1. Prey metabarcoding has become a popular tool in molecular ecology for resolving trophic interactions at high resolution, from various sample types and animals. To date, most predator-prey studies of small-sized animals (<1 mm) have met the problem of overabundant predator DNA in dietary samples by adding blocking primers/peptide nucleic acids. These primers aim to limit the PCR amplification and detection of the predator DNA but may introduce bias to the prey composition identified by interacting with sequences that are similar to those of the predator. 2. Here we demonstrate the use of an alternative method to explore the prey of small marine copepods using whole-body DNA extracts and deep, brute force metabarcoding of an 18S rDNA fragment. 3. After processing and curating raw data from two sequencing runs of varying depth (0.4 and 5.4 billion raw reads), we isolated 1.3 and 52.2 million prey reads, with average depths of ~15 900 and ~120 000 prey reads per copepod individual, respectively. While data from both sequencing runs were sufficient to distinguish dietary compositions from disparate seasons, locations and copepod species, greater sequencing depth led to better separation of clusters. 4. As computation and sequencing are becoming ever more powerful and affordable, we expect the brute force approach to become a general standard for prey metabarcoding, as it offers a simple and affordable solution to consumers that are impractical to dissect or unknown to science.

The biodiverse Neotropical ecoregion remains insufficiently assessed, poorly managed, and threatened by unregulated human activities. Novel, rapid and cost-effective DNA-based approaches are valuable to improve understanding of the biological communities and for biomonitoring in remote areas. Here, we evaluate the potential of environmental DNA (eDNA) metabarcoding for assessing the structure and distribution of fish communities by analysing sediments and water from 11 locations along the Jequitinhonha River catchment (Brazil). Each site was sampled twice, before and after a major rain event in a five-week period and fish diversity was estimated using high-through-put sequencing of 12S rRNA amplicons. In total, 252 Molecular Operational Taxonomic Units (MOTUs) and 34 fish species were recovered, including endemic, introduced, and previously unrecorded species for this basin. Spatio-temporal variation of fish assemblages was detected, richness during the first campaign was nearly twice as high as in the second sampling round; though peaks of diversity were primarily associated with only four locations. No correlation between β-diversity and longitudinal distance or presence of dams was detected, but low species richness observed at sites located near dams indicates that these anthropogenic barriers might have an impact on local fish diversity. Unexpectedly high α-diversity levels recorded at the river mouth suggest that these sections should be further evaluated as putative “eDNA reservoirs” for rapid monitoring. By uncovering spatio-temporal changes, unrecorded biodiversity components, and putative anthropogenic impacts on fish assemblages, we further strengthen the potential of eDNA metabarcoding as a biomonitoring tool, especially in regions often neglected or difficult to access.