNA metabarcoding reservoir PLOS ONE | https://doi.org/10.1371/journal.pone.0314210 January 24, 2025 2 / 22 Competing interests: The authors have declared that no competing interests exist.