Skip to main content

eDNA data set from 63 locations of Okinawa Island

Cite this dataset

Satoh, Noriyuki (2023). eDNA data set from 63 locations of Okinawa Island [Dataset]. Dryad.


Coral reefs have the highest biodiversity of any marine ecosystem in tropical and subtropical oceans. However, scleractinian corals, keystone organisms of reef productivity, are facing a crisis due to climate change and anthropogenic activities. A broad survey of reef-building corals is essential for world-wide reef preservation. To this end, direct observations made by coral-specialist divers can be supported by a more convenient method. We improved a recently devised environmental DNA (eDNA) metabarcoding method to identify more than 43 scleractinian genera by sampling 2 L of surface seawater above reefs. Together with direct observations by divers, we assessed the utility of eDNA at 63 locations spanning ~250 km near Okinawa Island. Slopes of these islands are populated by diverse coral genera, whereas shallow “moats” sustain fewer and less varied coral taxa. Major genera recorded by divers included Acropora, Pocillopora, Porites, and Montipora, the presence of which was confirmed by eDNA analyses. In addition, eDNA identified more genera than direct observations and documented the presence of new species. This scleractinian coral-specific eDNA method promises to be a powerful tool to survey coral reefs broadly, deeply, and cost-effectively.


(a) Sampling

Reef monitoring was conducted at 63 coral reef locations in the latter half of 2021 by simultaneous diving observations of corals and collection of surface seawater above these reefs. Names of monitored locations, survey dates, geomorphic classification, latitude, longitude, types of coral community, and major coral genera observed are summarized in electronic supplementary materials, table S1, figure 1, and figure S1. Most locations were reef slopes, 3–10 m in depth, and some were inner reef edge moats, 1–3 m in depth (figure 1; electronic supplementary material, table S1).

Two experienced coral-reef specialists snorkeled around reefs and recorded coral community types and dominant coral genera (species), according to the conventional Monitoring Sites 1000 Project conducted annually by the Ministry of the Environment of Japan ( Types of coral community included “multi-species mixed”, “specific species”, and “not determined”, where the first and second categories were distinguished by dominant species and their relative abundances (electronic supplementary materials, table S1 and figure S1). Dominant species included tabular Acropora, branching Pocillopora, and massive Porites. Monitoring of each site required about 30 min with repeated snorkeling. Each diver recorded coral genera based on his or her taxonomic experience. At the end of each day, the divers discussed their observations to reach a consensus regarding major coral genera observed at each site (electronic supplementary material, table S1). Coral coverage of individual reefs was beyond the scope of the present study and was not assessed

At the same time that divers were observing corals at a given coral-reef slope, 3 x 2 L of surface seawater were collected by eDNA researchers in a boat. At moats, the same number of samples was collected by throwing a bucket into the sea. To avoid contamination of seawater samples, containers and buckets were washed with fresh water prior to every sampling. Two-liter samples were filtered individually through 0.45-μm Sterivex filters (Merck), followed by addition of 1 mL of RNAlater (Qiagen) to the filtrate to prevent DNA degradation. Filters were maintained at 4℃ before transfer to a  -20℃ freezer in the laboratory. Due to heavy waves, we were unable to collect a seawater eDNA sample at site #42.

(b) eDNA extraction, PCR amplicons, and sequencing

eDNA in Sterivex filters was extracted following the Environmental DNA Sampling and Experiment Manual Version 2.1. Extracted eDNA samples were PCR-amplified using primers, Scle_12S_Fw (5′ -CCAGCMGACGCGGTRANACTTA-3′ ) and Scle_12S_Rv (5′-AAWTTGACGACGGCCATGC-3′), for mitochondrial 12S rRNA genes, as described in Shinzato et al. (2021). These primers were designed to identify 36 scleractinian coral genera, including Acanthastrea, Acropora, Anthemiphyllia, Astreopora, Coscinaraea, Crispatotrochus, Dipsastraea, Echinophyllia, Favites, Fimbriaphyllia, Fungia, Galaxea, Goniastrea, Goniopora, Heliofungia, Herpolitha, Hydnophora, Isopora, Leptoria, Lobophyllia, Montipora, Orbicella, Pachyseris, Pavona, Pectinia, Platygyra, Plesiastrea, Pocillopora, Polycyathus, Polyphyllia, Porites, Psammocora, and Turbinaria. All genera are common at Okinawa Archipelago (OA) except for Anthemiphyllia (, Crispatotrochus (, and Polycyathus (, though they are reportedly present at OA. Orbicella represents the Atlantic Ocean genus, of which Pacific Ocean species are Astrea and Leptastrea, and they are common at OA.

Genomic DNA isolated from an Acropora tenuis colony was used as a positive PCR control (data not shown). Tks GflexTM DNA Polymerase (Takara) was used for PCR amplification. PCR cycling conditions were 1 min at 94, followed by 35 cycles of 10 s at 98, 15 s at 60, and 30 s at 68, with an extension of 5 min at 68 in the final cycle. PCR products were extracted and cleaned with a FastGene Gel/PCR Extraction Kit (NIPPON Genetics). Amplicon sequencing libraries of cleaned PCR products were prepared using a KAPA Hyper Prep Kit (NIPPON Genetics Co., Ltd.) without fragmentation. Libraries were multiplexed and 300-bp paired-end reads were sequenced on a MiSeq platform (Illumina) using a MiSeq Reagent kit v3 (600 cycles). The number of sequence reads, total base-pair length, and average and maximum length of reads of each sample are shown in electronic supplementary material, table S2.

(c) Bioinformatic analysis

The analysis was carried out as described in Shinzato et al. (2021). Briefly, after removal of low-quality bases (Phred quality score <20) and Illumina sequencing adapters, the remaining sequences were merged using USEARCH, version 11.0.667. Then, de-noised (error-corrected) operational taxonomic units, called ZOTUs (zero-radius operational taxonomic units), were prepared for each sample. ZOTU sequences from all samples were concatenated and clustered using CD-HIT-EST version 4.6 with 100% nucleotide identity. Clustered, unique ZOTU sequences were used for the database for mapping. Merged sequences from each sample were mapped to the clustered ZOTUs and numbers of mapped sequences for each ZOTU were counted using the USEARCH “otutab” command with 99% percent identity (-id 0.99).

Identification of ZOTUs originating from scleractinians were selected based on criteria described in our previous study [15]. After selecting scleractinian ZOTUs, mapped reads from the same genera were combined. To remove possible errors and contamination, only genera with more than 0.1% of the total number of mapped reads in a given sample were considered. To infer similarities between samples and sampling locations, hierarchical clustering based on the ward D method was performed using percentages of coral genera. Numbers of ZOTUs at sampling locations are shown in electronic supplementary materials, table S2. A rough estimation suggested that 1% of ZOTUs was supported by 28.2 ± 12.3 sequence reads.

During direct observations, divers identified Stylophora (electronic supplementary material, table S1), which was not included in our previous eDNA study, since the mt genome sequence of Stylophora pistillata was deposited in the NCBI database only on June 6, 2021. Therefore, we updated the previously reported informatic tools so that Stylophora was also included in the present analysis. Simultaneously, an additional 8 genera were also included (Agaricia, Catalaphyllia, Diploastrea, Micromussa, Oulastrea, Paraechinophyllia, Physogyra, and Thalamophyllia). Therefore, this method is likely able to identify 45 genera, and 43 of these were detected in the present study (electronic supplementary material, figure S2).

Numbers of genera detected by direct observation and eDNA metabarcoding method were statistically compared for slopes and moats, respectively, using the Wilcoxon signed-rank test using R version 4.2.1 (R Development Core Team 2022). Overlap coefficients between genus detected by direct observation, and eDNA method was calculated for each sampling point.


Ministry of the Environment