The use of environmental DNA (eDNA) methods for community analysis has recently been developed. High-throughput parallel DNA sequencing (HTS), called eDNA metabarcoding, has been increasingly used in eDNA studies to examine multiple species. However, eDNA metabarcoding methodology requires validation based on traditional methods in all natural ecosystems before a reliable method can be established. To date, relatively few studies have performed eDNA metabarcoding of fishes in aquatic environments where fish communities were intensively surveyed using multiple traditional methods. Here, we have compared fish communities’ data from eDNA metabarcoding with seven conventional multiple capture methods in 31 backwater lakes in Hokkaido, Japan. We found that capture and field surveys of fishes were often interrupted by macrophytes and muddy sediments in the 31 lakes. We sampled 1 L of the surface water and analyzed eDNA using HTS. We also surveyed the fish communities using seven different capture methods, including various types of nets and electrofishing. At some sites, we could not detect any eDNA, presumably because of the polymerase chain reaction (PCR) inhibition. We also detected the marine fish species as sewage-derived eDNA. Comparisons of eDNA metabarcoding and capture methods showed that the detected fish communities were similar between the two methods, with an overlap of 70%. Thus, our study suggests that to detect fish communities in backwater lakes, the performance of eDNA metabarcoding with the use of 1 L surface water sampling is similar to that of capturing methods. Therefore, eDNA metabarcoding can be used for fish community analysis but environmental factors that can cause PCR inhibition, should be considered in eDNA applications.
Citation: Fujii K, Doi H, Matsuoka S, Nagano M, Sato H, Yamanaka H (2019) Environmental DNA metabarcoding for fish community analysis in backwater lakes: A comparison of capture methods. PLoS ONE 14(1): e0210357. https://doi.org/10.1371/journal.pone.0210357
Editor: Arga Chandrashekar Anil, CSIR-National Institute of Oceanography, INDIA
Received: July 28, 2018; Accepted: December 20, 2018; Published: January 31, 2019
Copyright: © 2019 Fujii et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The HTS data are deposited in DRA: DRA006606 (https://trace.ddbj.nig.ac.jp/DRASearch/submission?acc=DRA006606). All other data are deposited in the tables (Tables 2, 3, and 4).
Funding: The funders, Environment Research and Technology Development Fund (4-1602 to HD) and JST Core Research for Evolutional Science and Technology (JPMJCR13A2 to HY), had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ecological community evaluation is a critical step because it provides the basic information needed for biological conservation, for example the composition of fish communities in freshwater systems . Previously, fish capture methods such as the use of nets and other types of fishing gear/equipment have been used for community evaluation. However, each capture method has been shown to incompletely detect fish species in a community because of differences in the traits and habitats of fish. Thus, evaluation of fish communities should be completed using several capture methods . Some capture methods are difficult to employ in some ecosystems. For example, examining fish species in backwater environments is difficult because of limited access to pelagic areas, which is further complicated by the presence of macrophytes and muddy sediments. Using environmental DNA (eDNA) methods, especially DNA metabarcoding, may be a valuable new survey method for backwater habitats.
eDNA obtained directly from environmental samples can be used to evaluate species distributions. These methods have recently been developed and are considered to be useful techniques [3–8]. For example, in the past decade, many studies detected fish species [9, 10] and aquatic organisms [11–17] using eDNA. Recently, high-throughput parallel DNA sequencing (HTS) has been applied in eDNA studies to examine community composition from eDNA samples [3, 5, 18–24]. This eDNA technique with HTS sequencing and DNA-based species identification is called eDNA metabarcoding and is considered to be a useful method for assessing aquatic communities [19, 20].
eDNA metabarcoding has recently been applied in fish community surveys. For example, a universal polymerase chain reaction (PCR) primer for fish species, called MiFish (MiFish-U/E) was developed, whereby a hypervariable region of the mitochondrial 12S rRNA gene can be amplified . The versatility of these PCR primers using eDNA from four aquaria was tested with known species composition and natural seawater . These authors successfully detected eDNA from 232 fish species across 70 families and 152 genera in the aquaria and in the field, with a higher detection rate for species (>93%) in the aquaria. Moreover, using the MiFish primers and HTS, an investigation of marine fish communities in Maizuru Bay, Japan, detected a total of 128 fish species in the water samples [26, 27]. These studies indicate the great potential of eDNA metabarcoding as a useful tool for biodiversity assessment.
eDNA metabarcoding has been applied in fish biodiversity surveys, but testing and comparing its usefulness with traditional methods is necessary for the development of this technique as a conservation tool . The performance of eDNA metabarcoding has been tested in some studies and compared with that of capture methods [29, 30] or underwater visual consensus [26, 27], and it was found to have similar or higher performance than that of traditional methods. Comparisons of species detected using eDNA with those detected using multiple capture methods, which are generally used to investigate fish communities in aquatic habitats, are limited except for a study in a marine bay [26, 27]. Moreover, eDNA metabarcoding studies have primarily been conducted in marine , lake [31, 32], pond , and river ecosystems [34–37] but not in backwater ecosystems where there are many rare and endangered fish species . Therefore, a comparison of the performance of eDNA metabarcoding in assessing fish communities with that using traditional methods is necessary.
The objective of this study was to evaluate the performance of eDNA metabarcoding using HTS for fish communities in backwater lakes that are inhabited by rare and endangered fish species, and compare the ability of this methodology to detect species in a community with that of multiple capture methods, such as net sampling and electro-fishing. We conducted field surveys and sampling of eDNA in backwater lakes, including oxbow lakes that are isolated from rivers and backwater lakes, and thus, are postulated as the backwater of the rivers occurring from natural embankments . We also tested PCR inhibition effect on the metabarcoding analysis in the backwater lakes. Estimations of species distributions in these lakes using eDNA have not been widely attempted, and a direct capture of fish is difficult because of high densities of macrophytes on the surface.
Materials and methods
This project, including field survey and sampling for fish, was conducted in accordance with the Act on Welfare and Management of Animals, and Bylaw on Welfare and Management of Animals in Hokkaido, and the Regulations for Animal Experiments of the University of Hyogo. Permission for fish samplings in the backwaters and rivers in this study was granted by the Hokkaido government, Japan. An ethics statement is not required for this project.
We conducted a field survey and sampling in 31 backwater lakes in Hokkaido, Japan (Fig 1A). The study lakes were classified as oxbow (OL-, Fig 1B) and backwater lakes (BL-, Fig 1C). Most of the study lakes were interconnected by canals and rivers. These lakes host many rare and endangered fish species such as Lefua nikkonis and Rhynchocypris percnurus sachalinensis .
Traditional capture methods to estimate fish community composition
Traditional surveys for fish in the 31 lakes in our study were conducted, with moderate water levels, by the Hokkaido Regional Development Bureau between August 2 and September 13, 2016. The captures were conducted at shore sites. For captures, multiple fishing equipment were used including minnow traps, long-line fishing, gill nets, cast nets, dip nets, drag nets, and electro-fishing (Table 1). For minnow traps and long-line fishing, we used Pacific saury as bait. These traditional methods have been conducted for fish community surveys by the government such as the Ministry of Land, Infrastructure and Transport, Japan, and researchers in the habitats. The multiple fishing equipment may catch most of the fish species occupying the habitats. All capture methods were used for all sites and capture efforts were equal across all study lakes. Fish captured using fishing equipment were pooled and their abundances were calculated. Species were identified using pictorial keys . Because the individuals were juvenile and ammocoetes, some species were difficult to identify to species level and were included as a genus.
Water sampling and filtering
Using the aforementioned methods, we collected 1 L of surface water in bleached bottles from the shores of each lake where fish had previously been caught. We conducted water sampling, with moderate water levels in the lakes, from October 31 to November 4, 2016. Water samples were vacuum-filtered on-site onto 47 mm GF/F glass filters (pore size: 0.7 μm) (GE Healthcare, Little Chalfont, UK). During transport, the filters were stored in a cooler with ice packs. All filters were stored at -20°C within 12 h after filtration. A liter of Milli-Q water was used as the equipment control to monitor contamination during filtering in each site and during subsequent DNA extraction. The sampling bottles and filtering equipment (i.e., filter funnels and measuring cups) were cleaned using 10% commercial bleach (approximately 0.6% hypochlorous acid) and washed using DNA-free distilled water.
eDNA extraction from filter samples
To extract DNA from the filters, we followed previously described methods  using DNeasy Blood and Tissue Kit for DNA purification (Qiagen, Hilden, Germany) and Salivatte (Sarstedt, Numbrecht, Germany). The filter was placed in a salivated tube, and Buffer AL (400 μL) and proteinase K (20 μL) were added to the solution. The column was placed in a 56°C dry-oven for 30 min. After incubation, the salivated tube was centrifuged at 6,000 × g for 1 min to collect the DNA. To increase DNA yield from the filter, 220 μL of TE buffer was added to the filter and again centrifuged at 6,000 × g for 1 min. The collected DNA was purified using the DNeasy Blood and Tissue Kit following the manufacturer’s protocol. The extracted DNA samples (100 μL) were stored at –20°C until PCR assay.
Library preparation and MiSeq sequencing
A two-step PCR-procedure was used for library preparation of Illumina MiSeq sequencing. As the first step, a fragment of the mitochondrial 12S rRNA gene was amplified using the MiFish-U-F and MiFish-U-R primers  which were designed to contain Illumina sequencing primer regions and 6-mer Ns (forward: 5′-ACACTCTTTCCCTACACGACGCTCTTCCGATCTNNNNNN GTCGGTAAAACTCGTGCCAGC-3′, reverse: 5′-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTNNNNNN CATAGTGGGGTATCTAATCCCAGTTTG-3′). The italicized and normal letters represent MiSeq sequencing primers and MiFish primers, respectively, and the six random bases (N) were used to enhance cluster separation on the flow cells during initial base call calibrations on MiSeq. We used a KOD FX Neo polymerase (Toyobo, Osaka, Japan) for the first PCR to facilitate amplifications of DNA from crude extracts. The first PCR was performed with a 12 μL reaction volume containing 1× PCR Buffer for KOD FX Neo, 0.4 mM dNTP mix, 0.24 U KOD FX Neo polymerase, 0.3 μM of each primer, and 2 μL template . The thermal cycles of this step were as follows: initial denaturation at 94°C for 2 min, followed by 35 cycles of denaturation at 98°C for 10 s, annealing at 65°C for 30 s, and elongation at 68°C for 30 s, followed by final elongation at 68°C for 5 min. The first PCRs were performed using eight replicates to mitigate false negatives (PCR dropouts). Thereafter, individual 1st PCR replicates were pooled. The 30 μL of each PCR product was purified using AMPure XP (Beckman Coulter, Brea CA, USA) and eluted with 30 μL of sterilized water. The purified first PCR products were used as templates for the second PCR. The Illumina sequencing adaptors and the 8 bp identifier indices were added to the subsequent PCR process using a forward and reverse fusion primer: 5´-AATGATACGGCGACCACCGAGATCTACAXXXXXXXXACACTCTTTCCCTACACGACGCTCTTCCGATCT-3´ (forward) and 5´-CAAGCAGAAGACGGCATACGAGATXXXXXXXXGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT-3´ (reverse). The italic and normal letters represent MiSeq P5/P7 adapter and sequencing primers, respectively. The 8X bases represent dual-index sequences inserted to identify different samples. The second PCR was conducted with 12 cycles of a 12 μL reaction volume containing 1× KAPA HiFi HotStart ReadyMix, 0.3 μM of each primer, and 1.0 μL template from the first PCR production. The thermal cycle profile after an initial 3 min denaturation at 95°C was as follows: denaturation at 98°C for 20 s, annealing, and extension combined at 72°C (shuttle PCR) for 15 s, with the final extension at the same temperature for 5 min. The second PCR products were pooled in equal volumes and purified using AMPure XP as the first PCR. The purified PCR products were loaded on a 2% E-Gel SizeSelect (Thermo Fisher Scientific, Waltham, MA, USA) and the target size of the libraries (approximately 370 bp) was collected. The library concentration and quality were estimated by a Qubit dsDNA HS assay kit and a Qubit 2.0 (Thermo Fisher Scientific). The amplicon libraries were sequenced by 2 × 250 bp paired-end sequencing on the MiSeq platform at Ryukoku University using the MiSeq v2 Reagent Kit according to the manufacturer’s instructions. Note that the sequencing run contained a total of 351 libraries including 34 of our libraries and 317 libraries from other research projects.
Bioinformatic analysis for MiSeq sequencing
The processing formality of the MiSeq reads was evaluated by the FASTQC program (http://www.bioinformatics.babraham.ac.uk/projects/fastqc). After confirming a lack of technical errors in the MiSeq sequencing, low-quality tails were trimmed from each read using “DynamicTrim.pl” in the SOLEXAQA software package  with a cut-off threshold set at a Phred score of 10. The trimmed paired-end reads (reads 1 and 2) were then merged with each other. The assembled reads were further filtered by custom Perl scripts to remove reads with either ambiguous sites (Ns) or those exhibiting unusual lengths with reference to the expected size of the PCR amplicons (297 ± 25 bp). The software TagCleaner  was used to remove primer sequences with a maximum of three-base mismatches and transform the FASTQ format into FASTA.
The pre-processed reads with an identical sequence (i.e., 100% sequence similarity) were assembled using UCLUST . The number of identical reads was added to the header line of the FASTA formatted data file. The sequences represented by more than or equal to 10 identical reads were subjected to the downstream analyses.
The processed reads were subjected to local BLASTN searches on the comprehensive reference database of fish species that were previously established . The top BLAST hit with a sequence identity of ≥97% and the E-value threshold of 10−5 was applied to species detection of each sequence, but the species were mostly identified with ≥99% match (Fig 2). From the BLAST results, we identified the species using methods previously described . We also calculated the rate of shared sites by the number of sites with species detected by metabarcoding per number of sites with species detected by metabarcoding and traditional methods.
PCR inhibition test
To test for inhibition in the eDNA samples, 1 μL of the plasmid including the cytochrome b gene from Trachurus japonicus (1.5 × 102 copies), a marine fish, which was not found at the sites and detected by the eDNA metabarcoding, was added to the PCR template. This contained 900 nM of each primer and 125 nM of TaqMan probe in a 1× PCR master mix (KOD FX Neo) and 2 μL of the DNA solution. The total volume of each reaction mixture was 10 μL. The real-time PCR was performed by using quantitative real-time PCR (PikoReal real-time PCR, Thermo Fisher Scientific). The PCR (three replicates) was performed as follows: 2 min at 50°C, 10 min at 95°C, and 55 cycles of 15 s at 95°C, and 60 s at 60°C. The non-template control (NTC) was performed in three replicates per PCR. The results of the PCR were analyzed using PikoReal software v. 188.8.131.521 (Thermo Fisher Scientific). The primer and probe set used was that previously reported : forward primer: 5′-CAGATATCGCAACCGCCTTT-3′; reverse primer: 5′-CCGATGTGAAGGTAAATGCAA A-3′; probe: 5′-FAM-TATGCACGCCAACGGCGC CT-TAMRA-3′. The presence of PCR inhibitors was evaluated ΔCt (Ctpositive control−Ctsample). ΔCt of ≥3 cycles and is usually considered to be evidence of inhibition .
To evaluate the overlapping of the species lists between the captured fish and eDNA metabarcoding, we calculated the detection rate for eDNA metabarcoding. The equation was (the number of species found by both the capture and eDNA methods/number of all captured species) × 100. We conducted the community analysis for both the eDNA metabarcoding and catching data. We used the presence/absence data of both datasets and calculated the Jaccard similarity index and ordinated using non-metric multidimensional scaling (NMDS) with 100 iterations. We also performed an analysis of similarities (ANOSIM) for the Jaccard similarity index to test whether there was a statistically significant difference between two or more groups of sampling units. We tested the differences in communities between the eDNA metabarcoding and catching data (permutation = 999, α = 0.05). All statistical analyses were conducted in the “vegan” package of R v. 3.3.2 (R Core Team 2016) .
Traditional catching methods to estimate fish community
Using multiple capture methods, we found 26 taxa in nine families (Fig 2A). Cyprinidae was the most abundant family (12 species).
A MiSeq paired-end sequencing for the 34 PCR libraries (including 31 samples and three negative controls) yielded a total of 641.056 reads. We detected 24 taxa in eight families after following the pipeline procedure for eDNA metabarcoding, representing 85% of taxa that we found by the traditional methods (Fig 2B). We identified Carassius buergeri and C. auratus langsdorfii as Carassius spp. because of higher DNA-sequence similarities (<2 bp differences). We also detected two invaive species in this region, namely, Opsariichthys uncirostris and Zacco platypus. Using eDNA metabarcoding, we could not find three fish species that were detected by capture methods (i.e., Lethenteron sp., Ctenopharyngodon idella, and Hypomesus nipponensis at any site (Table 2).
In OL-14, OL-15, BL-1, and BL-5, we could not detect any fish species, probably because the PCR amplification almost failed due to PCR inhibitors. In fact, PCR inhibitor tests using real-time PCR showed remarkable inhibition (∆Ct > 3) in the samples compared with other sites and that showed modulate inhibition (∆Ct = 1.87) at BL-1 (Table 3). We also detected marine species using eDNA metabarcoding (Table 4). The number of reads for each of these species was one or two orders of magnitude less than that for expected species.
Comparisons of traditional capture and eDNA metabarcoding data
To compare between traditional capture methods and eDNA metabarcoding data, the taxonomic levels in the species list from traditional captures were adjusted to the lists from eDNA metabarcoding (Fig 2). Rates of shared sites were higher than 50% in 12 out of the 25 taxa (Fig 3 and Table 5). Using detection rates from eDNA metabarcoding, we compared detection among the sites (Fig 4). The detection rates were >50% in 22 out of 31 sites, including rates that were 100% at three sites (Fig 4). After the exclusion of the four sites where there were no eDNA detections (OL-14, OL-15, BL-1, and BL-5), the mean detection rate was 70.0%. However, the rate varied among the sites.
Comparing the ordination of fish communities between traditional capture and eDNA metabarcoding data
The NMDS ordinations for fish community similarities were similar in structure for both traditional capture and eDNA metabarcoding data (stress value = 0.14) (Fig 5). In fact, the communities were not significantly different between traditional capture and eDNA metabarcoding using ANOSIM (R = -0.0161, P = 0.585). The BL sites were clearly different from the OL sites in both survey methods.
We found that eDNA metabarcoding using only 1 L of water sample offered high detection rate (85%) of fish taxa when multiple capture methods and eDNA metabarcoding to evaluate fish communities in backwaters were compared. Multiple fish captures using seven types of equipment required three people to be present at each site each day, whereas eDNA sampling required only one person at each site, each day and only required a few minutes to complete. Here, we showed that only 1 L of water would be sufficient to evaluate fish community composition, and we also illustrated the similarity of fish communities across our study sites using NMDS and ANOSIM results. Thus, eDNA metabarcoding can reduce the cost and effort of surveying fish communities in backwaters and other freshwater habitats.
Comparison between the traditional capture and eDNA metabarcoding data
Some taxa that were captured using traditional methods were not detected by eDNA metabarcoding. Between 1 and 11 individuals caught using the seven capture methods were taxa that were not detected using eDNA. These were: Lethenteron sp., Ctenopharyngodon idella, and Hypomesus nipponensis. The low abundance of the taxa may be the result of a limited amount of eDNA in the water samples. Moreover, because Lethenteron sp. inhabits the sediment or sand bed of the lakes, an adequate amount of eDNA may not be available around the surface water , resulting in a lack of detection by eDNA metabarcoding. Such limitations of eDNA were also reported in previous real-time PCR based studies . In this study, we only collected 1 L-samples from the lakes without biological replications. Increasing replications induces the detection of fish species by eDNA metabarcoding , thus, the limited sampling may have also caused non-detection by eDNA metabarcoding.
Tribolodon brandtii was caught at only two sites, but eDNA metabarcoding detected the taxa at 14 sites, probably indicating that most of the individuals identified as Tribolodon sp. in the field actually represent T. brandtii. In fact, individuals of Tribolodon genus with <10 cm body length were not identified to the taxa level in the field because of difficulties in morphological identification. For example, to identify the genus Tribolodon in the field, the number of predorsal scale rows must be counted and therefore, eDNA metabarcoding is advantageous where such genera are difficult to identify. An additional advantage of eDNA metabarcoding occurs in the case of some taxa for which only juveniles and ammocoetes are available and cannot be identified to species by traditional methods. The eDNA metabarcoding can identify the species regardless of the maturity stage.
We were not able to detect eDNA from four sites using metabarcoding, namely, OL-14, OL-15, BL-1, and BL-5. The results of PCR inhibitor tests showed remarkable inhibitions at these four sites (Table 3). High humic acid in the surface water of the backwater lakes in the region has been described . Humic acid is a major PCR inhibitor that would reduce the number of present amplicons [49–52]. Therefore, the PCR inhibitor probably decreased the detection rate at these sites, despite the fact that we used the PCR master mix (KOD FX Neo) to amplify the DNA from samples with high PCR inhibitors, such as soil samples. Thus, we suggest the need for caution when applying eDNA metabarcoding in backwater habitats where many PCR inhibitors are present, such as in streams . In this study, we extracted DNA from water samples using DNeasy blood and tissue kits. The use of another DNA extraction kit, such as the PowerSoil kit (Qiagen), might reduce the amount of humic acid extracted from samples, mitigating the negative effects of PCR inhibitors, but further study is still needed.
We also found similarities between NMDS ordinations using traditional capture methods and eDNA metabarcoding, indicating that fish community structures across sites were not significantly different between the two methods. In fact, the taxa list showed a 70% overlap. Therefore, the community analysis using eDNA can be useful for spatially broad areas with remarkably little effort.
eDNA detections of fish species
We detected marine species using eDNA metabarcoding. In addition to marine species, Oncorhynchus keta, Oncorhynchus masou masou, Spirinchus lanceolatus, and Plecoglossus altivelis, that inhabit freshwater habitats, are often used in regional cuisine and therefore, sewage water may contain eDNA from these species. Thus, sewage water may be a source of the eDNA of unusual species in the freshwater of backwater lakes. Furthermore, marine species were exclusively detected in the OL-1 site because this site was an exception, with sewage water flowing directly into the lake. Thus, it is possible that samples collected in the lake water were contaminated by the marine fish-derived DNA from the sewage. It has been suggested that the eDNA of commonly consumed species may originate from wastewater contamination . It is, therefore, necessary to consider the source of DNA release to address the false-positive detections in eDNA metabarcoding, especially if it is from sewage water.
Through eDNA metabarcoding, we also detected two domestic invasive species in OL-12 that had never been found in the region: Opsariichthys uncirostris (1817 reads) and Zacco platypus (678 reads). The species Z. platypus was found in another region of Hokkaido Island, whereas O. uncirostris had never been found on Hokkaido Island. In this study, we caught seven domestic invasive species that had previously been found in this region, including Cyprinus carpio, Pseudorasbora parva, and Gnathopogon elongatus. Aside from these species, O. uncirostris and Z. platypus are possibly new invasive species to the region. Although further evaluation of capture is needed, eDNA metabarcoding can identify potentially invasive species in aquatic habitats.
In conclusion, eDNA metabarcoding of fish communities was performed similarly through multiple capture methods in backwater lakes. Traditional fish-capture methods, using nets or electro-fishing, require more effort and are more time-consuming in the field. Here, we showed that eDNA in 1 L water samples had a similar detectability to that of traditional methods of fish capture, suggesting the usefulness of eDNA in detecting fish community in the habitats. The eDNA metabarcoding can provide us with various information about fish communities, including species compositions of fish, emergence of new invasive species, and survival of locally extinct species. We also noted some disadvantages of the eDNA metabarcoding such as PCR inhibition for eDNA analysis and false positives of some marine species originating from wastewater contamination. Such disadvantages of eDNA methods should be considered in future applications.
We would like to thank Hokkaido Regional Development Bureau for providing the fish community data from multiple catching methods.
- 1. Elliott M, Whitfield AK, Potter IC, Blaber SJ, Cyrus DP, Nordlie FG, et al. The guild approach to categorizing estuarine fish assemblages: a global review. Fish Fish. 2007;8: 241–268.
- 2. Gabriel O, Lange K, Dahm E, Wendt T. Fish catching methods of the world. John Wiley & Sons; 2008.
- 3. Deiner K, Bik HM, Mächler E, Seymour M, Lacoursière‐Roussel A, Altermatt F, et al. Environmental DNA metabarcoding: Transforming how we survey animal and plant communities. Mol. Ecol. 2017;26: 5872–5895. pmid:28921802
- 4. Rees HC, Maddison BC, Middleditch DJ, Patmore JR, Gough KC. Review: the detection of aquatic animal species using environmental DNA—a review of eDNA as a survey tool in ecology. J Appl Ecol. 2014;51: 1450–1459.
- 5. Goldberg CS, Strickler KM, Pilliod DS. Moving environmental DNA methods from concept to practice for monitoring aquatic macroorganisms. Biol. Conserv. 2015;183: 1–3.
- 6. Thomsen PF, Willerslev E. Environmental DNA–an emerging tool in conservation for monitoring past and present biodiversity. Biol. Conserv. 2015;183: 4–18.
- 7. Andruszkiewicz EA, Starks HA, Chavez FP, Sassoubre LM, Block BA, Boehm AB. 2017. Biomonitoring of marine vertebrates in Monterey Bay using eDNA metabarcoding. PloS ONE. 2017;12: e0176343. pmid:28441466
- 8. Bista I, Carvalho GR, Walsh K, Seymour M, Hajibabaei M, Lallias D, et al. Annual time-series analysis of aqueous eDNA reveals ecologically relevant dynamics of lake ecosystem biodiversity. Nat Commun. 2017;8: 14087. pmid:28098255
- 9. Minamoto T, Yamanaka H, Takahara T, Honjo MN, Kawabata Z. Surveillance of fish species composition using environmental DNA. Limnology 2012;13: 193–197.
- 10. Mahon AR, Jerde CL, Galaska M, Bergner JL, Chadderton WL, Lodge DM, et al. Validation of eDNA surveillance sensitivity for detection of Asian carps in controlled and eld experiments. PLoS ONE 2013;8: e58316. pmid:23472178
- 11. Ikeda K, Doi H, Tanaka K, Kawai T, Negishi JN. Using environmental DNA to detect an endangered crayfish Cambaroides japonicus in streams. Conserv Gen Res. 2016;8: 231–234.
- 12. Pilliod DS, Goldberg CS, Arkle RS, Waits LP. Estimating occupancy and abundance of stream amphibians using environmental DNA from filtered water samples. Can J Fish Aquat Sci. 2013;70: 1123–1130.
- 13. Fukumoto S, Ushimaru A, Minamoto T. A basin-scale application of environmental DNA assessment for rare endemic species and closely related exotic species in rivers: a case study of giant salamanders in Japan. J Appl Ecol. 2015;52: 358–365.
- 14. Katano I, Harada K, Doi H, Souma R, Minamoto T. Environmental DNA method for estimating salamander distribution in headwater streams, and a comparison of water sampling methods. PLoS ONE. 2017. pmid:28520733
- 15. Tréguier A, Paillisson JM, Dejean T, Valentini A, Schlaepfer MA, Roussel JM. Environmental DNA surveillance for invertebrate species: advantages and technical limitations to detect invasive crayfish Procambarus clarkii in freshwater ponds. J Appl Ecol. 2014;51: 871–879.
- 16. Schmidt BR, Kery M, Ursenbacher S, Hyman OJ, Collins JP. Site occupancy models in the analysis of environmental DNA presence/absence surveys: a case study of an emerging amphibian pathogen. Methods Ecol Evol. 2013;4: 646–653.
- 17. Brozio S, Manson C, Gourevitch E, Burns TJ, Greener MS, Downie JR, Hoskisson PA. Development and application of an eDNA method to detect the critically endangered trinidad golden tree frog (Phytotriades auratus) in bromeliad. PLoS ONE 2017;12: e0170619. pmid:28199338
- 18. Ficetola GF, Miaud C, Pompanon F, Taberlet P. Species detection using environmental DNA from water samples. Biol Lett. 2008;4: 423–425. pmid:18400683
- 19. Jerde CL, Chadderton WL, Mahon AR, Renshaw MA, Corush J, Budny ML, et al. Detection of Asian carp DNA as part of a Great Lakes basin-wide surveillance program. Can J Fish Aquat Sci. 2013;70: 522–526.
- 20. Takahara T, Minamoto T, Doi H. Using environmental DNA to estimate the distribution of an invasive fish species in ponds. PLoS ONE 2013;8, e56584. pmid:23437177
- 21. Doi H, Takahara T, Minamoto T, Matsuhashi S, Uchii K, Yamanaka H. Droplet digital polymerase chain reaction (PCR) outperforms real-time PCR in the detection of environmental DNA from an invasive fish species. Env Sci Tech. 2015;49: 5601–5608. pmid:25850372
- 22. Uchii K, Doi H, Minamoto T. A novel environmental DNA approach to quantify the cryptic invasion of non-native genotypes. Mol Ecol Res. 2016;16: 415–422. pmid:26307935
- 23. Takahara T, Minamoto T, Yamanaka H, Doi H, Kawabata Z. Estimation of fish biomass using environmental DNA. PLoS ONE. 2012. pmid:22563411
- 24. Thomsen PF, Kielgast J, Iversen LL, Wiuf C, Rasmussen M, Gilbert MTP, et al. Monitoring endangered freshwater biodiversity using environmental DNA. Mol Ecol. 2012;21: 2565–2573. pmid:22151771
- 25. Miya M, Sato Y, Fukunaga T, Sado T, Poulsen JY, Sato K, et al. MiFish, a set of universal PCR primers for metabarcoding environmental DNA from fishes: detection of more than 230 subtropical marine species. Royal Soc Open Sci. 2015;2: 150088. pmid:26587265
- 26. Yamamoto S, Minami K, Fukaya K, Takahashi K, Sawada H, Murakami H, et al. Environmental DNA as a ‘Snapshot’ of fish distribution: a case study of Japanese Jack Mackerel in Maizuru Bay, Sea of Japan. PLoS ONE 2016;11: e0149786. pmid:26933889
- 27. Yamamoto S, Masuda R, Sato Y, Sado T, Araki H, Kondoh M, et al. Environmental DNA metabarcoding reveals local fish communities in a species-rich coastal sea. Sci Rep. 2017;7. pmid:28079122
- 28. Thomsen PF, Kielgast J, Iversen LL, Møller PR, Rasmussen M, Willerslev E. Detection of a diverse marine fish fauna using environmental DNA from seawater samples. PLoS ONE 2012;7: e41732. pmid:22952584
- 29. Hänfling B, Handley LL, Read DS, Hahn C, Li J, Nichols P, et al. Environmental DNA metabarcoding of lake fish communities reflects long-term data from established survey methods. Mol Ecol. 2016;25: 3101–3119. pmid:27095076
- 30. Shaw JLA, Clarke LJ, Wedderburn SD, Barnes TC, Weyrich LS, Cooper A. Comparison of environmental DNA metabarcoding and conventional fish survey methods in a river system. Biol Conserv. 2016;197: 131–138.
- 31. Eichmiller JJ, Bajer PG, Sorensen PW. The relationship between the distribution of common carp and their environmental DNA in a small lake. PLoS ONE 2014;9: e112611. pmid:25383965
- 32. Darling JA1, Mahon AR. From molecules to management: adopting DNA-based methods for monitoring biological invasions in aquatic environments. Environ Res. 2011. pmid:21353670
- 33. Doi H, Uchii K, Takahara T, Matsuhashi S, Yamanaka H, Minamoto T. Use of droplet digital PCR for estimation of fish abundance and biomass in environmental DNA surveys. PLoS ONE 2015;10: e0122763. pmid:25799582
- 34. Yamanaka H, Minamoto T. The use of environmental DNA of fishes as an efficient method of determining habitat connectivity. Ecol Ind. 2016;62: 147–153.
- 35. Wilcox TM, McKelvey KS, Young MK, Sepulveda AJ, Shepard BB, Jane SF, et al. Understanding environmental DNA detection probabilities: A case study using a stream-dwelling char Salvelinus fontinalis. Biol Conserv. 2016;194: 209–216.
- 36. Doi H, Inui R, Akamatsu Y, Kanno K, Yamanaka H, Takahara T, et al. Environmental DNA analysis for estimating the abundance and biomass of stream fish. Freshw Biol. 2017;6: 30–39.
- 37. Deiner K, Fronhofer EA, Mächler E, Walser JC, Altermatt F. Environmental DNA reveals that rivers are conveyer belts of biodiversity information. Nat Commun. 2016;7: 12544. pmid:27572523
- 38. Hayashida K, Hirayama A, Ueda H. Changes in fish fauna in oxbow lakes on the ishikari river and the influence of invasive fish species. Annu J Hydraul Eng. 2010;54. (in Japanese).
- 39. Takeuchi K, Maruoka N, Daimon S. A study on nature restoration in the lower reaches of the Ishikari River. Report of Riverfront Japan Research Institute. 2003;14.(in Japanese).
- 40. Nakabo T. Fishes of Japan with pictorial keys to the species, 3rd ed. Tokai University Press; 2013 (in Japanese).
- 41. Sato H, Sogo Y, Doi H, Yamanaka H. Usefulness and limitations of sample pooling for environmental DNA metabarcoding of freshwater fish communities. Sci Rep. 2017;7. pmid:29093520
- 42. Cox MP, Peterson DA, Biggs PJ. SolexaQA: at-a-glance quality assessment of Illumina second-generation sequencing data. BMC Bioinform. 2010;11: 485. pmid:20875133
- 43. Schmieder R, Lim YW, Rohwer F, Edwards R. TagCleaner: Identification and removal of tag sequences from genomic and metagenomic datasets. BMC Bioinform. 2010;11: 341. pmid:20573248
- 44. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 2010;26: 2460–2461. pmid:20709691
- 45. Hartman LJ, Coyne SR, Norwood DA. Development of a novel internal positive control for Taqman based assays. Mol Cell Probes. 2005;19: 51–59. pmid:15652220
- 46. R Core Team. 2017 R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
- 47. Dougherty MM, Larson ER, Renshaw MA, Gantz C A, Egan SP, Erickson DM, Lodge DM. Environmental DNA (eDNA) detects the invasive rusty crayfish Orconectes rusticus at low abundances. J Appl Ecol. 2016;53: 722–732. pmid:27773942
- 48. Tani M, Kondo R, Tsutsuji K. Water quality in marshes and streams in the coastal peatland at the Pacific seaboard of the Tokachi district. J Jap Soc Water Env. 1998;22: 232–237.
- 49. Lakay FM, Botha A, Prior BA. Comparative analysis of environmental DNA extraction and purification methods from different humic acid-rich soils. J Appl Microb. 2007;102: 265–273. pmid:17184343
- 50. Opel KL, Chung D, McCord BR. A study of PCR inhibition mechanisms using real time PCR. J Forensic Sci. 2010;55: 25–33. pmid:20015162
- 51. Abbaszadegan M, Huber MS, Gerba CP, Pepper IL. Detection of enteroviruses in groundwater with the polymerase chain reaction. Appl Environ Microbiol. 1993;59: 1318–1324. pmid:7685998
- 52. Ijzerman MM, Dahling DR, Fout GS. A method to remove environmental inhibitors prior to the detection of waterborne enteric viruses by reverse transcription-polymerase chain reaction. J Virol Method. 1997;63: 145–153.
- 53. Jane SF, Wilcox TM, McKelvey KS, Young MK, Schwartz MK, Lowe WH, Letcher BH, Whiteley AR. Distance, flow and PCR inhibition: eDNA dynamics in two headwater streams. Mol Ecol Res. 2015;15: 216–227. pmid:24890199
- 54. Stoeckle MY, Soboleva L, Charlop-Powers Z. Aquatic environmental DNA detects seasonal fish abundance and habitat preference in an urban estuary. PLOS ONE 2017; 12: e0175186. pmid:28403183