https://orcid.org/0000-0002-6020-334X
Figures
Abstract
Ocean acidification (OA), driven by rising atmospheric CO2, presents a serious threat to marine biodiversity, especially within coral reef ecosystems. Natural analogue sites, such as the high-pCO2 seep at Iōtorishima Island in Japan, offer insights into future conditions. This study investigated the holobiont communities of Symbiodiniaceae and bacteria in the zoantharian Palythoa tuberculosa at Iōtorishima and compared them to specimens from control sites in Okinawa and Hawaiʻi. Using amplicon sequencing of the dinoflagellate internal transcribed spacer 2 (ITS2) region of ribosomal DNA and microbial 16S rRNA gene, we detected significant shifts in both Symbiodiniaceae and bacterial communities under high-pCO2 conditions at Iōtorishima. Specifically, P. tuberculosa at the seep site had reduced Symbiodiniaceae diversity, predominatly featuring Cladocopium C1 and C3 types. Additionally, its bacterial communities showed lower richness with distinct taxonomic profiles, including increased levels of Mollicutes and Vibrio spp. These results highlight the potentially adverse effects of OA on hexacoral holobionts and emphasize the need for detailed, high-resolution studies across various holobiont species and geographic locations. The shifts observed specifically in Symbiodiniaceae and bacterial communities at the Iōtorishima seep suggest that holobionts may exhibit plasticity in response to environmental stress, which has implications for resilience and adaptation of zoantharians and other reef organisms amid climate change. This research provides crucial baseline data for predicting future coral reef compositions in an OA-affected world.
Citation: Huerlimann R, Wee HB, Alves dos Santos M, Kise H, Mizuyama M, Dudoit ‘, et al. (2025) Shifts in coral reef holobiont communities in the high-CO2 marine environment of Iōtorishima Island. PLOS Clim 4(7): e0000665. https://doi.org/10.1371/journal.pclm.0000665
Editor: Karl D. Castillo, University of North Carolina at Chapel Hill, UNITED STATES OF AMERICA
Received: November 12, 2024; Accepted: June 4, 2025; Published: July 3, 2025
Copyright: © 2025 Huerlimann 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 bacterial 16S rRNA data and the Symbiodiniaceae ITS data has been deposited on FigShare under DOI: 10.6084/m9.figshare.28756385 The bacterial data includes raw fastq files, a metadata table, a table with taxonomic assignments, and an ASV table with raw, unfiltered read counts. The Symbiodiniaceae data contains raw fastq files and a metadata table.
Funding: This project contributes towards the International CO2 Natural Analogues (ICONA) Network. Generally, the work at Iōtorishima was supported by an OIST KICKS grant entitled “Japanese volcanic CO2 vents—natural laboratories to study the behaviour and adaptation of marine organisms to acidifying oceans” (Awarded to TR, HK1, and JDR). Part of the funding came from the Japan Society for the Promotion of Science (JSPS) Core-to-Core Program (JPJSCCA20210006 to JDR, TR and HK1). JDR was partially supported by a JSPS Grant-in-Aid for Transformative Research Areas entitled “Environmental, ecological, and genetic observations of coral reef Symbiodiniaceae-host holobiont symbioses” (23H03821 to JDR). HK2 was also supported in part by JSPS KAKENHI grant (23KJ2206 to HK2) from the Japan Society for the Promotion of Science. HK2, MM, GYS and AI were supported by Environmentally-conscious Developments and Technologies (E-code) research laboratory at the National Institute of Advanced Industrial Science and Technology (AIST). (HK1: Haruko Kurihara; HK2: Hiroki Kise)
Competing interests: The authors have declared that no competing interests exist.
Introduction
Under increasing anthropogenic emissions of CO2, future oceans are expected to have higher pCO2 levels and lower pH compared to today, a process called ocean acidification (OA). Over the past thirty years, numerous studies have documented a wide range of potential negative impacts of OA on various marine organisms (e.g., [1–5]). However, significant knowledge gaps remain, and a comprehensive understanding of future oceans ecosystem is still a subject of debate [6]. One promising approach to improve our predictions of future changes in marine ecosystems is the study of natural analogues.
Natural analogues of future oceans are typically areas where current environmental conditions resemble those projected for the future exhibiting higher temperature and lower pH [6]. To isolate the effect of OA itself, locations with CO2 seeps resulting in higher pCO2 conditions and lower pH [5,7], such as Vulcano Island in Italy [7] and Ambitle in Papua New Guinea [8], can be investigated. In addition, some enclosed bays with limited water exchange combined with decaying biomass also result in higher pCO2 conditions and lower pH, as well as having higher seawater temperatures. Notable examples of enclosed bay natural analogue sites include Nikko Bay in Palau [9] and Bourake in New Caledonia [10].
Shallow water subtropical and tropical coral reef ecosystems harbor the highest levels of marine biodiversity [11], primarily constructed by the skeletal secretions of organisms, such as zooxanthellate scleractinian corals, coralline algae, and foraminifera (e.g., Jindrich [12]). These bioengineers, along with their calcium carbonate and aragonite skeletons, may be particularly susceptible to increased pCO2 concentrations and decreased pH levels [1]. Therefore, research at natural analogue sites within coral reefs in subtropical and tropical regions is especially crucial for gaining knowledge to better protect marine biodiversity in the face of future environmental conditions.
Research on subtropical and subtropical natural analogue sites has been relatively limited so far, and the findings vary depending on the species and study [6]. Benthic studies have primarily focused on scleractinian corals [6], which is understandable given their key role as ecosystem engineers. Some studies have reported a general decline in scleractinians, accompanied by increases in other anthozoans such as soft corals [13], zoantharians [14], and sea anemones [15]. However, these other groups have not been investigated as extensively as scleractinians. Research on the photosymbionts of these anthozoans, the Symbiodiniaceae dinoflagellates [16], has produced contrasting results. While some studies found that Symbiodiniaceae diversity remained unaffected by natural analogue conditions (e.g., in scleractinians; [17]), others have indicated negative impacts on diversity [18]. Additionally, there have been very few studies on other holobiont components such as bacteria, protists, or viruses at natural analogues (e.g., [19]). Research of bacterial communities in scleractinian corals has been more common, but results have been variable, with one study showing a decline in alpha-diversity and an increase in Endozoicomonas as a potential way to cope with heat stress in three coral genera (Acropora, Pocillopora and Porites; [20]), while another study showed an increase in alpha- and beta-diversity under combined heat stress and disturbance (Porites and Montipora; [21]) and a third study showed no changes under near-future OA conditions (Acropora and Seriatopora; [22]). These contrasting results and the lack of studies including non-scleractinians indicate a data gap. Thus, there is an urgent need for advanced, large-scale, and multi-faceted datasets focusing on non-scleractinian benthos from coral reef natural analogues, especially given the apparent resilience of some of these groups to low pH conditions [13–15].
One such natural analogue site is located at the uninhabited Iōtorishima Island in the Ryukyu Islands of southern Japan. The Iōtorishima site features a CO2 seep over a shallow coral reef [13], where environmental conditions within the inner lagoon vary based on seep activity, seasons, and tides, and sometimes leading to high pCO2 levels [13]. Initial studies at Iōtorishima revealed an increase in soft coral abundance [13] at the high-CO2 site, while more recent research has documented numerous zoantharians [23] and the “living fossil” octocoral Nanipora [24], related to blue corals. These finding suggests that understudied non-scleractinian groups may often thrive in marginal and unique environments [14,15], highlighting the need for further investigation of these taxa.
In this study, we used high-throughput sequencing to examine the Symbiodiniaceae and bacterial communities of the zoantharian Palythoa tuberculosa at Iōtorishima, Okinawa Island, and Hawaiʻi. By comparing the composition of microbes associated with P. tuberculosa across these regions, we gain a valuable insights into how holobionts respond to low pH and high pCO2 levels.
Materials and methods
Iōtorishima site and specimen collection
Iōtorishima Island (hereafter Iōtorishima) is a small uninhabited island in the middle Ryukyu Islands in the East China Sea. The island is approximately 65 km to the west of Tokunoshima Island of Kagoshima, and is the northernmost island in Okinawa Prefecture. Iōtorishima was inhabited until 1958 when volcanic activity led to the evacuation of residents. In recent years, despite being isolated and uninhabited, the island has attracted marine scientific research due to a CO2 vent in a shallow fringing reef on the southeast coast of the island, acting as a natural analogue for OA [13,18,23,24].
As detailed in [24], Iōtorishima was visited from 14 to 16 September 2020 (Fig 1, Table 1). Field surveys focused on the shallow CO2 seep at depths of 0.5 to 2 m within the shallow coral reef lagoon [13], centered around 27° 52′ 11.8″ N, 128° 14′ 01.8″ E. During benthic surveys, Palythoa tuberculosa (Esper, 1805) specimens (n = 14) were collected within the shallow CO2 lagoon by snorkeling (high pCO2). We also collected specimens of the same species from a control site (n = 7) within the lagoon at the same depths with lower CO2 levels; this was the same as described in [18]. P. tuberculosa is easily identifiable in the field by its ‘immersae’ colony morphology [25,26]. Tissue samples were individually preserved in >95% ethanol and stored at -20°C until further analyses.
Map data from OpenStreetMap (https://www.openstreetmap.org/copyright).
The map was manually created by HBW in QGIS (3.34 Prizren. 2024. QGIS Geographic Information System. QGIS Association. http://www.qgis.org) based on data from OpenStreetMap (https://www.openstreetmap.org/copyright).
Mizugama (Okinawa) site
Palythoa tuberculosa samples were collected on 17 August 2017 (n = 13) and 9 April 2018 (n = 16 specimens) at the Mizugama seawall reef (Fig 1, Table 1). The specimens were collected via SCUBA diving or reef walking at low tide. Specimens were collected at different distances from the Hija River mouth and at various depths: specimens No. 1–5 (0 m from river mouth, 10 m depth), No. 6–10 (0 m from river mouth, 2 m depth), No. 11–15 (tidal pool inside river mouth, 0.5 m depth), No. 16–20 (500 m from river mouth, 10 m depth), No. 21–25 (500 m from river mouth, 2 m depth). Collected specimens were preserved in ethanol 99% and transferred to the laboratory and stored at -20°C until subsequent analyses.
Hawaiian Islands sites
We sampled 25 individuals of Palythoa tuberculosa from the north shore of Oʻahu and Kona coast of Hawaiʻi Island (Fig 1, Table 1) between July 2018 to September 2020 at depths of 1–2 m. Collections were made via snorkeling or SCUBA diving, and tissue samples were individually preserved in salt-saturated DMSO (20% DMSO, 0.25M EDTA, pH 8.0 saturated with NaCl, as per [27]) buffer and stored at room temperature until further analyses.
DNA extraction and high-throughput sequencing
Genomic DNA of Palythoa tuberculosa collections was extracted using the Qiagen DNEasy PowerSoil kit (Tokyo, Japan) following the manufacturer’s protocols. The concentration of the genomic DNA samples was analyzed using a Qubit Fluorometer with the Qubit 1X dsDNA HS Assay kit (Invitrogen, Japan).
For Symbiodiniaceae, amplicon libraries were generated from the internal transcribed spacer 2 region of ribosomal DNA (ITS2 rDNA). The first PCR was performed with the primer set SYM_VAR_5.8S2/SYM_VAR_REV [28] with an overhang adapter sequence for the MiSeq platform (Illumina, Los Angeles, CA, USA). Each 20 µl PCR included the following components: 1.0 µl of temperate DNA, 1.2 µl of each forward and reverse primer, 0.2 µl of TaKaRa Ex Taq™, 2.0 µl of 10 × Ex Taq Buffer, 1.6 µl of dNTP mixture (Takara Bio Inc., Shiga, Japan), and 12.8 µl of nuclease-free water. The PCRs were run with a temperature profile of 98°C for 2 min, followed by 35 cycles of 98 °C for 10 s, 56°C for 30 s, and 72°C for 30 s, with a final extension at 72 °C for 7 min. Generated amplicons were purified using Ampure XP beads (Beckman Coulter, Brea, CA, USA). Index PCR was performed to add Illumina sequencing adaptors and sample index sequences, followed by second purification using Ampure XP beads. Quantification of the concentration of purified index PCR products was measured by the Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). These PCR products were pooled at equimolar concentrations. Amplified PCR products were sequenced on an Illumina MiSeq platform at the National Institute of Advanced Industrial Science and Technology, using a V2-500 cycle kit to generate 2 × 250 bp paired-end reads.
For bacteria, we amplified the V3-V4 region of the bacterial 16S rRNA gene using the primer set P341F/P805R [29] with overhang adaptor sequences. Each 20 µl PCR included the following components: 1.0 µl of temperate DNA, 0.4 µl of each forward and reverse primer, 0.4 µl of MightyAmp DNA Polymerase, 10 µl of 2x MightyAmp Buffer (Takara Bio Inc., Shiga, Japan), and 7.8 µl of nuclease-free water. The PCRs were run with a temperature profile of 98 °C for 2 min, followed by 10 cycles of 98 °C for 10 s, 65–56°C for 15s (decrease temperature by 1°C per cycle) and 68°C for 1 min, and 25 cycles of 98 °C for 10 s, 55°C for 15s and 68°C for 1 min, with a final extension at 68°C for 2 min. Index PCR was performed to add Illumina sequencing adaptors and sample index sequences. Index PCR products were purified using Ampure XP beads and quantified using the Qubit dsDNA HS Assay Kit, followed by pooling at equimolar concentrations. The purified library was sequenced on an Illumina Miseq platform, using a V3-600 cycle kit to generate 2 × 300 bp paired-end reads.
Bioinformatical analyses
For Symbiodiniaceae, the demultiplexed paired-end reads were analyzed in local using SymPortal analytical framework [30], a platform for phylogenetically resolving Symbiodiniaceae taxa using ITS2 amplicon data. Sequence quality control (QC) of paired-end reads was performed utilizing mothur 1.39.5 [31], blast+ suite of executables [32] and minimum entropy decomposition [33] to filter artifactual and non-Symbiodiniaceae sequences from dataset. Post QC sequences from each sample were loaded into the SymPortal database to identify the specific sets of sequences called defining intragenomic variants (DIVs).
For bacteria, the demultiplexed paired-end sequence reads were analyzed in the QIIME2 v2022.8.3 framework [34], using several plug-in programs [35]. End sequences, including primer and adopter sequences, were trimmed from the 5’ and 3’ ends using Cutadapt v4.1 [36] (via q2-cutadapt) for up to 10 repeats. If the primer sequences were not found or the trimmed length of the sequences was less than 100 bp, the reads were discarded. Quality control, including quality filtering, denoising, correction of reading errors, merging of paired-end reads, and removal of non-biological sequences (including chimeras), was performed by DADA2 v4.1.3 [37] (via q2-dada2) with the following custom values. For denoising, the maximum expected error values, ‘maxEE’ [38], were set to 2 for forward reads and 5 for reverse reads. The truncated length at the 3’ end of forward and reverse reads was independently determined by the base position corresponding to a Phred quality score of less than 20 in the first quartile of total reads. After quality control, a count table of representative sequences was generated by dereplication of the same amplicon sequence variant (ASV).
Taxonomic assignment to the ASV was performed using a pre-trained Naive Bayes classifier [39] with reference to the 16S ribosomal RNA sequence database SILVA v138.1 [40,41] curated by RESCRIPt [42], and then operational taxonomic units (OTUs) were determined for each ASV using scikit-learn [43], a machine learning classifier plugin in Qiime2.
The taxonomic assignment identified one genus as “Candidatus Hepatoplasma”; however, in our experience this is a misidentification due to the uncertainty within this group of marine bacteria. The alignment of our ASV sequences against sequences within this group showed a close relationship with the poorly researched clade Metamycoplasmataceae [44]. Considering the importance of this group in our findings, coupled with the uncertainty in the taxonomy, we decided to rename the ASVs as “Unknown Genus within Mollicutes” or in places “Mollicutes” for short (see S1 Fig). Future work will hopefully further resolve this important bacterial group.
The average number of reads per sample was 44,344 ± 26,807 (mean ± stdev) before filtering, and 38,181 ± 26,535 (mean ± stdev) after filtering. Two samples from Okinawa Mizugama and one sample from Hawaiʻi Kona were removed due to low read counts (Mizugama: 1,856 and 61 reads, Kona: 136 reads) and distant clustering from other related samples. Other filtering included removing ASVs (amplicon sequence variants) with no taxonomic assignment at the phylum level (i.e., “NA”), filtering ASV assigned to chloroplast at the order level (since they are of plant origin), mitochondria at the family level (since they are of eukaryotic origin, potentially including Palythoa), and Cutibacterium at the genus level (human contamination). To generate the bar plot in Fig 4 without introducing bias from varying sequencing depth, the relative abundance of each sample was calculated individually. The samples were then grouped by location, and the overall relative abundance was calculated as a percentage.
Statistical analyses
The NGS reads of family-level taxonomic clustering of bacteria and genus-level clustering of Symbiodiniaceae for Palythoa tuberculosa were optimised and organised based on ASV and DIV, respectively. One table each was produced for bacteria and Symbiodiniaceae data, and the reads were transformed into percentage abundance of composition for each specimen/individual. The specimens were categorized based on the locations they were collected: the acidified coral reef sites of Iōtorishima (seep), and the non-acidified coral reef sites of Iōtorishimal (control), Okinawa (Mizugama), and Hawaiʻi (Kona, Hawaiʻi Island and North Shore, Oʻahu).
Both bacteria (Family, Species, and ASVs) and Symbiodiniaceae datasets were analyzed using R (version 4.3.2; [45] via RStudio V2023.06.1 Build 524 [46] (RStudio, 2023). Alpha-diversity was analyzed using the phyloseq package [47].
Ordinated (Bray-Curtis Dissimilarity) tests of beta-diversity dispersion and distribution (permutest) based on locations were conducted for both datasets using relative abundances. Permutative ANOVA (PERMANOVA) analyses via adonis2 (vegan package, version 2.6.4, [48,49]) were conducted to test the dissimilar distance of composition among the sites, followed by a post-hoc test (pairwiseAdonis package). Similarity Profiling (SIMPER) analyses were conducted to test the apriori grouping of the compositions via clustering. The distance differences among the specimens and locations were represented in Non-parametric multidimensional scaling (nMDS).
Richness (called Observed in phyloseq) and evenness were calculated using phyloseq. In order to calculate Evenness, the Shannon diversity index was divided by log(Observed). Both alpha diversity measures were significant for Shapiro-Wilk normality test, therefore the non-parametric Kruskal-Wallis rank sum test and pairwise comparisons using Wilcoxon rank sum exact test were used for the statistical analysis.
Figures were created via R (base and ggplot2 package [50]). All post-hoc tests were conducted with Holm’s correction method. The alpha value of all statistical tests was set at α = 0.05.
Results
Alpha diversity - Bacteria
In terms of alpha diversity, there was a significant difference (Kruskal-Wallis chi-squared = 28.2, df = 4, p-value = 1.115e-05) in the bacterial richness at the ASV level between all locations. A post-hoc comparison showed that the Iōtorishima Seep location had a significantly lower richness than any other site (Fig 2). In contrast, the post-hoc analysis of the evenness showed no significant differences between locations (Fig 2), despite an overall significant result (Kruskal-Wallis chi-squared = 13.0, df = 4, p-value = 0.01124). Other diversity indices metrics can be found in S2 Fig.
(statistical significance: * = 0.05, ** = 0.01, *** = 0.001; outliers are marked with filled circles).
Beta diversity - Bacteria
For the bacterial beta diversity, there were significant differences among the locations regarding the family composition of bacteria hosted by Palythoa tuberculosa (PERMANOVA: R2 = 0.204, F = 4.032, p < 0.001). Tests of composition dissimilarity between groups (with Holm’s correction) showed only Hawaiʻi Island Kona recorded no significant differences in bacterial family composition with Iōtorishima Control and Oʻahu North Shore (Table 2A, pairwise Adonis: p = 0.356). The species composition of bacteria in P. tuberculosa was significantly different among the locations (PERMANOVA: R2 = 0.196, F = 3.831, p < 0.001). Pairwise tests showed only Hawaiʻi Island Kona and Okinawa Mizugama bacterial species compositions were not significantly different (Table 2B, pairwise Adonis: R2 = 0.053, p = 0.053). Lastly, the highest resolution was found at the ASV level, which showed significant differences among (PERMANOVA: R2 = 0.164, F = 3.0872, p < 0.001) and between locations (Table 2C, pairwise Adonis: p < 0.010).
nMDS showed more defined separations of dissimilarities among the locations as the OTU resolution increased from family (Fig 3A, stress = 0.204) to ASV (Fig 3E, stress = 0.216). Iōtorishima Seeps were observed to have more dissimilarities with Iōtorishima Control and Hawaiʻi Island Kona for species (Fig 3C, stress = 0.219); and ASVs saw the same separation among the locations in addition to statistically significant differences with Okinawa Mizugama. This nMDS also showed that the site closest to Iōtorishima Seeps was Oʻahu North Shore. nMDS graphs clearly showed that the Iōtorishima Seep had the most distinct composition among all the locations in the study. However, there were no significant beta-dispersions of the diversity among the locations regarding family (Permutest: Df = 4, F = 0.812, p = 0.519), species (Permutest: Df = 4, F = 0.847,p = 0.479), or ASV (Permutest: Df = 4, F = 0.315,p = 0.858).
Blue = Iōtorishima Control, Red = Iōtorishima Seep, Green = Okinawa Mizugama, Purple = Hawaiʻi Island Kona, Pink = Oʻahu North Shore. Ovals represent ellipsoid hulls enclosing all points within each location.
Mizugama was included as a reference for a highly disturbed area and exhibited the largest environmental variation (see [51]). An analysis of the two sampling years (2017 and 2018) and different depths (Shallow: ≤ 2 m and Deep 8–11 m) showed no strong separation (S3 Fig, nMDS stress = 0.188). However, a statistical analysis showed a significant difference for year (PERMANOVA: R2 = 0.148, F = 3.986, p = 0.001), but not depth (PERMANOVA: R2 = 0.068, F = 1.680, p = 0.091). Despite this significant difference, we decided to combine the data for Mizugama for the analysis. The two main reasons for this were that 1) the Mizugama location has been extensively described in previous publications [51] and is not the focus of the current research; and 2) this location was included to provide the contrast of a highly disturbed area. Lastly, as the beta-diversity analysis of the bacterial community (Fig 3) shows, Mizugama formed a distinct cluster within the larger dataset.
Bacterial community composition
At the phylum level, the bacterial composition was dominated by Proteobacteria, irrespective of location (Fig 4). Unique to the Iōtorishima Seep was the high abundance of Firmicutes, the absence of Cyanobacteria, and the increased abundance of Actinobacteria, comparable to as observed at Oʻahu North Shore.
In terms of family composition, the Iōtorishima Seep location had a high abundance of Corynebacteriaceae, Entomoplasmatales, Moraxellaceae, Staphylococcaceae and Vibrionaceae, especially when compared to the Iōtorishima Control site (Fig 4). On the other hand, the Iōtorishima Control site had a higher relative abundance of Kiloniellaceae and Rhodobacteraceae.
The bacterial composition at genus level showed a high abundance of Mollicutes, Corynebacterium, Moraxella, and Vibrio at the Iōtorishima Seep site (Fig 4). Interestingly, Vibrio was highest at Okinawa Mizugama, which was the most heavily urbanized site, followed by Iōtorishima Seep, and Oʻahu North Shore, while being absent in the remaining two sites. Mizugama showed a unique accumulation of Spiroplasma. As well, the Iōtorishima Control in particular, but to a lower degree also the Iōtorishima Seep, showed a comparatively high concentration of Tistlia, which was uncommon at the other sites.
Beta diversity - Symbiodiniaceae
Most locations were dominated by various types of Cladocopium, with the exception of the Seep and Mizugama, which also showed the presence of several Durusdinium types (S4 Fig). Statistically, there were significant differences among the locations for the Symbiodiniaceae OTU composition (PERMANOVA: R2 = 0.351, F = 8.122, p < 0.001). Tests of composition dissimilarity between groups showed only Iōtorishima Seep P. tuberculosa hosted significantly different Symbiodiniaceae compared to all other locations (Table 3, pairwise Adonis: p < 0.001). Hawaiʻi Island Kona specimens had significantly different Symbiodiniaceae composition with the other Pacific locations, except for with Oʻahu North Shore (Table 3, pairwise Adonis: R2 = 0.152, p = 0.192) Other pairwise comparisons showed no significant differences between locations (Table 3, pairwise Adonis: p > 0.192).
nMDS (Fig 5, stress = 0.094) showed overlapping Symbiodiniaceae composition between Okinawa Mizugama specimens with other locations. While the Iōtorishima Seep Symbiodiniaceae were completely separated from those of Hawaii main island, there was some similarity with other locations (S5 Fig). Furthermore, Hawaiʻi Island Konai specimens had the lowest variation within a site. While agreeing with the beta-dispersal output (Fig 5), Okinawa Mizugama had the highest median of departure from centroid in diversity with a larger dispersal on nMDS.
Purple = Hawaiʻi Island Kona, Blue = Iōtorishima Control, Red = Iōtorishima Seep, Green = Okinawa Mizugama. Ovals represent ellipsoid hulls enclosing all points within each location.
The beta-diversity dispersion of Symbiodiniaceae hosted by Palythoa tuberculosa within each location indicated that the median OTU diversity was closely similar in Iōtorishima Seep, Okinawa Mizugama, and Hawaiʻi Island Kona (Fig 5). Permutation tests for the size of dispersion relative to the centroid showed significance among the locations (Permutest: Df = 4, F = 4.6146, p = 0.004). Post hoc tests on the diversity dispersion of Symbodiniaceae showed significant differences only for the very low OTU diversity variation of Hawaiʻi Island Kona compared to other locations (p < 0.05).
SIMPER analyses showed Symbiodiniaceae Cladocopium were the main driver of the dissimilarity in Symbiodiniaceae OTU composition, specifically with Cladocopium C1, C3, C71 and C1n as the main drivers of differences. Furthermore, C1 and C3 were the dominant drivers of differences between the Iōtorishima Seep and other locations.
Discussion
Our analyses of the Symbiodiniaceae and bacterial compositions in the zoantharian Palythoa tuberculosa from the high pCO2 site at Iōtorishima, compared to a control site on the same island, as well as sites from other locations in Okinawa and Hawaiʻi, revealed both expected and surprising findings, which we explain below. As with any extensive high-throughput sequencing study, some results initially seemed perplexing. However, when interpreted in the context of exisiting knowledge about holobiont Symbiodiniaceae and bacterial compositions, this research provides a strong foundation for future studies and offers important hypotheses on how high pCO2 levels may impact zooxanthellate anthozoan holobionts.
Firstly, the findings from Symbiodiniaceae analyses in this study warrant discussion. P. tuberculosa at the Iōtorishima seep exhibited a mixture of Cladocopium (former Symbiodinium ‘clade C’, [17]) of types C1/C3 and Durusdinium (former Symbiodinium ‘clade D’, [17]) of type D2. This is in contrast to the Iōtorishima control and all other control sites which were dominated by various types of Cladocopium. These findings in the control sites are similar to previous research with P. tuberculosa across the Indo-Pacific Ocean [51–54]. Despite most specimens hosting Cladocopium, the variation among Cladocopium types was the primary factor driving differences among sites, including the nMDS analyses. Specifically, Cladocopium C1 and C3-related types were identified to be the main contributors to the distinct Symbiodiniaceae communities at Iōtorishima compared to other sites. These results demonstrated that the high pCO2 Symbiodiniaceae communities at Iōtorishima were distinct from those at all other sites, with minimal overlap among ellipses. This is particularly notable considering that the “control” specimens at Iōtorishima were collected less than 200 m away from the high CO2 specimens, yet they were more similar to Symbiodiniaceae communities from Oʻahu North Shore and Hawaiʻi Island, over 7,000 km away, as well as to specimens in Okinawa, more than 170 km away. This lower anthozoan community diversity of anthozoans thriving in more stressful environments aligns with recent findings showing higher symbiont variability in corals with lower resilience to stressful environments [55]. This recent research has suggested hosting higher symbiont diversity may come with costs due to competitive interactions among different symbionts, thus lowering holobiont performance [52], and this could possibly also be the case in this study for P. tuberculosa at Iōtorishima seeps.
Based on our results, we conclude that high levels of pCO2 are one of the main causes of differences observed in symbiont and bacterial communities. Combining our results with previous findings at the same seep [18], we hypothesize that elevated CO2 levels may lead to reduced Symbiodiniaceae variability, at least in some anthozoan hosts like P. tuberculosa. Similar findings have also been observed in three species of scleractinian corals in Bourake, New Caledonia, which showed more homogenous Symbiodiniaceae communities compared to colonies at control sites [56]. While the mechanism behind these changes in Symbiodiniaceae diversity under high CO2 levels remains to be unclear, it could be due to the aforementioned costs of competitive interactions among different symbiont types [52], or possiblly due to the endosymbiotic Symbiodiniaceae experiencing higher rates of photosynthesis under lower pH conditions, contributing to stress resistance [56]. Regardless of the mechanism, our results support the hypothesis proposed by Tanvet et al. [56] that “[...] corals from distinct environments often have unique symbiotic partners that could be crucial to support their survival […]”. A deeper understanding of the environmental preferences and adaptations of various Cladocopium types within P. tuberculosa and other host species could help researchers confirm or refute these hypotheses. Recent research on P. tuberculosa around Okinawa has confirmed the presence of different lineages with hypothesized preferences for specific environments [51,57], further supporting this idea.
The bacterial results also reveal interesting patterns. Bacteria, a group that has existed for at least 3.5 billion years, display immense diversity across various phyla, each with its unique biology and a broad range of functions within ecosystems. Although Palythoa species have been reported associated with over 30 bacteria families, there was a distinct bacterial composition associated with specimens from the Indo-Pacific or Atlantic oceans [54]. This diversity was evident in our results, showing that the P. tuberculosa holobiont can encompass a highly diverse assemblage of bacteria. For instance, our P. tuberculosa specimens contained 394 bacterial families in total, with 12 families each contributing ≥ 1% relative abundance. Similarly high levels of bacterial diversity have been observed in many cnidarians, particularly in zooxanthellate anthozoan and scleractinian holobionts (e.g., reviewed in [58]). Therefore, the highly diverse bacterial communities from P. tuberculosa in this study are not unexpected.
However, a closer examination of the bacterial communities reveals several intriguing findings, especially concerning the high pCO2 site. Notably, the bacterial community richness within P. tuberculosa from the high pCO2 site was significantly lower than at any other site. Although family-level analyses did not show many differences across sites, more taxonomically refined analyses at species and ASV levels did reveal distinctions. Additionally, a detailed examination of bacterial community composition, even at the phylum level, indicated potentially important differences (Fig 4). For instance, at the phylum level, the Iōtorishima high pCO2 samples had a higher abundance of Firmicutes compared to other sites. At the family level, greater presence of Entoplasmatales was observed, and at the genus level there was a higher abundance of an unknown genus within Mollicutes. Mollicutes genotypes have been widely reported in various marine species (e.g., seastars [59]) and terrestrial invertebrate (e.g., isopods, [60]). They have also been shown to influence seasonal differences in bacterial communities in rock oysters, and have been implicated in variations in disease resistance [61].
In anthozoans, the few reports on Mollicutes present a different picture compared to our current findings. Mollicutes species have been documented in octocorals from the Great Barrier Reef, where their abundances remained stable over time, even before, during, and after a bleaching event [62]. Similarly, they were reported as stable in Mediterranean gorgonian octocorals over time [63], leading Steinberg et al. (2023) to hypothesize that Mollicutes may play a significant functional role in the microbiome of the host colonies. However, our results contrast with these previous reports. Mollicutes were found in relatively low numbers in both Hawaiʻi and Okinawan P. tuberculosa specimens and were almost absent (< 0.01% relative abundance) at the control site at Iōtorishima, but they were highly abundant (25.6% relative abundance) at the high pCO2 Iōtorishima site. Although Mollicutes may encompass various species or functional groups, and despite the differences from previous anthozoan studies, our findings support a link between this taxon and pH or CO2 levels. It is clear further research is needed into this intriguing matter.
Noteworthy are the changes in the abundance of the family Vibrionaceae, particulary the genus Vibrio. High abundances of Vibrio were observed at Mizugama (15.1% relative abundance), Iōtorishima Seep (6.7% relative abundance), and Oʻahu North Shore (6.0% relative abundance), while Iōtorishima control and Hawaiʻi Island exhibited much lower levels (< 1% relative abundance). Some members of Vibrionaceae are considered potential opportunistic and pathogenic bacteria, linked to several coral diseases (reviewed in [64]). Their increased presence has been associated with environmental changes [65], particularly in immune-comprised hosts (e.g., [66]). Urbanization may also be a factor, particularly at the Mizugama site located at the mouth of the Hija River, which has been reported to be among the most polluted of Okinawa’s rivers [67]. Indeed, the Mizugama site has previously been judged as moderately to highly impacted by anthropogenic impacts [68,69].
Research on natural analogues has gained momentum in recent years due to growing concerns about the impacts of climate change on marine ecosystems [6]. On coral reefs, most of the research has focused on scleractinian corals (e.g., [70]) and fish (e.g., [71]). However, an increasing body of literature suggests that as zooxanthellate scleractinian corals decline, other benthic groups may become dominant on reefs. Reports indicate the potential spread of such “non-coral” reefs due to environmental shifts, involving a wide variety of taxa, including sponges [72], sea anemones [15], corallimorpharians [73], soft corals [13,69], algae [74], and zoantharians [14,23,75]. Unfortunately, for many of these groups, essential biological data such as growth rates, tolerances, sexual reproductive traits, and even accurate taxonomic identification are often lacking [76]. This data gap hinders our ability to accurately model or predict which benthic organisms might replace scleractinian corals under various conditions, severely limiting our capacity to forecast the coral reefs. Therefore, further research in this area is urgently needed [76].
At the same time, it is important to recognize the significant diversity within each of these understudied groups, and researchers should avoid treating them as a single entity [76,77]. In this study, we focused on P. tuberculosa, a common and widely distributed Indo-Pacific species [78]. Recent research suggests that this species is flexible in terms of its heterotrophic performance [79] and symbiont composition, capable of hosting different Cladocopium lineages depending on local conditions, which allows it to thrive in a wide range of habitats [51,57]. However, it would be unwise to assume that all zoantharians, including those common on coral reefs (e.g., genus Zoanthus), will behave similarly, even under low pH conditions. There is also the possibility that P. tuberculosa represents multiple closely related lineages [80].
Finally, there is an even greater scarcity of data on the holobionts of these groups. In fact, there are very few published holobiont datasets for both Symbiodiniaceae and bacteria in P. tuberculosa (with the exception of [81]) and none for natural analogues. Given the abundance of unique “understudied” taxa at natural analogues, it is crucial to urgently aquire data on these holobionts, along with the basic biological information mentioned eariler. Simultaneously, efforts should be made to gather more comprehensive holobiont datasets at higher resolutions than are currently available, covering a broad geographic range.
In conclusion, marginal environments such as the high pCO2 reef at Iōtoroshima have distinct diversity and community compositions of Symbiodiniaceae and bacteria associated with P. tuberculosa, compared with multiple reference sites. Our use of multiple control sides ranging from Okinawa to the Hawaiian islands adds to the robustness of the results. Furthermore, the present study is unique in that it focuses on the holobiont of a zoantharian, which are an understudied taxa, especially in the context of natural analogues of OA. Our findings indicate that different components of the P. tuberculosa holobiont are affected in distinct ways and detecting these variations often requires high-resolution methods and comparisons across both narrow (e.g., nearby control sites) and wide scales (hundreds to thousands of km). To fully understand and better predict the impacts of OA on coral reef ecosystems, it is essential to investigate a wider diversity of organisms, as this work on Palythoa zoantharians has done.
Supporting information
S1 Fig. The tree was generated using Mollicutes ASV (identified as Candidatus Hepatoplasma by DADA2) and published data [42).
The tree was constructed using MAFFT (v7.525) with the following command: “mafft --adjustdirection --globalpair --maxiterate 1000” and IQ-TREE2 (v2.2.2.5) with the GTR + F + R3 model and 1,000 bootstrap iterations.
https://doi.org/10.1371/journal.pclm.0000665.s001
(PNG)
S2 Fig. Diversity indices metrics for Observed, Chao1, ACE, Shannon, Simpson, Inverse Simpson and Fisher as calculated by the phyloseq package.
https://doi.org/10.1371/journal.pclm.0000665.s002
(PNG)
S3 Fig. nMDS of the Mizugama location only, coloured by year of sampling (2017 and 2018) and Depth (Shallow: ≤ 2 m and Deep: 8–11 m).
https://doi.org/10.1371/journal.pclm.0000665.s003
(PNG)
S4 Fig. Symbiodiniaceae community structure across all locations expressed in relative abundance.
https://doi.org/10.1371/journal.pclm.0000665.s004
(PNG)
S5 Fig. Similarity Profiling (SIMPROF) dendrogram representing the composition of Symbiodiniaceae by Palythoa tuberculosa at different Locations (Iōtorishima Control = Blue, Iōtorishima Seep = Red, Okinawa Mizugama = Green, Hawaiʻi Island Kona = Purple, Oʻahu North Shore = Pink).
https://doi.org/10.1371/journal.pclm.0000665.s005
(PNG)
Acknowledgments
We thank the captain and crew of the Yosemiya III, and cruise members Y. Ide (Oceanic Planning Corp.), H. Takamiyagi (OIST), and H. Kayanne (U. Tokyo) for their support and advice at Iōtorishima. We also thank Akito Shima for making the mycoplasma trees.
References
- 1. Hoegh-Guldberg O, Mumby PJ, Hooten AJ, Steneck RS, Greenfield P, Gomez E, et al. Coral reefs under rapid climate change and ocean acidification. Science. 2007;318(5857):1737–42. pmid:18079392
- 2. Havenhand JN, Buttler F-R, Thorndyke MC, Williamson JE. Near-future levels of ocean acidification reduce fertilization success in a sea urchin. Curr Biol. 2008;18(15):R651–2. pmid:18682203
- 3. Heuer RM, Grosell M. Physiological impacts of elevated carbon dioxide and ocean acidification on fish. Am J Physiol Regul Integr Comp Physiol. 2014;307(9):R1061-84. pmid:25163920
- 4. Rodolfo-Metalpa R, Houlbrèque F, Tambutté É, Boisson F, Baggini C, Patti FP, et al. Coral and mollusc resistance to ocean acidification adversely affected by warming. Nat Clim Change. 2011;1(6):308–12.
- 5. Rodolfo‐Metalpa R, Lombardi C, Cocito S, Hall‐Spencer JM, Gambi MC. Effects of ocean acidification and high temperatures on the bryozoan Myriapora truncata at natural CO2 vents. Marine Ecol. 2010;31(3):447–56.
- 6. Leung JYS, Zhang S, Connell SD. Is ocean acidification really a threat to marine calcifiers? A systematic review and meta-analysis of 980+ Studies spanning two decades. Small. 2022;18(35):e2107407. pmid:35934837
- 7. Boatta F, D’Alessandro W, Gagliano AL, Liotta M, Milazzo M, Rodolfo-Metalpa R, et al. Geochemical survey of Levante Bay, Vulcano Island (Italy), a natural laboratory for the study of ocean acidification. Mar Pollut Bull. 2013;73(2):485–94. pmid:23465567
- 8. Pichler T, Biscéré T, Kinch J, Zampighi M, Houlbrèque F, Rodolfo-Metalpa R. Suitability of the shallow water hydrothermal system at Ambitle Island (Papua New Guinea) to study the effect of high pCO2 on coral reefs. Mar Pollut Bull. 2019;138:148–58. pmid:30660256
- 9. Golbuu Y, Gouezo M, Kurihara H, Rehm L, Wolanski E. Long-term isolation and local adaptation in Palau’s Nikko Bay help corals thrive in acidic waters. Coral Reefs. 2016;35:909–18.
- 10. Leiva C, Torda G, Zhou C, Pan Y, Harris J, Xiang X, et al. Rapid evolution in action: environmental filtering supports coral adaptation to a hot, acidic, and deoxygenated extreme habitat. Glob Chang Biol. 2025;31(3):e70103. pmid:40028829
- 11.
Reaka-Kudla ML. The global biodiversity of coral reefs: a comparison with rain forests. Biodiversity II: Understanding and protecting our biological resources. 1997: 551.
- 12. Jindrich V. Structure and diagenesis of recent algal-foraminifer reefs, Fernando de Noronha, Brazil. J Sedimentary Res. 1983;53(2):449–59.
- 13. Inoue S, Kayanne H, Yamamoto S, Kurihara H. Spatial community shift from hard to soft corals in acidified water. Nat Clim Change. 2013;3(7):683–7.
- 14. Reimer JD, Agostini S, Golbuu Y, Harvey BP, Izumiyama M, Jamodiong EA. High abundances of zooxanthellate zoantharians (Palythoa and Zoanthus) at multiple natural analogues: potential model anthozoans? Coral Reefs. 2023;42(3):707–15.
- 15. Suggett DJ, Hall-Spencer JM, Rodolfo-Metalpa R, Boatman TG, Payton R, Tye Pettay D, et al. Sea anemones may thrive in a high CO2 world. Glob Chang Biol. 2012;18(10):3015–25. pmid:28741826
- 16. LaJeunesse TC, Parkinson JE, Gabrielson PW, Jeong HJ, Reimer JD, Voolstra CR, et al. Systematic revision of symbiodiniaceae highlights the antiquity and diversity of coral endosymbionts. Curr Biol. 2018;28(16):2570-2580.e6. pmid:30100341
- 17. Noonan SHC, Fabricius KE, Humphrey C. Symbiodinium community composition in scleractinian corals is not affected by life-long exposure to elevated carbon dioxide. PLoS One. 2013;8(5):e63985. pmid:23717522
- 18. Wee HB, Kurihara H, Reimer JD. Reduced Symbiodiniaceae diversity in Palythoa tuberculosa at a heavily acidified coral reef. Coral Reefs. 2019;38:311–9.
- 19. Kenkel CD, Moya A, Strahl J, Humphrey C, Bay LK. Functional genomic analysis of corals from natural CO2 -seeps reveals core molecular responses involved in acclimatization to ocean acidification. Glob Chang Biol. 2018;24(1):158–71. pmid:28727232
- 20. Maher RL, Schmeltzer ER, Meiling S, McMinds R, Ezzat L, Shantz AA. Coral microbiomes demonstrate flexibility and resilience through a reduction in community diversity following a thermal stress event. Front Ecol and Evol. 2020;8:555698.
- 21. McDevitt-Irwin JM, Garren M, McMinds R, Vega Thurber R, Baum JK. Variable interaction outcomes of local disturbance and El Niño-induced heat stress on coral microbiome alpha and beta diversity. Coral Reefs. 2019;38:331–45.
- 22. Webster NS, Negri AP, Botté ES, Laffy PW, Flores F, Noonan S, et al. Host-associated coral reef microbes respond to the cumulative pressures of ocean warming and ocean acidification. Sci Rep. 2016;6:19324. pmid:26758800
- 23. Reimer JD, Wee HB, López C, Beger M, Cruz ICS. Widespread Zoanthus and Palythoa dominance, barrens, and phase shifts in shallow water subtropical and tropical marine ecosystems. Oceanography Marine Biol. 2021;:533–57.
- 24. Reimer JD, Kurihara H, Ravasi T, Ide Y, Izumiyama M, Kayanne H. Unexpected high abundance of aragonite-forming Nanipora (Octocorallia: Helioporacea) at an acidified volcanic reef in southern Japan. Mar Biodivers. 2021;51(1):19.
- 25. Pax F. Studien an westindischen Actinien. Zool Jahrb Suppl. 1910;11:157–330.
- 26. Reimer JD. Key to field identification of shallow water brachycnemic zoanthids (Order Zoantharia: Suborder Brachycnemina) present in Okinawa. Galaxea J Coral Reef Studies. 2010;12(1):23–9.
- 27. Gaither M, Szabó Z, Crepeau M, Bird C, Toonen R. Preservation of corals in salt-saturated DMSO buffer is superior to ethanol for PCR experiments. Coral Reefs. 2011;30:329–33.
- 28. Hume BCC, Ziegler M, Poulain J, Pochon X, Romac S, Boissin E, et al. An improved primer set and amplification protocol with increased specificity and sensitivity targeting the Symbiodinium ITS2 region. PeerJ. 2018;6:e4816. pmid:29844969
- 29. Takahashi S, Tomita J, Nishioka K, Hisada T, Nishijima M. Development of a prokaryotic universal primer for simultaneous analysis of Bacteria and Archaea using next-generation sequencing. PLoS One. 2014;9(8):e105592. pmid:25144201
- 30. Hume BCC, Smith EG, Ziegler M, Warrington HJM, Burt JA, LaJeunesse TC, et al. SymPortal: A novel analytical framework and platform for coral algal symbiont next-generation sequencing ITS2 profiling. Mol Ecol Resour. 2019;19(4):1063–80. pmid:30740899
- 31. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol. 2009;75(23):7537–41. pmid:19801464
- 32. Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009;10:421. pmid:20003500
- 33. Eren AM, Morrison HG, Lescault PJ, Reveillaud J, Vineis JH, Sogin ML. Minimum entropy decomposition: unsupervised oligotyping for sensitive partitioning of high-throughput marker gene sequences. ISME J. 2015;9(4):968–79. pmid:25325381
- 34. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37(8):852–7. pmid:31341288
- 35. Hamamoto K, Mizuyama M, Nishijima M, Maeda A, Gibu K, Poliseno A, et al. Diversity, composition and potential roles of sedimentary microbial communities in different coastal substrates around subtropical Okinawa Island, Japan. Environ Microbiome. 2024;19(1):54. pmid:39080706
- 36. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17(1):10–2.
- 37. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3. pmid:27214047
- 38. Edgar RC, Flyvbjerg H. Error filtering, pair assembly and error correction for next-generation sequencing reads. Bioinformatics. 2015;31(21):3476–82. pmid:26139637
- 39. Bokulich NA, Kaehler BD, Rideout JR, Dillon M, Bolyen E, Knight R, et al. Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2’s q2-feature-classifier plugin. Microbiome. 2018;6(1):90. pmid:29773078
- 40. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41(Database issue):D590–6. pmid:23193283
- 41. Yilmaz P, Parfrey LW, Yarza P, Gerken J, Pruesse E, Quast C, et al. The SILVA and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids Res. 2014;42(Database issue):D643–8. pmid:24293649
- 42. Robeson MS 2nd, O’Rourke DR, Kaehler BD, Ziemski M, Dillon MR, Foster JT, et al. RESCRIPt: Reproducible sequence taxonomy reference database management. PLoS Comput Biol. 2021;17(11):e1009581. pmid:34748542
- 43. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O. Scikit-learn: machine learning in python. J Machine Learn Res. 2011;12:2825–30.
- 44. Vohsen SA, Gruber-Vodicka HR, Herrera S, Dubilier N, Fisher CR, Baums IB. Discovery of deep-sea coral symbionts from a novel clade of marine bacteria with severely reduced genomes. Nat Commun. 2024;15(1):9508. pmid:39496625
- 45.
R Core Team R. R: A language and environment for statistical computing. 2013.
- 46. Racine JS. RStudio: a platform-independent IDE for R and Sweave. JSTOR. 2012.
- 47. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4):e61217. pmid:23630581
- 48. Dixon P. Vegan, a package of R functions for community ecology. J Vegetation Sci. 2003;14(6):927–30.
- 49. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’hara R. Package ‘vegan’. Community Ecol Package. 2013:1–295.
- 50. Wickham H. ggplot2. Wiley Interdisciplinary Rev. 2011;3(2):180–5.
- 51. Wee HB, Kobayashi Y, Reimer JD. Effects of temperature, salinity, and depth on Symbiodiniaceae lineages hosted by Palythoa tuberculosa near a river mouth. Mar Ecol Prog Ser. 2021;667:43–60.
- 52. Reimer JD, Takishita K, Maruyama T. Molecular identification of symbiotic dinoflagellates (Symbiodinium spp.) from Palythoa spp. (Anthozoa: Hexacorallia) in Japan. Coral Reefs. 2006;25:521–7.
- 53. Mizuyama M, Masucci GD, Reimer JD. Speciation among sympatric lineages in the genus Palythoa (Cnidaria: Anthozoa: Zoantharia) revealed by morphological comparison, phylogenetic analyses and investigation of spawning period. PeerJ. 2018;6:e5132.
- 54. Santos ME, Reimer JD, Kiriukhin B, Wee HB, Mizuyama M, Kise H. Coral microbiomes from the Atlantic and Indo-Pacific oceans have the same alpha diversity but different composition. bioRxiv. 2024.
- 55. Howe-Kerr LI, Bachelot B, Wright RM, Kenkel CD, Bay LK, Correa AMS. Symbiont community diversity is more variable in corals that respond poorly to stress. Glob Chang Biol. 2020;26(4):2220–34. pmid:32048447
- 56. Tanvet C, Camp EF, Sutton J, Houlbrèque F, Thouzeau G, Rodolfo-Metalpa R. Corals adapted to extreme and fluctuating seawater pH increase calcification rates and have unique symbiont communities. Ecol Evol. 2023;13(5):e10099. pmid:37261315
- 57. Noda H, Parkinson JE, Yang S-Y, Reimer JD. A preliminary survey of zoantharian endosymbionts shows high genetic variation over small geographic scales on Okinawa-jima Island, Japan. PeerJ. 2017;5:e3740. pmid:29018596
- 58.
Pogoreutz C, Voolstra CR, Rädecker N, Weis V, Cardenas A, Raina JB. The coral holobiont highlights the dependence of cnidarian animal hosts on their associated microbes. Cellular dialogues in the holobiont. CRC Press; 2020: 91–118.
- 59. Loudon AH, Park J, Parfrey LW. Identifying the core microbiome of the sea star Pisaster ochraceus in the context of sea star wasting disease. FEMS Microbiol Ecol. 2023;99(3):fiad005. pmid:36690340
- 60. Fraune S, Zimmer M. Host-specificity of environmentally transmitted Mycoplasma-like isopod symbionts. Environ Microbiol. 2008;10(10):2497–504. pmid:18833647
- 61. Nguyen VK, King WL, Siboni N, Mahbub KR, Dove M, O’Connor W. The Sydney rock oyster microbiota is influenced by location, season and genetics. Aquaculture. 2020;527:735472.
- 62.
Steinberg RK, Dafforn KA, Johnston EL, Ainsworth T. Octocoral Symbiodiniaceae, but not bacterial communities, are resistant to compositional changes from heat stress on a subtropical reef. EcoEvoRxiv. 2023. https://doi.org/10.32942/X2X61F
- 63. van de Water JAJM, Voolstra CR, Rottier C, Cocito S, Peirano A, Allemand D, et al. Seasonal stability in the microbiomes of temperate gorgonians and the red coral Corallium rubrum across the Mediterranean Sea. Microb Ecol. 2018;75(1):274–88. pmid:28681143
- 64. Mohamed AR, Ochsenkühn MA, Kazlak AM, Moustafa A, Amin SA. The coral microbiome: towards an understanding of the molecular mechanisms of coral-microbiota interactions. FEMS Microbiol Rev. 2023;47(2):fuad005. pmid:36882224
- 65. Brumfield KD, Chen AJ, Gangwar M, Usmani M, Hasan NA, Jutla AS, et al. Environmental factors influencing occurrence of Vibrio parahaemolyticus and Vibrio vulnificus. Appl Environ Microbiol. 2023;89(6):e0030723. pmid:37222620
- 66. Rubio-Portillo E, Gago JF, Martínez-García M, Vezzulli L, Rosselló-Móra R, Antón J, et al. Vibrio communities in scleractinian corals differ according to health status and geographic location in the Mediterranean Sea. Syst Appl Microbiol. 2018;41(2):131–8. pmid:29338888
- 67. Ramos AA, Inoue Y, Ohde S. Metal contents in Porites corals: anthropogenic input of river run-off into a coral reef from an urbanized area, Okinawa. Mar Pollut Bull. 2004;48(3–4):281–94. pmid:14972580
- 68. DiBattista JD, Reimer JD, Stat M, Masucci GD, Biondi P, De Brauwer M, et al. Environmental DNA can act as a biodiversity barometer of anthropogenic pressures in coastal ecosystems. Sci Rep. 2020;10(1):8365. pmid:32433472
- 69. Lalas JAA, Jamodiong EA, Reimer JD. Spatial patterns of soft coral (Octocorallia) assemblages in the shallow coral reefs of Okinawa Island, Ryukyu Archipelago, Japan: dominance on highly disturbed reefs. Regional Stud Marine Sci. 2024;71:103405.
- 70. Godefroid M, Dupont S, Metian M, Hédouin L. Two decades of seawater acidification experiments on tropical scleractinian corals: Overview, meta-analysis and perspectives. Mar Pollut Bull. 2022;178:113552. pmid:35339865
- 71. Priest J, Ferreira CM, Munday PL, Roberts A, Rodolfo‐Metalpa R, Rummer JL. Out of shape: Ocean acidification simplifies coral reef architecture and reshuffles fish assemblages. J Animal Ecol. 2024.
- 72. Bell JJ, Davy SK, Jones T, Taylor MW, Webster NS. Could some coral reefs become sponge reefs as our climate changes? Global Change Biol. 2013;19(9):2613–24.
- 73. Tkachenko KS, Wu BJ, Fang LS, Fan TY. Dynamics of a coral reef community after mass mortality of branching Acropora corals and an outbreak of anemones. Marine Biol. 2007;151:185–94.
- 74. Agostini S, Harvey BP, Wada S, Kon K, Milazzo M, Inaba K, et al. Ocean acidification drives community shifts towards simplified non-calcified habitats in a subtropical-temperate transition zone. Sci Rep. 2018;8(1):11354. pmid:30054497
- 75. Cruz ICS, Loiola M, Albuquerque T, Reis R, Nunes J de ACC, Reimer JD, et al. Effect of phase shift from corals to Zoantharia on reef fish assemblages. PLoS One. 2015;10(1):e0116944. pmid:25629532
- 76. Otis N, Reimer JD, Kawamura I, Kise H, Mizuyama M, Obuchi M. Variation in species and functional composition of octocorals and zoantharians across a tropical to temperate environmental gradient in the Indo-Pacific. Coral Reefs. 2024:1–14.
- 77. Cannon SE, Donner SD, Liu A, González Espinosa PC, Baird AH, Baum JK, et al. Macroalgae exhibit diverse responses to human disturbances on coral reefs. Glob Chang Biol. 2023;29(12):3318–30. pmid:37020174
- 78. Hibino Y, Todd P, Ashworth CD, Obuchi M, Reimer JD. Monitoring colony colour and zooxanthellae (Symbiodinium spp.) condition in the reef zoanthid Palythoa tuberculosa in Okinawa, Japan. Marine Biol Res. 2013;9(8):794–801.
- 79. Santos MEA, Baker DM, Conti-Jerpe IE, Reimer JD. Populations of a widespread hexacoral have trophic plasticity and flexible syntrophic interactions across the Indo-Pacific Ocean. Coral Reefs. 2021;40:543–58.
- 80. Dudoit AA, Santos ME, Reimer JD, Toonen RJ. Phylogenomics of Palythoa (Hexacorallia: Zoantharia): probing species boundaries in a globally distributed genus. Coral Reefs. 2022;41(3):655–72.
- 81. Paulino GVB, Broetto L, Pylro VS, Landell MF. Compositional shifts in bacterial communities associated with the coral Palythoa caribaeorum due to anthropogenic effects. Mar Pollut Bull. 2017;114(2):1024–30. pmid:27889074