Specimens were collected by snorkeling using a chisel and a hammer from Ama Beach, Zamami Island, Ryukyu Archipelago, Japan (26°23'N; 127°29'E) at a depth of 1 m in July 2012 (Suppl. material 1). Colonies were attached to the bottoms (=downward facing sides) of carbonate stones. Digital images were also taken in situ to record the appearance of living colonies. Specimens were fixed in 99% ethanol immediately after collection.

Digital images were utilized to examine the color and shape of living colonies and polyps (Fig. 1). Skeletons were soaked in household bleach containing sodium hypochlorite for 15–20 min, followed by rinsing with distilled water and air-drying, to remove soft tissues. Skeletal specimens were stuck to aluminum specimen mounts by carbon double-faced tape, then examined and imaged with a scanning electron microscope (SEM) VE-8800 (Keyence, Osaka, Japan). CT images of ethanol-preserved specimens were taken with micro CT (in vivo micro X-ray CT system R_mCT2, Rigaku, Tokyo, Japan) and examined with the DICOM imaging application OsiriX v. 5.9 32-bit (Rosset et al. 2004) to investigate the structure of the skeleton in a non-invasive manner. By using this application, areas of particular CT-value in organisms can be visualized: e.g. showing skeleton (high CT-value) and hiding soft tissue (low CT-value), or vice versa. To check for the presence of sclerites, small pieces of EtOH preserved colonies were dissolved with household bleach on a well-slide and examined with an optical microscope. Additionally, several polyps were dissolved on carbon double-faced tape. Deposits from washing solution of whole colonies were retrieved and carefully rinsed, followed by mounting on carbon double-faced tape. These specimens were examined by SEM. X-ray diffraction analyses were performed with an X-ray diffractometer RINT Ultima/PC (Rigaku, Tokyo, Japan) at the Instrumental Research Center, University of the Ryukyus, to examine the crystal forms of calcium carbonate (aragonite/calcite) of the skeleton. Skeletal specimens were smashed to powder, mounted on well-washed glass slides and analyzed using Cu Kα radiation (40 kV, 30 mA), scanning between a 2θ angle ranging from 25° to 37° at 0.02° steps. The calcite/aragonite component (%) of samples was determined by comparing with data of standard samples of 100% calcite/aragonite. For comparison, a sample skeleton of the blue coral Heliopora coerulea was prepared and analyzed in the same manner.

Figure 1. 

Living colony of Nanipora kamurai attached to the bottoms (=downward facing side) of calcium-carbonate stone at Ama Beach, Zamami Island, Okinawa, Japan, 16 July 2012. Scale bar: approximately 5 mm.

DNA was extracted from tentacles and anthocodial tissue of ethanol-preserved samples by a guanidine extraction protocol following Sinniger et al. (2009). PCR amplifications were performed using HotStarTaq DNA polymerase (Qiagen, Tokyo, Japan) according to the manufacturer’s instructions. The mitochondrial mismatch repair protein (mtMutS) was amplified by semi-nested PCR following the procedure of McFadden et al. (2006), as no visible PCR product was yielded by a single PCR reaction. First, primers ND42599F (5’-GCCATTATGGTTAACTATTAC-3’) (France and Hoover 2002) and Mut-3458R (5’-TSGAGCAAAAGCCACTCC-3’) (Sánchez et al. 2003) were used to amplify the 5’ end of mtMutS, and then a second PCR reaction was run using internal forward primer ND42625F (5’-TACGTG GYACAATTGCTG-3’) (Lepard 2003) and reverse primer Mut-3458R. Both PCR reactions were performed under the following conditions: 15 min at 94 °C; 35 cycles of 1.5 min at 94 °C, 1.5 min at 58 °C, and 1 min at 72 °C; and a final extension of 5 min at 72 °C. Mitochondrial cytochrome oxidase subunit I (COI) was amplified by the primers COII-8068 (5’-CCATAACAGGACTAGCAGCATC-3’) (McFadden et al. 2004) and COI-OCTr (5’-ATCATAGCATAGACCATACC-3’) (France and Hoover 2002) under the following conditions: 5 min at 95 °C; 35 cycles of 1 min at 94 °C, 1 min at 40 °C, and 1.5 min at 72 °C; and then a final extension of 7 min at 72 °C. The nuclear ribosomal gene complex of the 3’ end of the 18S subunit, internal transcribed spacer 1 (ITS-1), 5.8S subunit, ITS-2, and the 5’ end of the 28S subunit was amplified by the primers 1s (5’-GGTACCCTTTGTACACACCGCCCGTCGCT-3’) and 2ss (5’-GCTTTGGGCTGCAGTCCCAAGCAACCCGACTC-3’) (Chen et al. 1996) under the following conditions: 15 min at 95 °C; 35 cycles of 0.5 min at 94 °C, 1 min at 52 °C, and 1.5 min at 72 °C; and a final extension of 5 min at 72 °C. Amplified products were visualized with 1.0% agarose gel electrophoresis. Positive PCR products were cleaned up by Exonuclease I and Shrimp Alkaline Phosphatase (Takara, Shiga, Japan) before sequencing.

Sequencing was performed by Fasmac (Kanagawa, Japan). Cycle sequencing was performed in both directions using the forward and reverse primers separately with BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) under reaction conditions according to the manufacturer’s instructions. Reaction products were analyzed on an ABI PRISM 3700 DNA Analyzer (Applied Biosystems). The sequences were analyzed by 4Peaks Version 1.7.2 software (mekentosj.com, Amsterdam, Netherlands).

By using Se-AL v2.0a11 software (Rambaut 2002), the nucleotide sequences of mtMutS, COI, and ITS1-5.8s-ITS2-28S from specimens obtained in the present study were separately aligned with sequences of Heliopora and other octocoral species retrieved from GenBank (Suppl. material). The alignments were checked by eye and manually edited to remove any ambiguous sites (e.g. double peaks) before phylogenetic analyses. For each alignment, none or only one to two base pairs were edited in this manner. Consequently, three aligned data sets were generated: 1) 861 sites of 33 sequences (mtMutS), 2) 735 sites of 12 sequences (COI), and 3) 697 sites of 5 sequences (ITS1-5.8s-ITS2-28S). The alignment data are available on request from the corresponding author. Additional octocoral sequences retrieved from GenBank are shown in Supplementary Table 2.

Maximum-likelihood (ML) analyses with PhyML Online (Guindon and Gascuel 2003) of these datasets were independently performed using input trees generated by BIONJ (Gascuel 1997) with the general time reversible (GTR) model. PhyML bootstrap trees (1000 replicates) were constructed using the same parameters as the individual ML trees. Genetic distances were calculated using Kimura’s two-parameter model (Kimura 1980).

Bayesian trees were reconstructed by running the program MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003) within the program Geneious version 8.0.4 (Restricted) created by Biomatters (available from http://www.geneious.com/). One cold and three heated Markov chain Monte Carlo (MCMC) chains with default-chain temperatures were run for 1,000,000 generations, sampling log-likelihoods (InLs), and trees at 1000-generation intervals (1,000 InLs and trees were saved during MCMC). The first 100,000 generations of all runs were discarded as “burn-in” for the dataset and remaining 900 trees were used to obtain posterior probabilities and branch-length estimates, respectively. Neighbor-joining (NJ) trees were also reconstructed by using CLC Free Workbench 4 software (CLCbio.com, Aarhus North, Denmark) (1000 replicates).

Results of phylogenetic analyses apparently show that Nanipora kamurai is more closely related to Heliopora coerulea than to other groups of octocorals. The genetic distance between Nanipora and Heliopora as shown in branch lengths in the mtMutS tree (Fig. 12) is comparable to what is typically seen among different families of octocorals (e.g. McFadden et al. 2006) (see also Suppl. materials 3, 4). This is the first time the existence of an extant relative of H. coerulea has been confirmed using molecular phylogenetic analyses, as no phylogenetic research on Epiphaxum spp. has yet been conducted.

Lonsdale (1850), who described the fossil genus Epiphaxum from Chalk (Upper Cretaceous) in Sussex, United Kingdom, felt that Epiphaxum was not related to any extant family. According to Lozouet and Molodtsova (2008), Epiphaxum was later placed by Voigt (1958) into Clavulariidae, which had been defined by the presence of cylindrical anthosteles and connecting stolons. Finally, Bayer and Muzik (1977) placed this genus in the order Helioporacea by examining the type of calcium-carbonate present in skeletons.

The general colony shape of N. kamurai closely resembles encrusting and stoloniferous species of Epiphaxum (Bayer, 1992). As well, the basic structure of the skeleton (simple and tubular calyx, indentations on the top of the calyx) is common among Nanipora and Epiphaxum spp. However, longitudinal grooves on the surface of calyces’ skeleton, common to every Epiphaxum spp. (see Bayer and Muzik 1977; Bayer 1992; Lozouet and Molodtsova 2008), were not observed in N. kamurai. Instead, the entire surface of the Nanipora skeleton is covered by a reticulate pattern (Fig. 4). The various descriptions of genus Epiphaxum mention possession of calcite calcium-carbonate sclerites (Bayer and Muzik 1977; Bayer 1992; Lozouet and Molodtsova 2008), although actually in some species of Epiphaxum the presence of sclerites has not been confirmed due to the absence of soft-tissue in the holotype specimen (E. septifer Bayer, 1992), or due to the type specimen being fossilized (E. arbuscula Bayer, 1992) (Bayer 1992; Lozouet and Molodtsova 2008). The stoloniferous colony is one of the simplest colony morphologies in Octocorallia and is found among many unrelated groups of octocorals, but many other morphological characters apparently suggest the close relatedness of N. kamurai and Epiphaxum. In particular, considering the complete lack of sclerites, placement of N. kamurai in a new genus inside Lithotelestidae is much more appropriate than a major modification of genus Epiphaxum and inclusion of N. kamurai within Epiphaxum. Molecular phylogenetic data for Epiphaxum spp. are necessary to construct the complete phylogeny of Helioporacea.

The blue coral Heriopora coerulea, the sole member of Helioporidae, has been considered to be extraordinarily distinct among octocorals. When Bayer (1981) reorganized the entire subclass Octocorallia into only three orders, one order was Helioporacea. However, McFadden et al. (2006) showed that H. coerulea sequences were placed within a large Calcaxonia–Pennatulacea clade. In this study, similar to McFadden et al. (2006), H. coerulea and Nanipora kamurai sequences fell into the same sub-major clade of Octocorallia. However, although no recent molecular study supports the phylogenetic distinctiveness of Helioporacea as an order within subclass Octocorallia, the order Helioporacea is still used. The taxonomic status of this order needs to be re-examined with careful morphological and molecular phylogenetic comparisons. The discovery and confirmation of the phylogenetic relationship of N. kamurai to Helioporacea in this study should contribute to this reassessment.

Extant Heliopora coerulea is restricted to the Indo-Pacific, and extant species of Epiphaxum are found from the Caribbean and the western Indian Ocean (Madagascar), while fossil species of both genera have been found sporadically but widely from Europe. So far, Nanipora has only been found in Zamami Island, Okinawa, Japan, although surveys are needed to confirm its exact distribution. As indicated in Lozouet and Molodtsova (2008), recent H. coerulea is considered to be a relict of fossil Heliopora species distributed throughout the Tethys Ocean. Considering the morphological characters of these genera, we can hypothesize that Helioporidae branched first from a common ancestor, subsequently followed by the division of Epiphaxum and Nanipora, although the geographical timing of this split is unknown. As pointed out by Lozouet and Molodtsova (2008), the discontinuous distribution of Epiphaxum across the Pacific and the Atlantic implies that at least Epiphaxum already existed before the closure of the Tethyan connection. Additionally, the morphological distinctiveness of Nanipora strongly supports the hypothesis that Nanipora and Epiphaxum radiated before the disjunction of the Atlantic and the Pacific, rather than Nanipora radiating from Indo-West Pacific Epiphaxum. Molecular phylogenetic analyses of Atlantic and Pacific Epiphaxum and corroborated studies with paleontology should help confirm the evolutionary history of this unique group.

Detailed habitat information of extant species of Epiphaxum is unknown as all known living specimens were obtained by dredging or trawling (Bayer and Muzik 1977; Bayer 1992). Lozouet and Molodtsova (2008) indicted fossil species of Epiphaxum (E. arbuscula) were strongly related to submarine canyons, which also included fossil fauna found in muddy sea-floor at depths corresponding to the outer continental shelf or upper slope, submarine cave environments, and rocky circa-littoral cliffs. Considering this information together with the cryptic habitat of Nanipora kamurai (bottom side of carbonate stones), Epiphaxum and Nanipora appear to not reside in shallow well-lit subtropical and tropical environments like Heliopora, but instead in shaded or cryptic environments. Detailed surveys of such cryptic environments may lead to the discovery of other unknown Epiphaxum or Nanipora species.

It is astonishing that a unique, relict species such as Nanipora kamurai was found from shallow waters. Generally, relict species are thought to be most commonly found in stable and undisturbed environments, such as abyssal waters, as demonstrated by the discovery of Coelacanthiformes in Africa and Indonesia (Smith 1939; Erdmann et al. 1998). Although not from abyssal depths, extant Epiphaxum spp. were also found from quite deep habitats (E. micropora: 50–400 m, E. breve: 183 m, E. septifer: 200–360 m). The discovery of N. kamurai in this study demonstrates the importance of the study of cryptic anthozoan fauna in shallow coral reef areas (see also Fujii and Reimer 2011), not only for a more correct understanding of coral reef biodiversity, but also to make progress in clarifying the phylogeny and taxonomy of octocorals.