CONCERN - Well-defined objectives are essential for informing appropriate sampling design and effort. Assessing species abundance requires more intensive sampling than species presence, and testing for species absence requires even more (see GLFC STP - Guidance for environmental DNA sampling and effort).

CONCERN - If elements of the sampling design are not sufficient to meet the defined purpose/objective(s), this can be corrected (if feasible) by either revising the design before sampling begins, or filling gaps by additional sampling if appropriate. If neither is possible, communicate limits to data interpretation during project completion.

FOR YOUR CONSIDERATION - If elements of the sample design are not clear, seek clarification to ensure that spatial coverage, effort, sampling depths, and timing (single or repeated) are appropriate for meeting the survey objectives. Random, regular, or stratified sampling may be most appropriate for a survey depending on the objectives. Since species occupy different habitats, consideration should also be given to sampling at different depths and/or habitats to target specific species assemblages or life stages. Consideration should also be given to the timing of sampling effort(s) based on the survey objectives.

FOR YOUR CONSIDERATION - Although these are not essential, each of these approaches will help inform sampling effort and strengthen the interpretation value of eDNA results.

FOR YOUR CONSIDERATION - Paired sampling with eDNA and other sampling methods is not required, but is advisable for early stages of assessment methods to compare capture or detection efficiency of different sampling methods.

CONCERN - Well-established SOPs are essential for ensuring data quality and reliability at all stages, as well as for training field crews. In the absence of SOPs for all stages of data collection, particularly for steps taken to minimize or prevent sample contamination, eDNA results should be considered suspect.

CONCERN - Many environmental variables influence eDNA concentrations and detections (see GLFC STP- eDNA Agency Fact Sheet; Abbott et al. 2021; Ruppert et al. 2019).

CONCERN - If information on numbers and types of negative controls is not reported, the eDNA results should not be considered reliable.

CONCERN - Cross-contamination (or the presumption of cross-contamination) is an ever-present issue and it makes interpretation of the results very challenging. Agencies need to be very aware that strict adherence to sample collection and handling protocols is critical, especially with such a sensitive tool.

CONCERN - The physical layout of specialized PCR laboratories is designed to prevent contamination (see Abbott et al. 2021, Box 6, p. 13). It is highly advisable to have physically separated, dedicated work spaces for each stage of sample processing (sample storage, DNA extraction, qPCR setup, and amplification). Providing information on lab layout and sample workflow is relevant for assessing risks of false positive results. As eDNA metabarcoding is even more vulnerable to contamination than targeted eDNA assays, it is essential that extreme care be taken to avoid contamination and that SOPs describe precautions taken to avoid contamination.

CONCERN - Several approaches used together will significantly reduce the risk of contamination throughout the entire workflow (see Abbott et al. 2021, Box 6, p. 13). This is essential information for processing labs to provide in order for managers to assess the reliability of results. These precautions are especially important in labs that carry out other genetic or tissue-based analyses, as sample contamination is a higher risk and can be more difficult to detect in eDNA metabarcoding studies.

CONCERN - Reporting the number and type of negative controls used in the study, as well as the test results from these controls, are critical to data interpretation and acceptance (see Abbott et al. 2021, Glossary, p. 28 for types of "Blanks"). Including negative controls at multiple stages and reporting the results from them should be considered mandatory. If these data are not reported, the eDNA results should not be accepted.

CONCERN - This information is essential for robust, reliable interpretation of eDNA metabarcoding results.

CONCERN - Without sequence information and read depth (i.e., the number of sequence copies detected), eDNA metabarcoding should be questioned. False positive results can result from bioinformatic assumptions, unrecognized gaps in reference sequence libraries, and unrecognized limits of taxonomic resolution for an assay.

CONCERN - Information should be provided on which reference sequence databases (e.g., GenBank, EMBL, or comparable publicly accessible sequence databases) and matching criteria (e.g., homologous sequence lengths, percent similarity) were used to identify species.

CONCERN - The criteria used to document species presence/detection need to be clearly stated, as false positive results are a known challenge for eDNA metabarcoding.

CONCERN - Ruling out sample contamination by reporting a) sequence results from negative controls, and b) analysis details to check for sequencing artifacts (chimeric sequences, tag jumping, etc.) are important for improving confidence in the sequencing results. Spatial and temporal detection patterns (consistency among field replicate samples, temporal trends consistent with species ecology, movements, and habitat use, etc.) are similarly important for interpreting the results.

CONCERN - It is important that reference sequence datasets be as complete (geographically and taxonomically representative) as possible to ensure reliable matching of eDNA metabarcoding sequence data to their source species. For species with intraspecific (within-species) variation in the sequence amplified by the eDNA assay, the reference dataset should also include sequence variation within, as well as among, species.

CONCERN - Documentation of how data queries have been done is important for assessing whether appropriate analyses were performed. Reporting sequence read lengths and percent similarity of eDNA metabarcoding sequences to species in the reference database(s) is valuable to ensure putative matches are relevant and realistic.

CONCERN - The purpose and objectives for eDNA surveys should be clearly defined to help inform sampling design and interpretation of results. Including relevant spatial and temporal scales in defining objectives is very useful for informing the sampling design.

CONCERN - Reporting the sampling design (type, coverage, and effort) is essential for interpretation of eDNA results as well as using the data to inform management decisions and actions. Including negative controls at multiple stages is vital. Without these, the study should be considered invalid and the eDNA results should not be accepted.

FOR YOUR CONSIDERATION - It is critical for sampling design to reflect the study objectives in order to achieve informative outcomes. Confirming species absence requires more intensive sampling than abundance estimation, which in turn requires greater sampling effort than assessing species presence. The sampling design should include appropriate spatial and temporal scales, as well as appropriate sampling effort or intensity (number of replicates, repeat sampling, etc.) for each scale of interest.

FOR YOUR CONSIDERATION - Habitat occupancy modelling, detection probability estimation, or power analyses are not essential for designing eDNA sampling efforts, but are valuable for both informing sampling design and strengthening interpretation of results.

FOR YOUR CONSIDERATION - Paired sampling with eDNA and other sampling methods is not required, but is advisable for early stages of assessment methods to compare capture or detection efficiency of different sampling methods.

CONCERN - If relevant SOPs are not available, for new projects it is generally advisable to perform a pilot study to determine if sampling effort and methodologies will be effective in addressing the project's objectives. Note that this is a requirement for several national standards (e.g., U.S. federal agencies, Canada, Australia, and New Zealand).

CONCERN - The dilution and persistence rate (and thus detectability) of eDNA varies considerably depending on a number of environmental factors (see Abbott et al. 2021, Box 3, p. 9). By affecting eDNA concentration in environmental samples, variables such as water depth, flow, and direction—among others—can significantly influence detection outcomes and signal strength, and therefore are important to collect and report.

CONCERN - Because of the sensitivity of eDNA, field negative controls (or "field blanks") are essential to identify possible contamination of equipment or other sources that could lead to false positives (see Abbott et al. 2021, Box 5, p. 11; Glossary, p. 28). Including negative controls at multiple stages and reporting the results from them should be considered mandatory. If these are not included or reported, the eDNA results should not be accepted.

CONCERN - Adherence to strict collection and sample handling protocols is critical (see Abbott et al. 2021, Box 6, p. 13).

CONCERN - The physical layout of specialized PCR laboratories is designed to prevent contamination (see Abbott et al. 2021, Box 6, p. 13). It is highly advisable to have physically separated, dedicated work spaces for each stage of sample processing (sample storage, DNA extraction, qPCR setup, and amplification). Providing information on lab layout and sample workflow is relevant for assessing risks of false positive results.

CONCERN - Several approaches used together will significantly reduce the risk of contamination throughout the entire workflow (see Abbott et al. 2021, Box 6, p. 13). This is essential information for processing labs to provide in order for managers to assess the reliability of the results.

CONCERN - In addition to field blanks, negative controls should be inserted into each step of the eDNA workflow to identify the source of contamination, should it be introduced (see Abbott et al. 2021, Glossary, p. 28 for types of "Blanks"). Including negative controls at multiple stages and reporting the results from them should be considered mandatory. If these are not included or reported, the eDNA results should not be accepted.

CONCERN - If eDNA service providers are using proprietary assays for which all details are not disclosed, critical assay elements such as target specificity and sensitivity, along with supporting evidence, still need to be shared to permit objective interpretation of results (see Abbott et al. 2021, Appendix 3, p. 41). Published assays, even if validated by others in a particular geographic area (e.g., with a different suite of co-occurring species) and environmental sample types, still need to be verified (see see Abbott et al. 2021, Box 8, p. 19; Figure 3, p. 20). Reliable interpretation of eDNA data is dependent on robust validation of the eDNA assay (see eDNA Validation Scale).

CONCERN - Although reporting of some data is optional, there are key elements that need to be reported for meaningful interpretation of results (see Abbott et al. 2021, Section IVG, p. 23). As well as providing information on potential false positive results, reporting the data from the negative controls is needed for detection of the "Limit of Blanks" or LOB—e.g., "noise" vs. the true signal from known negative samples (see CSA W219:23 document and Lesperance et al. 2021).

CONCERN - Because criteria for acceptance of species detections will vary depending on project objectives, management risk tolerance, and other factors, they should be carefully considered and communicated (see Abbott et al. 2021, Box 10, p. 23; Figure 4, p. 25).

CONCERN - Documentation is critical. For example, if the amplicon has been sequenced and confirmed to be the target species, this helps rule out cross-reactivity of the assay with a co-occurring species.