Navigating the Murky Waters: Variation in Low Abundance Sequence Filtration in Fish <scp>eDNA</scp> Metabarcoding

ABSTRACT Environmental DNA (eDNA) metabarcoding is a valuable tool for assessing fish communities and informing environmental management strategies. Well‐defined and informed methodologies are necessary to increase the repeatability and accuracy of eDNA results. This review evaluates 87 fish eDNA metabarcoding studies with a focus on low abundance sequence filtration methods used. This study aims to reveal the variety of methodological approaches used in eDNA metabarcoding and to provide recommendations based on these findings. A rubric of 32 criteria was developed to standardize the evaluation process, focusing not only on low abundance sequence filtration methods, but also on the incorporation of replicates, controls, primer validations, data availability, and other best practice criteria. We found diverse approaches to low abundance sequence filtering which showed little justification for threshold selection. While most studies incorporated some form of negative control or replicate, their implementation and reporting were inconsistent. There was also limited use of positive controls and primer validation throughout the studies. We recommend the adoption of various practices: (1) increasing the use of controls and replicates, (2) providing rationale behind low abundance filtering criteria or omitting it entirely, (3) completing analyses to validate primers, (4) improving the completion and appropriate communication of methods and results, and (5) making raw sequence data publicly available. Refined methodologies in eDNA metabarcoding research are imperative to ensure the reproducibility and accuracy of fish community assessment and environmental management practices. This is especially important as this tool continues to integrate into conservation and management efforts.