Ethics statement
Field research, including sediment collection, was approved by the Harbormaster of Maizuru Bay (Permission Number 31 issued on July 1, 2016). All the experiments were performed in accordance with the guidelines on the Regulation on Animal Experimentation of Kyoto University (https://www.kyoto-u.ac.jp/en/research/research-compliance-ethics/animal-experiments), the Kyoto Prefecture Fishery Management Rules (http://www.pref.kyoto.jp/reiki/reiki_honbun/a300RG00000634.html), and the ARRIVE guidelines (https://arriveguidelines.org). Fish (15 individuals of jack mackerel juveniles) were anesthetized prior to length and weight measurements using 0.05% of 2-phenoxyethanol, and all individuals recovered from the anesthesia. No fish were sacrificed or injured in the present study. The research plan was approved by the institutional review boards at the Maizuru Fisheries Research Station (MFRS) of Kyoto University.
Detection of eDNA in the water and sediment in experimental tanks
On July 6, 2016, natural marine sediment was collected at a depth of 47 m off the shore of Kyoto in the Sea of Japan (35.5544° N, 135.3210° E), using a Smith–McIntyre bottom sampler. Approximately 100 L of sediment was collected from 13 casts and was preserved in four large, covered containers at room temperature until the experiment began. Four sub-samples, 3 g from each container, were used for eDNA extraction and detection of jack mackerel using the method described below, and we confirmed that none of them contained the DNA of this species. The median particle diameter of the sediment was 47.7 μm, and mud content was 61.9% based on analyses using a laser diffraction particle size analyzer (SALD-2200, Shimadzu, Kyoto, Japan). Jack mackerel juveniles were collected by hook-and-line fishing from the pier of the MFRS. This species is the most abundant fish in this area and is typically found in waters 14 °C or warmer43.
Four 200 L polycarbonate tanks (66 cm in bottom diameter) were set in the rearing facility of MFRS; three were used as test tanks, and the fourth was used as a control (blank) tank. The marine sediment (24 L, 7 cm in thick layer) was placed in each tank. Fine-filtered seawater was provided 2 d after the sediment had settled. The seawater used was pumped from 6 m depth offshore from the MFRS and filtered by passing through coarse polyvinyl fabric and fine sand of ca. 0.6 mm in diameter (5G-ST, Nikkiso Eiko, Japan; www.nikkiso-eiko.co.jp). Water was supplied at the rate of 490 mL min−1 (four cycles of circulation per day) and was drained from the center of each tank, filtered through a 2 mm mesh net. Aeration was performed at a rate of 600 mL min−1. This flow-through system was maintained throughout the experimental period.
Five individual jack mackerel (70.7 ± 4.4 mm in total length and 3.26 ± 0.67 g in wet mass, mean ± SD) were introduced in each of the three test tanks on August 8, 7 days after the sediment was introduced. Defrosted krill Euphausia pacifica (5 g per tank) was fed to the fish between 16:00 and 17:00 every day. Fish were removed from the tanks 14 days after introduction using two hand nets, taking care not to disturb the sediment. Water temperature was recorded using a digital thermometer at 10:00 every day while the fish were kept in the tanks, for the following 14 days after their removal, and once a week thereafter. The water temperature ranged from 24.9 to 29.5 °C (mean = 27.9 °C) during the first four weeks of the experiment and from 9.4 to 27.9 °C (mean = 17.8 °C) during the following months. These conditions are similar to the natural condition that would be undergone by eDNA in sediment; for the last 19 years, the recorded bottom water temperature in the area from which the jack mackerel had been collected ranged from 8.5 to 29.6 °C, with a mean of 18.2 °C (Masuda 200843 with updated data). All rearing equipment was either newly purchased or bleached with 0.1% sodium hypochlorite and rinsed well with tap water before use.
Water and sediment samples were collected immediately before the introduction of fish (day 0) and on days 1, 2, 4, 7, and 14 after their introduction. Sampling was also conducted on days 0, 1, 2, 4, 7, and 14, as well as in months 1, 2, 4, 8, and 12 after the removal of fish. Three 1 L water samples and three sediment samples (3 g) were collected from each tank on each sampling day. Water was collected from the drainage outlet in plastic bottles, and sediment was collected in petri dishes (inner diameter of 58 mm and depth of 21 mm). The sediment was collected by pushing the open end of a petri dish onto the surface of the sediment and securing the cover from underneath. Sediment sampling was conducted with a pair of prebleached long-sleeved gloves. One sample was obtained from the central tank area, and another two from near the peripheral tank area. Repetitive collection from the same location was avoided by marking each sampling location with a piece of PVC pipe (similar in diameter to the petri dishes and 3 cm in height).
Water was filtered using glass fiber filters (0.7 μm mesh; GF/F 47 mm, GE Healthcare Japan, Tokyo, Japan). This mesh size, along with 0.45 μm, are two of the most commonly used filters in macroorganism eDNA studies44. The amount of eDNA detected using a 0.7 μm mesh is equivalent to that by a 0.45 μm mesh30. Contamination was evaluated by filtering 1 L of reverse osmosis water at the end of each sampling day. The filtered paper was wrapped in aluminum foil and preserved at − 20 °C.
Sediment core sampling
Sediment core samples were collected at four locations (St. 1–4) in and around Nishi-Moune Bay, Kesennuma, Miyagi, Japan (38.8919–38.8932°N, 141.6235–141.6262°E; Fig. 1) on May 20, 2017. St. 1 was in the inner part of the bay where the tsunami impact was assumed to be the highest, with a run-up height of 15 m. St. 2 was located along a shallow rocky shore where the tsunami impact was limited. St. 3 was located at the mouth of the bay, and St. 4 was outside the bay. Average depths of the seafloor where cores were collected were 8.1, 9.6, 23.0, and 14.0 m at stations 1, 2, 3, and 4, respectively. Seafloor temperatures ranged from 9.9 to 11.5 °C. An acrylic pipe (inner diameter of 54 mm, length of 50 cm, and thickness of 3 mm) was pushed into the bottom sediment by a scuba diver. A silicon cap (59 and 52 mm in upper and lower diameter, respectively, and 45 mm in height) was placed on the top of the pipe, and the diver slowly pulled the pipe up and put another cap on the bottom. Three cores were collected from each location and transferred to a boat at the sea surface. Sediment core samples were kept vertical to avoid disturbing the layers and protected from direct sunlight. Cores were immediately transferred to the laboratory within 10 min, and were prepared for the cutting process.
The core samples (1 cm thickness) were cut by layers as follows: after removing the bottom cap, a core sample pipe was placed on a stage that pushed the sediment inside. Seawater in the upper part of the pipe was discarded until the top of the sediment appeared on the surface. A thin acrylic plate was used to cut the core, and the cut specimen was placed in a small vinyl bag and preserved at -20 °C. All 12 collected cores were used for eDNA analysis, and one at St. 1 (inner bay) and all three at St. 3 (bay mouth) were used for the analysis of PAHs.
DNA extraction
DNA extraction from the glass fiber filter was performed following the method described in Yamamoto et al.6 using a DNeasy Blood and Tissue Kit (Qiagen, Hilde, Germany) and a Salivette tube (Sarstedt, Nümbrecht, Germany). Total eDNA was eluted in 100 μL AE buffer and preserved at − 20 °C.
DNA extraction from sediment was conducted using a combination of alkaline DNA extraction45 and ethanol precipitation, using a commercial soil DNA extraction kit (Power Soil DNA Isolation Kit, QIAGEN, Hilden, Germany), as described in Sakata et al.26. Wet sediment (ca. 3 g) was placed in a 15 mL tube. Triplicate samples were obtained from each petri dish in the tank experiment, and a single sample was obtained from each layer in the sediment cores. We added 6 mL of 0.33 M NaOH and 3 mL of 10 mM TE buffer (pH = 6.7) to the tube and mixed well using Voltex. The samples were incubated at 94 °C for 50 min, and during this time, they were inverted at 15 and 30 min of incubation. After the incubation, the samples were cooled for several minutes and then centrifuged at 5,000 × g for 30 s. Supernatants (7.5 mL) were collected in 50 mL tubes and 7.5 mL of 1 M Tris HCL buffer (pH = 9.0 in the tank experiment and pH = 6.7 in the core samples), 1.5 mL of 3 M sodium acetate (pH = 5.2), and 30 mL of 99.5% ethanol were added, and mixed well by inversion. Ethanol precipitation was achieved by incubating the mixture for 1 h at − 20 °C. As a negative control of extraction, 3 mL of pure water was treated in the same manner. Sediment in the tank experiment was processed up to the ethanol precipitation on the same day as sampling, whereas core samples were defrosted at room temperature prior to analysis and then preserved as precipitate.
The ethanol-precipitated sediment sample was centrifuged at 5,350 × g for 20 min, after which the supernatant was discarded. The precipitate was moved to the PowerBead Tube of the Power Soil Isolation Kit using a microspatula. The debris left in the centrifuged tube was also transferred by dissolving it in 100 μL of pure water. The following procedure was performed according to the protocol of Power Soil. The total eluted DNA (100 μL) was stored at − 20 °C. All the spatulas were bleached prior to use, and brand-new centrifugation tubes were used for the procedure.
Quantitative PCR
DNA was quantified using real-time TaqMan PCR with a LightCycler 96 Real-Time PCR System (Roche, Basel, Switzerland). Species-specific sets of primers and probes were used to quantify the eDNA of jack mackerel, moon jellyfish, and sea nettle (Supplementary Table S5). For the specimens in the tank experiment, each reaction contained 2 μL of extracted eDNA solution, a final concentration of 900 nM of forward and reverse primers, and 125 nM of TaqMan probe in 1 × PCR master mix (FastStart Essential DNA Master; Roche, Basel, Switzerland). PCR was performed under the following conditions: 10 min at 95 °C, 50 cycles of 10 s at 95 °C, and 1 min at 60 °C. For the core samples, each reaction contained 5 μL of extracted eDNA solution, a final concentration of 900 nM of forward and reverse primers, and 125 nM of TaqMan probe in 1 × TaqMan Environmental Master Mix 2.0 (Thermo Fisher Scientific, Massachusetts, USA). PCR was performed under the following conditions: 2 min at 50 °C, 10 min at 95 °C, 60 cycles of 15 s at 95 °C, and 1 min at 60 °C. PCR was performed in triplicates for each extracted DNA sample. Triplicates of pure water instead of the eDNA solution were used for each PCR performance as a PCR negative control. All PCR negative controls were below the detection level.
As a standard for quantification, we used a linearized plasmid containing synthesized artificial DNA fragments of the cytochrome b (CytB) gene sequence of jack mackerel or cytochrome C oxidase subunit I (COI) gene sequences of moon jellyfish and Pacific sea nettle, including target regions. The dilution series of 3.0 × 101–3.0 × 104 was run in PCR in triplicate to obtain quantification curves. Quantification was accepted only when the fitted R2 value was above 0.99 on the quantification curve. The average of the PCR replicates was used to represent the eDNA concentration in each sample. eDNA concentrations were expressed as the number of copies per gram of samples in both water and sediment. As contamination precautions, water filtration, DNA extraction, and PCR reactions were performed in separate rooms, and persons entering one of the above three rooms were not permitted to enter the other rooms.
Evaluation of PCR inhibitors
Sediment often contains chemicals that inhibit the PCR process. An analysis using an internal positive control (IPC) was conducted to confirm that the eDNA extraction kit successfully removed such inhibitive chemicals. DNA of lambda phage that was not present in the environment was used as the IPC46. Water and sediment samples (n = 12 for each) in the experimental tanks on day 14 after the introduction of fish were used for this experiment. We placed 300 copies of lambda phage DNA in the extracted eDNA with the primer–probe set in the test group, whereas pure water (instead of extracted eDNA) was placed in the control group (n = 3). PCR amplification of the test and control groups was compared, defining delta Ct as the difference in the number of threshold cycles (Ct values) in the PCR between samples with and without extracted eDNA. Delta Ct in the water samples ranged from − 0.49 to + 2.93 cycles, and from − 0.39 to + 0.28 cycles in the sediment samples (Supplementary Table S6). These values were less than + 3 cycles, previously proposed as criteria of inhibition12, and thus the inhibition was negligible in the present method.
Analysis of polycyclic aromatic hydrocarbons (PAHs) for detecting tsunami signature
The sampled sediment cores (one at St. 1 and three at St. 3) were analyzed to quantify PAHs as a tsunami signature. Specimens from every two layers were used for the analysis.
Five hundred microliters of mixed acetone solution containing 5 μg mL−1 each of naphthalene-d8, acenaphthene-d10, fluorene-d10, anthracene-d10, fluoranthene-d10, pyrene-d10, and chrysene-d12 as surrogate standards was added to a centrifuge tube containing 1 g of sediment. The analytes were extracted twice by shaking for 10 min with acetone (10 mL). Supernatants mixed with 60 mL of saturated NaCl solution were transferred to a separatory funnel. The analytes were extracted twice with 10 mL hexane, and the organic layer was combined. This layer was then dried over anhydrous Na2SO4 and concentrated to trace level using a rotary evaporator. The solution was concentrated to 1 mL under a nitrogen atmosphere and cleaned using a Florisil Sep-Pak column (Waters Association Co., Ltd.). The Florisil Sep-Pak cartridge for clean-up was washed with 10 mL of hexane. A hexane solution containing the analytes followed by 10 mL of hexane/acetone (99/1) solution were passed through the prewashed cartridge. After the addition of 100 μL of 1 mg L−1 atradine-d5 as an internal standard, the eluate was carefully evaporated with a stream of nitrogen up to 1 mL. The analytes were determined using gas chromatography–mass spectrometry (GC/MS).
A Hewlett-Packard 6890 series gas chromatograph equipped with a mass spectrometer (5973 N) was used for PAH analysis. The separation was carried out in a capillary column coated with 5% phenyl methyl silicone (J&W Scientific Co., 30 m length × 0.25 mm i.d., 0.25 μm film thickness). The column temperature was maintained at 50 °C for the first minute and then increased to 290 °C at 20 °C min−1 and to 310 °C at 10 °C min−1. Finally, the column temperature was maintained at 310 °C for 10 min. The interface temperature, ion source temperature, and ion energy were 280 °C, 230 °C, and 70 eV, respectively. Selected ion monitoring was operated under this program. The monitoring ions of 128 (127) for naphthalene, 152 (151) for acenaphthylene, 153 (152) for acenaphthene, 166 (165) for fluorene, 178 (176) for phenanthrene and anthracene, 202 (203) for fluoranthene and pyrene, 228 (229) for benzo[a]anthracene and chrysene, 252 (253) for benzo[b]fluoranthene, 252 (281) for benzo[k]fluoranthene, and benzo[a]pyrene, 276 (207) for dibenzo[a,h]anthracene, indeno[1,2,3-cd]pyrene, and benzo[g,h,i]perylene, were used to quantify the concentrations of PAHs; qualifier ions are indicated in parentheses. One microliter of the sample was injected by splitless injection.
Data analysis
Concentration of eDNA in the water and sediment of experimental tanks after the introduction of fish was analyzed by repeated-measures (rm) ANOVA; ‘days after the introduction of fish’ was defined as the explanatory variable, ‘concentration of eDNA’ as the response variable, and the ‘triplicates of petri dishes’ as a random factor. Then, eDNA concentrations among days were compared using Tukey’s HSD test. Homoscedasticity in eDNA content was improved by log 10 (x + 1) transformation. The decrease in eDNA in water and sediment samples after the removal of fish was also analyzed by rm ANOVA in both the test and control tanks. A comparison of the eDNA concentrations between the test and control tanks was also conducted by rm ANOVA after the removal of fish. All analyses were performed in R ver. 3.4.2 (using the packages of lmerTest and multcomp)47,48,49.
Concentration of eDNA in water and sediment samples after introduction and removal of fish was fitted to eight candidate models as log X (y = a + b * ln(x)), log Y (y = exp(a + b * x)), asymptotic (y = a * x/(1 + b * x)), reciprocal (y = a + b/x), power law (y = a * x ^ b), exponential (y = a * exp(b * x)), and exponential decay (y = a + b * exp(c * x)) using the “nls” function of R. Models with the lowest AIC values were listed, and regression lines were drawn by Kaleida Graph 4.5 (Hulinks, Tokyo, Japan).
We tested whether the concentration of jellyfish eDNA was highest in the layers immediately above the signature of the tsunami in the sediment cores collected at St. 3. The depth of peak PAHs was identified in each core, and this was considered to represent the timing of the tsunami. Core samples of eDNA were then divided into the following three parts: (1) upper, including the upper half of the core above the PAH peak, representing recent sedimentation; (2) middle, including the lower half of the core above the PAH peak, representing sedimentation immediately after the tsunami; and (3) lower, including the layers of PAH peak and below, representing sedimentation at the timing of or prior to the tsunami. Concentrations of jellyfish eDNA of each species were compared among these three parts by nested ANOVA (layers nested in triplicate cores) followed by Tukey’s HSD test.
Source: Ecology - nature.com
