Improving metabarcoding taxonomic assignment: A case study of fishes in a large marine ecosystem.

DNA metabarcoding is an important tool for molecular ecology. However, its effectiveness hinges on the quality of reference sequence databases and classification parameters employed. Here we evaluate the performance of MiFish 12S taxonomic assignments using a case study of California Current Large Marine Ecosystem fishes to determine best practices for metabarcoding. Specifically, we use a taxonomy cross-validation by identity framework to compare classification performance between a global database comprised of all available sequences and a curated database that only includes sequences of fishes from the California Current Large Marine Ecosystem. We demonstrate that the regional database provides higher assignment accuracy than the comprehensive global database. We also document a tradeoff between accuracy and misclassification across a range of taxonomic cutoff scores, highlighting the importance of parameter selection for taxonomic classification. Furthermore, we compared assignment accuracy with and without the inclusion of additionally generated reference sequences. To this end, we sequenced tissue from 597 species using the MiFish 12S primers, adding 252 species to GenBank's existing 550 California Current Large Marine Ecosystem fish sequences. We then compared species and reads identified from seawater environmental DNA samples using global databases with and without our generated references, and the regional database. The addition of new references allowed for the identification of 16 additional native taxa representing 17.0% of total reads from eDNA samples, including species with vast ecological and economic value. Together these results demonstrate the importance of comprehensive and curated reference databases for effective metabarcoding and the need for locus-specific validation efforts.

The importance of making testable predictions: A cautionary tale.

We found a startling correlation (Pearson ρ > 0.97) between a single event in daily sea surface temperatures each spring, and peak fish egg abundance measurements the following summer, in 7 years of approximately weekly fish egg abundance data collected at Scripps Pier in La Jolla California. Even more surprising was that this event-based result persisted despite the large and variable number of fish species involved (up to 46), and the large and variable time interval between trigger and response (up to ~3 months). To mitigate potential over-fitting, we made an out-of-sample prediction beyond the publication process for the peak summer egg abundance observed at Scripps Pier in 2020 (available on bioRxiv). During peer-review, the prediction failed, and while it would be tempting to explain this away as a result of the record-breaking toxic algal bloom that occurred during the spring (9x higher concentration of dinoflagellates than ever previously recorded), a re-examination of our methodology revealed a potential source of over-fitting that had not been evaluated for robustness. This cautionary tale highlights the importance of testable true out-of-sample predictions of future values that cannot (even accidentally) be used in model fitting, and that can therefore catch model assumptions that may otherwise escape notice. We believe that this example can benefit the current push towards ecology as a predictive science and support the notion that predictions should live and die in the public domain, along with the models that made them.