Pathway to Increase Standards and Competency of eDNA Surveys (PISCeS)—Advancing collaboration and standardization efforts in the field of eDNA

Twelve years have passed since the first published study using environmental DNA (eDNA) for conservation purposes—detection of the invasive American bullfrog (Lithobates catesbeianus) in controlled and field conditions as proof of concept (Ficetola, Miaud, Pompanon, & Tab erlet, 2008). Since then, the number of published studies describing species presence/absence detections using eDNA analysis has been increasing at a rapid rate (Tsuji, Takahara, Doi, Shibata, & Yamanaka, 2019). eDNA analyses have been applied in diverse systems (including terrestrial, freshwater and marine environments) and to answer a variety of questions, from the detection of target organisms using quantitative PCR (qPCR) (e.g. Buxton, Groombridge, Zakaria, & Griffiths, 2017; Currier, Morris, Wilson, & Freeland, 2018; Davy, Kidd, & Wilson, 2015; Hunter et al., 2015; Lacoursière-Roussel, Rosabal, & Bernatchez, 2016), droplet digital PCR (ddPCR) (e.g., Baker, Steel, Nieukirk, & Klinck, 2018; Doi et al., 2015; Evans, Shirey, Wieringa, Mahon, & Lamberti, 2017; Mauvisseau et al., 2019), and even gene editing methods (e.g., CRISPR-Cas, Williams et al., 2019) to community structure studies using metabarcoding with high-throughput sequencing (HTS) (e.g., Balasingham, Walter, Mandrak, & Heath, 2018; Günther, Knebelsberger, Neumann, Laakmann, & Martínez Arbizu, 2018; Kelly, Port, Yamahara, & Crowder, 2014; Leduc et al., 2019; Valentini et al., 2016). Recently, the field of eDNA has gained more attention from conservationists, recognizing not only its benefits, but also remaining challenges. For instance, although many methodological approaches have been developed for collecting and processing eDNA samples (Thomas, Howard, Nguyen, Seimon, & Received: 25 May 2020 | Accepted: 28 May 2020 DOI: 10.1002/edn3.112

, improved understanding of the ecology of eDNA-its origin, state, transport, and fate (Barnes & Turner, 2016)-is needed to inform best practices of eDNA collection, analyses, and study design. Ultimately, a better understanding of eDNA ecology and developing best practices will aid to predict taxa presence/absence and distributions. Despite this and other challenges (e.g., PCR inhibition), the application of eDNA-based technology is already beginning to influence conservation and management decisions (e.g., Kelly, Port, Yamahara, Martone, et al., 2014). However, there is considerable work to be done before eDNA surveys can be fully accepted by policy makers.
Only when eDNA surveys become broadly accepted by regulatory agencies will the industry sector justify directing budgets away from conventional field surveys. Hence, a continued collaboration between researchers, regulators, and other stakeholders is necessary to promote methods standardization and a framework to assess the quality of methods employed in order to support the acceptance of eDNA surveys (Helbing & Hobbs, 2019).

| VALIDATI ON OF EDNA DE TEC TI ON S
A still-growing area of eDNA research is the proof of concept for targeted and metabarcoding analytical approaches for detecting specific species and whole communities, respectively. Gasparini et al. (2020) validated the utility of an eDNA qPCR assay to monitor the imperilled wavy-rayed lampmussel (Lampsilis fasciola). Using caged L. fasciola individuals placed in an area where the species was known to be absent, they evaluated eDNA detection rates at several downstream sampling distances. Their study demonstrated that unionid mussels could be successfully detected in lotic systems (Gasparini et al., 2020). However, the authors warn of reduced eDNA detection when species abundance is low and/or when sam-  to produce eDNA species detections in as quick as one hour (Thomas et al., 2019) in a field setting. Their technique was validated in laboratory and field conditions using the invasive freshwater New Zealand mudsnail (Potamopyrgus antipodarum) as target species. According to their study, the on-site eDNA-Sampler/Biomeme system generates consistent results comparable to a laboratory-based protocol (i.e., both protocols achieved target detectability at the same eDNA dilution levels) (Thomas et al., 2019). Similarly, Skinner et al. (2020) conducted mesocosm experiments to determine the accuracy and precision of eDNA detections of a marine species-at-risk (Striped bass Morone saxatilis) using a handheld qPCR device compared with a conventional laboratory-based qPCR protocol. In their study, the authors demonstrated the success of both platforms in detecting eDNA from the target species (Skinner et al., 2020). However, unlike Thomas et al. (2019), Skinner et al. (2020) showed lower estimates of eDNA concentration from the portable handheld compared with the laboratory-based qPCR platforms. In their study, the authors also used a time series experiment to determine the persistence/decay rate of eDNA signal demonstrating that striped bass eDNA can be reliably detected in a water sample within 24 hr of collection. Lastly, despite the variation in eDNA concentrations between the handheld and laboratory-based qPCR platforms, both showed a positive relationship between eDNA concentrations and striped bass stocking densities. Their results validate the utility of on-site eDNA methods for identifying spatial and temporal distributions of the marine fish (Skinner et al., 2020).

| S TANDARDIZ ATION OF EDNA PR AC TI CE S
Current published eDNA studies are highly variable regarding field sampling and laboratory methods, assay development, data interpretation, and metadata reported. Research in this special issue highlight specific areas that lack standard methods and reporting across the field of eDNA (Klymus et al., 2019;Nicholson et al., 2020;Roux et al., 2020). Klymus et al. (2019) emphasized the need for improved accuracy in interpreting eDNA results and propose achieving this through standardized methods and reporting for the limit of detection (LOD) and limit of quantification (LOQ) in single-species qPCR eDNA assays. Well-defined qPCR assay performance metrics will also facilitate the assessment of eDNA data across studies and provide resource managers with a rigorous foundation for decisionmaking (Klymus et al., 2019). Another area where standardization is deficient is in metadata reporting. Nicholson et al. (2020) reviewed 160 eDNA freshwater studies and assessed whether or not they could be replicated using the reported data. The authors evaluated, and identified important data gaps, in 10 metadata categories: time, space, environmental conditions, sampling method, filtration method, sanitation, controls, extraction method, PCR conditions, and statistical analyses. According to their review, the best reported parameter (in 99.2% of the analyzed papers) was habitat type within the metadata category space. Other most commonly reported parameters are PCR data (i.e., PCR cycles, primer sequences), extraction equipment, negative controls, and sampling information such as sample volume, filter size, filter preservation method, year of sampling, and sampling location. Environmental conditions at the time of sampling (i.e., UV, precipitation, and wind) are the least reported parameters among studies (Nicholson et al., 2020). This review demonstrates the high variability in metadata reported among eDNA studies and the need for standardized data collection and reporting protocols.
In an effort to advance eDNA protocol standardization, Roux et al. (2020)analytically validated a qPCR assay using an approach consistent with that recommended by the World Organization for Animal Health (OIE) for assays used to detect globally listed animal diseases (International Office of Epizootics Aquatic Animal Health Standards Commission, 2016). The assay was thoroughly laboratory-validated for the detection of highly invasive European green crab from water samples and bulk DNA extracted from plankton samples and was field-tested in both Canada and Australia (Roux et al., 2020).
This study provides an example of how established guidelines from another field of study can be leveraged to promote the uptake of consistent and defensible approaches among eDNA practitioners.
Uniformity in the development and validation of eDNA qPCR assays will enable defensible comparisons among studies, improve replicability of methods among labs, and facilitate clear communication of assay characteristics to end-users.

| CHALLENG E S IN EDNA RE S E ARCH
Three main challenges in eDNA research are identified in this special issue: PCR inhibition, appropriate selection of primers for eDNA metabarcoding, and absence of standardization of eDNA methods. PCR inhibition occurs when environmental compounds (e.g., sediments, suspended particles, inorganic substances) or a high density of target DNA, interfere with PCR at the molecular level and prevent or significantly diminish amplification of the target region (Lance & Guan, 2019;Skinner et al., 2020). PCR inhibition was also the main drawback of the on-site qPCR detection approaches tested by Thomas et al. (2019) andSkinner et al. (2020).
Lack of amplification alone does not provide sufficient information to distinguish whether there is inhibition, absence of target DNA, or another process at work. In this special issue, PCR inhibition is demonstrated in both freshwater (Gasparini et al., 2020;Hobbs et al., 2020;Thomas et al., 2019) and marine studies (Roux et al., 2020;Skinner et al., 2020), with encouragement for the use of an internal positive control (IPC) to identify inhibition (see Gasparini et al., 2020;Klymus et al., 2019;Skinner et al., 2020;Thomas et al., 2019). Specially, inhibition can be identified when a delay or complete absence of IPC amplification occurs in a specific reaction well (Gasparini et al., 2020;Skinner et al., 2020;Thomas et al., 2019). Other suggested approaches to test for inhibition presented here are by spiking the eDNA samples with an artificial positive control (gBlock) (Roux et al., 2020) and by using an IntegritE-DNA TM test (Hobbs et al., 2020). The most common and overall successful approach to remove inhibitors is by diluting the problematic samples (Gasparini et al., 2020;Skinner et al., 2020;Thomas et al., 2019). Although there are ways of dealing with PCR inhibition, an important issue remains which is the variability in inhibition among sampling sites and seasons. This suggest heterogeneous patterns of inhibitors in the environment (Gasparini et al., 2020;Hobbs et al., 2020;Roux et al., 2020;Thomas et al., 2019) which should be further investigated to accurately interpret the results from eDNA surveys.
Another important challenge identified in this special issue is the appropriate selection of primers in eDNA metabarcoding studies to maximize taxa recovery (Morey et al., 2020;Mychek-Londer et al., 2019). Choosing a specific set of primers can affect taxa recovery in community studies and even universal primers can be subject to primer bias (i.e., preferential binding to specific taxa) as pointed out Lastly, the absence of methodological and reporting standards in eDNA research has been remarked in several articles presented in this special issue (Klymus et al., 2019;Nicholson et al., 2020;Roux et al., 2020). The acceptance of eDNA surveys by government and industry highly depends on the establishment of standardized protocols. The challenges identified in this special issue reflect issues frequently discussed among eDNA researchers and are important research areas. Methodological development and standardization in the field of eDNA should include field survey design and bioinformatic methods for processing high-throughput sequence data from eDNA metabarcoding studies. This discussion will continue as novel research emerges and further discussion forums and collaborations among academics, regulators, and industry are established.

| MOVING EDNA COLL ABOR ATIONS FORWARD
Biomonitoring surveys using eDNA are an innovative option that can improve taxon detectability, minimize cost, and increase the timeframes available for environmental surveys. eDNA can complement conventional survey methods thereby improving conservation outcomes. Adoption of eDNA surveys by regulated communities is currently held between research and regulatory acceptance. Notably, eDNA technologies are still viewed by regulators as "emerging and developing" (EPRI, 2018). The field of eDNA needs to advance to a consensus on methods standardization from researchers, and then, government agencies may be in a better position to confidently accept the results from eDNA surveys. Once industry has authorization from regulatory agencies to use this survey method, the widespread adoption, and consistent application of eDNA will result in increased benefits for biodiversity conservation.

ACK N OWLED G EM ENTS
We appreciate the work of all the contributing authors as well as the many reviewers of these manuscripts. Additional thanks to Jocelyn Kelly for her valuable comments on early versions of this manuscript.