Introduction
Effective monitoring of species distribution and abundance is the first step in the conservation of biodiversity and ecosystems (Margules & Pressey, 2000), as well as the proper management of fishery resources (Jackson et al., 2001). However, traditional methods that rely on capturing and morphological identification of species require substantial effort and cost, resulting in insufficient and biased monitoring and damage to individuals and their habitats (Thomsen & Willerslev, 2015). To overcome these limitations, analysis of environmental DNA (eDNA), which is defined as the total pool of DNA isolated from environmental samples (Pawlowski et al., 2020), has been developed (Ficetola et al., 2008; Minamoto et al., 2012; Deiner et al., 2017a). Macro-organisms such as fish are reported to produce eDNA from mucus, scale, feces, and gametes (Barnes & Turner, 2016). The PCR-based detection of eDNA in water samples enables non-invasive and cost-effective surveillance of species distribution and composition in aquatic ecosystems (Takahara et al., 2013; Yamamoto et al., 2017; Lawson Handley et al., 2019); thus, eDNA analysis is a promising tool for biological conservation and fishery resource management.
To achieve high accuracy and reliability of eDNA detection and quantification, eDNA must be preserved as soon as possible after water sampling because of its rapid degradation. There are a variety of preservation strategies for aqueous eDNA (Table 1), which primarily depend on whether water filtration is performed in the field (on-site) or in the laboratory (in-lab). On-site filtration allows the immediate storage of filter samples via the addition of a buffer (Renshaw et al., 2015; Spens et al., 2017) or desiccation (Thomas et al., 2019), whereas in-lab filtration is generally feasible to maximize the number of sampling sites per survey, as water filtration in the field is not required. In case of in-lab processing, to suppress eDNA degradation during transportation to the laboratory, water samples have been chilled and frozen (Takahara et al., 2015; Jo et al., 2020a), precipitated using organic solvents (Doi et al., 2017; Ladell et al., 2019), and directly added with Longmire’s buffer after collection (Williams et al., 2016).
Recently, benzalkonium chloride (BAC) has been used as an inexpensive and simple preservative for macrobial eDNA in water samples (Yamanaka et al., 2017). BAC is a cationic surfactant that inhibits bacterial function by adsorbing onto their cell surfaces (Ziani et al., 2011). The preservation strategy does not necessarily require elaborate work (e.g., the use of a pipette) and an equipment to chill the sample (e.g., cooler box and refrigerator). Yamanaka et al. (2017) reported that the addition of BAC at a final concentration of only 0.01% preserved 92% of bluegill sunfish (Lepomis macrochirus ) eDNA in water samples after 8 hours at ambient temperature (~ 25 °C) compared to only 14% in untreated water samples. Moreover, BAC addition allowed the retention of 50% of target eDNA in water samples after 10 days compared to non-detection in untreated water. To the best of our knowledge, BAC treatment is among the most suitable eDNA preservation strategies to maximize both the number of sampling sites and sampling volume (hundreds to thousands of milliliters). It allows intensive monitoring of species distribution and abundance via eDNA analysis over a short period of time.
Nevertheless, the performance of BAC in eDNA preservation has not necessarily been evaluated fully because most eDNA studies using BAC targeted short fragments (up to 200 bp) of mitochondrial DNA (mtDNA) in freshwater ecosystems (e.g., Sakata et al., 2017; Yamanaka et al., 2017; Hayami et al., 2020). Therefore, the present study investigated the performance of BAC in preserving eDNA in water samples from three aspects: (i) genetic region, (ii) DNA fragment size (i.e., the length of PCR amplicon), and (iii) marine ecosystems. First, given the possibility and prospect of using nuclear DNA (nuDNA) and longer DNA fragments in eDNA analyses for population-level inferences, such as population status and genetic diversity (Deiner et al., 2017b; Sigsgaard et al., 2020), it is important to verify whether BAC can be effective in preserving them from degradation. Moreover, BAC has only been applied to brackish water in a single experiment by Takahara et al. (2020); however, no study has examined its performance in eDNA preservation targeting seawater samples. Some water chemistry parameters, such as pH, salinity, and ionic content, are generally higher in marine systems than in freshwater systems (Okabe & Shimazu, 2007; Collins et al., 2018), which may affect the performance of BAC in eDNA preservation. Using Japanese jack mackerel (Trachurus japonicus ), an economically important marine fish in East Asia, including Japan, we examined the preservative performance of BAC targeting different fragment sizes of nuDNA and mtDNA in seawater samples. Furthermore, Yamanaka et al. (2017) anticipated that BAC should enable the preservation of community information inferred by eDNA metabarcoding; however, this has not been verified yet. Thus, we performed eDNA metabarcoding using MiFish primers (Miya et al., 2015) and examined whether BAC could be effective in preserving genetic information of fish communities.