Conclusion
Our study demonstrates that a portfolio of metabarcoding markers
provides sufficient taxonomic coverage to detect and identify all of the
>100 marine and freshwater taxa collected worldwide and
combined here into heterogeneous and complex mixtures. Although
multi-locus metabarcoding has become increasingly common for
biodiversity assessments, where studies rely on different loci (COI,
18S, 16S, 12S) to resolve different taxonomic groups, our focus was to
employ multiple markers to provide greater taxonomic coverage and
increased taxonomic resolution across a single group: fishes (and
tangentially, other aquatic taxa). We recommend that investigators seek
to optimize their own portfolios of markers suited to the taxonomic and
geographic scope of their work, but our findings demonstrate that this
approach can enable powerful inferences about biodiversity.
Unfortunately, in order to achieve quantitative inferences about
relative abundance of each species in a complex mixture, we must
recommend further exploration of capture-based approaches (e.g., Mariac
et al., 2018) or leveraging genome-wide data (e.g., skim-seq; Bohmann,
Mirarab, Bafna, & Gilbert, 2020; Chua et al., 2021; Kobus et al.,
2020). Accurately detecting and then quantitatively accounting for the
relative abundance species in complex mixtures is a critical next step
for ecological and forensic studies using eDNA.