Introduction

Some cnidarians are known to live on the shells of live gastropods. These symbiotic associations between cnidarians and shelled gastropods have been reported in several studies such as between the actiniarian Allantactis parasitica Danielssen (1890), and Buccinum undatum Linnaeus, 1758 (Mercier & Hamel, 2008), and between the hydrozoan Cytaeis capitata (Puce et al., 2004), and Nassarius globosus (Quoy & Gaimard, 1833) (Puce et al., 2004).

However, there have been few detailed studies of the associations between cnidarians and live gastropod mollusks with an external shell, particularly in comparison to cnidarian association with gastropod shells occupied by hermit crabs (e.g., Brooks & Gwaltney, 1993; Gusmão et al., 2020; Williams & McDermott, 2004). Several hypotheses have been suggested to explain symbioses between cnidarians and live gastropod mollusks. For instance, the symbiosis between A. parasitica and B. undatum provides advantages to both partners, including mobility, substrate, and food acquisition to A. parasitica, while B. undatum may receive protection from A. parasitica through camouflage as well as from the cnidarian’s nematocysts (Mercier & Hamel, 2008). However, overall, respective benefits for gastropod mollusks and cnidarians have generally not been well examined (Puce et al., 2008), and there are many understudied cnidarian taxa associated with gastropod mollusks. Therefore, basic fundamental studies in terms of taxonomy and ecology on such understudied cnidarians are needed.

One such understudied taxon is the order Zoantharia Rafinesque, 1815. Symbiotic associations between several species within the zoantharian family Epizoanthidae Delage & Hérouard, 1901, and gastropods have been documented (e.g., Haddon & Duerden, 1896; Lwowsky, 1913; Carlgren, 1924; Reimer et al., 2010; Kise & Reimer, 2019) at mesophotic and deeper depths (Reimer et al., 2019). Within Epizoanthidae, some species of the genera Epizoanthus Gray, 1867, and Paleozoanthus Carlgren, 1924, are known to be associated with shelled gastropods, such as E. egeriae Haddon & Duerden, 1896, associated with Murex spp., E. indicus (Lwowsky, 1913) associated with Borsonia symbiotes (Wood-Mason & Alcock, 1891) from the Indo-West Pacific (see Kise & Reimer, 2019), E. mediterraneus Carlgren, 1935, associated with Murex spp. from the Mediterranean Sea, and Paleozoanthus reticulatus Carlgren, 1924, associated with Granulifusus rubrolineatus (Sowerby II, 1870) from off South Africa. Although many Epizoanthus species associated with gastropods have been reported on, the monotypic species Paleozoanthus reticulatus has never subsequently been reported since its original description. In summary, our knowledge of Epizoanthidae-mollusk associations is still fragmentary (Reimer et al., 2010).

In this study, to help address this knowledge gap, we examined Epizoanthidae specimens associated with gastropods collected from the northwest Pacific Ocean, as well as a single specimen in the collection of the KwaZulu-Natal Museum in South Africa. Our morphological observations combined with molecular phylogenetic analyses led us to formally describe some Pacific Ocean specimens as a new zoantharian species, Epizoanthus protoporos sp. nov. In addition, we examine the symbiosis between E. protoporos sp. nov. and its host G. niponicus. Furthermore, we report on the existence of P. reticulatus for the first time since its original description, and add information on this species’ morphological characteristics.

Materials and methods

Specimen collection

Regarding specimens from Japan, three Epizoanthus specimens associated with Granulifusus niponicus were collected at depths of 250 to 300 m via trawl net on the fishing trawler Jinsho-maru from the Sea of Kumano, Mie, Japan (33°54′44.6″N–33°56′0.03.7″N, 136°17′47.8″E–136°19′37.6″E) by Moritaki on December 26, 2016, and March 17, 2019. After observation of living specimens, specimens were initially fixed in 5–10% seawater formalin and were then later preserved in 70% ethanol for morphological observations. Subsamples were preserved in 99.5% ethanol for molecular analyses. Newly collected specimens were deposited in the National Science Museum, Tsukuba, Ibaraki, Japan.

Furthermore, we examined a single specimen associated with Granulifusus rubrolineatus in the Mollusca collections at KwaZulu-Natal Museum (NMSA-P1196), collected from off South Africa.

Morphological observation

External morphological characters of the preserved specimen were examined using in situ images and dissecting microscope. Internal morphological characters were examined by histological sections; 10–15 mm thickness of serial section were made with microtome LEICA RM2145 (Leica, Germany) and stained with hematoxylin and eosin after decalcification with Morse solution for 48 h (1:1 vol; 20% citric acid: 50% formic acid) and desilication with 20% hydrofluoric acid for 18–24 h. Classification of marginal muscle shapes followed Swain et al. (2015). Cnidae analyses were conducted using undischarged nematocysts and spirocysts from tentacles, column, actinopharynx, and mesenterial filaments using a Nikon Eclipse80i stereomicroscope (Nikon, Tokyo). Cnidae sizes were measured using ImageJ v1.45 s (Rasband, 2012). The reported frequencies are the relative amounts based on numbers from all slides in the cnidae analyses. Cnidae classification generally followed England (1991) and Ryland and Lancaster (2004) exception for the treatment of basitrichs and microbasic b-mastigophores as mentioned in Kise et al. (2019).

DNA extraction, PCR amplification, and sequencing

Total DNA was extracted from tissue by using a spin-column DNeasy Blood and Tissue Extraction kit following the manufacturer’s instructions (Qiagen, Hilden, Germany). PCR amplification using Hot Star Taq Plus Master Mix Kit (Qiagen, Hilden, Germany) and TaKaRa Ex Taq™ (Takara Bio Inc., Japan) was performed for each of COI (mitochondrial cytochrome oxidase subunit I), mt 12S-rDNA (mitochondrial 12S ribosomal DNA), mt 16S-rDNA (mitochondrial 16S ribosomal DNA), 18S-rDNA (nuclear 18S ribosomal DNA), and ITS-rDNA (nuclear internal transcribed spacer region of ribosomal DNA). COI was amplified using the universal primer set LCO1490 and HCO2198 (Folmer et al., 1994) following the protocol by Montenegro et al. (2015). mt 12S-rDNA was amplified using the primer sets ANTMTf and ANTMTr (Chen et al., 2002), and 12S1a and 12S3r (Sinniger et al., 2005), following protocols by Chen et al. (2002) and Sinniger et al. (2005), respectively. mt 16S-rDNA was amplified using the primer set (Sinniger et al., 2010) and 16SbmoH (Sinniger et al., 2005), following the protocol by Sinniger et al. (2005). 18S-rDNA was amplified using the primer set 18SA and 18SB (Medlin et al., 1988) and sequenced using 18SL, 18SC, 18SY, and 18SO (Apakupakul et al., 1999), following the protocol by Swain (2010). ITS-rDNA was amplified using the primer set ITSf and ITSr (Swain, 2009), following the protocol by Swain (2010). All PCR products were purified with shrimp alkaline phosphatase (SAP) and Exonuclease I (Takara Bio Inc., Shiga, Japan) at 37 °C for 40 min followed by 80 °C for 20 min. Cleaned PCR products were sequenced in both directions by Fasmac (Kanagawa, Japan) and Macrogen (Tokyo, Japan). Obtained sequences in this study were deposited in GenBank under accession numbers ON007050 – ON074339 (Table S1).

Molecular phylogenetic analyses

Sequences were initially aligned in Geneious v10.2.3 (Kearse et al., 2012). Thereafter, the sequences were manually trimmed and realigned using MAFFT (Katoh & Standley, 2013) with the auto algorithm under default parameters for all genetic markers. A minimum of data from three markers was established as the threshold to include or exclude OTUs from the final combined dataset (Table S1). All aligned datasets are available from Figshare (https://doi.org/10.6084/m9.figshare.19545556). Phylogenetic analyses were performed on the concatenated dataset using Maximum likelihood (ML) and Bayesian inference (BI). ModelTest-NG v0.1.6 (Darriba et al., 2019) under the Akaike information criterion was used to select the best fitting model for each molecular marker independently for both ML and BI analyses. The best selected models for ML and BI analyses were TPM2uf + I + G (GTR + G) for COI, TPM3uf + G (HKY + G for BI) for mt 12S-rDNA, TIM3 + G (GTR + G) for mt 16S-rDNA, HKY + I for 18S-rDNA, TVM + G (GTR + G for BI) for ITS-rDNA, and TIM1 + I + G (GTR + I + G for BI) for 28S-rDNA. Independent phylogenetic analyses were performed using model partitioning per each region in RAxML-NG v0.9.0 (Kozlov et al., 2019) for ML, and MrBayes v3.2.6 (Ronquist & Huelsenbeck, 2003) for BI. RAxML-NG was configured to use 12,345 initial seeds, search for the best tree among 100 preliminary parsimony trees, branch length was scaled and automatically optimized per partition, and model parameters were also optimized. MrBayes was configured following the models and parameters as indicated by ModelTest-NG, 4 MCMC heated chains were run for 5,000,000 generations with a temperature for the heated chain of 0.2. Chains were sampled every 200 generations. Burn-in was set to 1,250,000 generations at which point the average standard deviation of split frequency (ASDOSF) was steadily below 0.01. Parazoanthidae genera sequences were used as the outgroup in ML and BI analyses (Table S1).

Results

Taxonomic account.

Order Zoantharia Rafinesque, 1815

Suborder Macrocnemina Haddon & Shackleton, 1891

Family Epizoanthidae Delage & Hérouard, 1901.

Genus Paleozoanthus Carlgren, 1924

Type species. Paleozoanthus reticulatus Carlgren, 1924, by original designation and monotypy.

Paleozoanthus reticulatus — Carlgren (1938: 103), fig. 53.

Diagnosis. Macrocnemic zoantharians with a simple mesogleal muscle, and both complete and incomplete mesenteries fertile.

Paleozoanthus reticulatus Carlgren, 1924

Figure 1a–c

Fig. 1
figure 1

External morphology of Palezoanthus reticulatus and Epizoanthus protoporos sp. nov. a, b Overall view of the preserved specimen of P. reticulatus (NMSA-P1196). c Original drawing of P. reticulatus (Carlgren, 1924). d, e Overall view of the preserved specimen of Epizoanthus protoporos sp. nov. (holotype: NSMT-Co1797). f A living colony on Granulifusus niponicus in aquarium. Scale bars: 1 cm (a, b, d–f)

Full size image

Material examined. NMSA-P1196, collected from off the Mgazi River mouth, South Africa (43°0′09.0″S, 29°30′05.0″E) at a depth of 100 m, collected by National Research Institute for Oceanology of the Council for Scientific and Industrial Research, 15 July 1982, fixed in 75% ethanol, deposited in the KwaZulu-Natal Museum, South Africa.

Description. External morphology. Colonial macrocnemic zoantharian associated with gastropod Granulifusus rubrolineatus. The examined specimen consists of six truncated cone-like shaped polyps, 1.7–3.1 mm in height and 3.5–5.5 mm in diameter when contracted (Fig. 1a, b). The six polyps connected by thin mesh-like coenenchyme and regularly arranged on the shell margin. Coenenchyme completely covering the gastropod shell (Fig. 1a). No polyps attached on the aperture of the G. rubrolineatus shell (Fig. 1b). Surface of column rough, and ectoderm and mesoglea of scapus and coenenchyme encrusted with numerous sand and silica particles. Tentacles in two rows, number not available, but estimated as 20–24 based on numbers of capitulary ridges. Capitulary ridges present, 10–12 in number, visible in contracted polyps.

Internal morphology. We could not obtain cross-sections or images to observe internal morphology due to the poor condition of the specimen.

Cnidae. Basitrichs and microbasic b-mastigophores, holotrichs, and spirocysts (Fig. 2, Table 1).

Fig. 2
figure 2

Cnidae in the tentacles, column, actinopharynx, and mesenterial filaments of Palezoanthus reticulatus and Epizoanthus protoporos sp. nov. (holotype: NSMT-Co 1797). Abbreviations: HL: holotrich large, HS: holotrich small, O: basitrichs and microbasic b-mastigophores, PM: microbasic p-mastigophores, S: spriocysts

Full size image
Table 1 Cnidae types and sizes observed in Paleozoanthus reticulatus and Epizoanthus protoporos sp. nov. Frequency: relative abundance of cnidae type in decreasing order; numerous, common, occasional, rare (n = number of cnidae)
Full size table

Distribution. The type locality of Paleozoanthus reticulatus is the Agulhas Bank, South Africa (35°16′00.0″S, 22°26′07.0″E). With the current specimen, we confirmed the presence of this species in the Eastern Cape, South Africa, as the examined specimen was collected from off the Mgazi River mouth, South Africa.

Associated host. Paleozoanthus reticulatus is associated with Granulifusus rubrolineatus.

Remarks. Paleozoanthus reticulatus was described in brief by Carlgren (1924) based on a single specimen collected from the Agulhas Bank, off South Africa (Fig. 1c). The original description is not very complete, although it includes internal and external morphology, as Carlgren (1924) examined a single specimen that was poorly preserved. In general, gametes of zoantharians develop only on complete mesenteries (Ryland, 1997), while P. reticulatus is unique in having gametes develop in both complete and incomplete mesenteries (Carlgren, 1924, 1938). The present study could not confirm the presence of fertilized incomplete mesenteries due to the poor condition of the examined specimen, although in other regards the examined specimen corresponds well to the description of Carlgren (1924). Although the capitulary ridges of examined specimens are 10–12, the original description reported 24 capitulary ridges with 24 tentacles and mesenteries. The numbers of the capitulary ridges are usually half the tentacles and mesenteries (Swain et al., 2016), and therefore, the number of the capitulary ridges determined in this study is likely correct. Although P. reticulatus had been treated as a taxon inquirendum due to a lack of information (Reimer & Sinniger, 2020), this study reports on the existence of P. reticulatus for the first time since its original description, based on our identification made with the associated gastropod species and the zoantharian’s external morphological characters.

Genus Epizoanthus Gray, 1867

Type species. Dysidea papillosa (Johnston, 1842), by monotypy (see also Opinion, 1689; ICZN, 1992).

Duseideia papillosa — Johnston (1842: 190–191), fig. 18, Mammillifera incrustata — Düben & Koren (1847: 268), Sidisia barleei — Gray (1858: 557–560), pl 5, fig. 8, Zoanthus couchii — Landsborough (1852: 225), Zoanthus incrustatus — Sars (1860), 141, Epizoanthus americanus — Verrill (1864: 34, 45), Epizoanthus incrustatus — Haddon & Shackleton (1898: 636–616), pl 58, fig. 1- 22, pl 59, fig. 2, pl 60, 1, Epizoanthus papillosum — Cutress & Pequegnat (1960: 98).

Diagnosis. Macrocnemic zoantharians with simple mesogleal muscle, readily distinguishable from Palaeozoanthus by the presence of non-fertile micromesenteries (Sinniger & Häussermann, 2009).

Epizoanthus protoporos sp. nov.

Figure 1d–f, Fig. 4a–f

Material examined. Holotype: Sea of Kumano, Mie, Japan (33°56′03.7″N, 136°19′37.6″E), 300 m depth, December 26, 2016, NSMT-Co 1797. Paratypes: Sea of Kumano, Mie, Japan (33°54′44.6″N, 136°17′47.8″E), 250 m depth, March 17, 2019, NSMT-Co 1798; Sea of Kumano, Mie, Japan (33°54′44.6″N, 136°17′47.8″E), 250 m depth, March 17, 2019, NSMT-Co 1799.

Description. External morphology. Colonial macrocnemic zoantharian associated with gastropod Granulifusus niponicus (Smith, 1879). The holotype is a colony consisting of seven polyps on a G. niponicus shell (Fig. 1d–f). Polyps of living holotype truncated and cone-like in shape, and 2.0–3.9 mm in height, 9.1–14.5 mm in diameter when expanded (Fig. 1f). Polyps of preserved holotype dark beige in coloration, 2.7–4.7 mm in height from coenenchyme, 5.3–9.1 mm in diameter when contracted. Polyps partially connected by thin coenenchyme and regularly arranged on the shell margin (Fig. 1d, e). No polyps attached on the aperture of the G. niponicus shell (Fig. 1e). Ectoderm and mesoglea of scapus and coenenchyme heavily encrusted with numerous sand and silica particles, while ectoderm and mesoglea of capitulum encrusted with a small amount of sand and silica particles. Tentacles in two rows, 28–32 in number, light beige and/or pale red in coloration. Tips of tentacles usually cream in coloration. Tentacles thick and longer than expanded oral disk diameter, 3.5–14.1 mm in length and 1.0–2.6 in diameter. Capitulary ridges present and strongly pronounced when contracted, 14–16 in number. Oral disk small, 5.0–6.4 mm in diameter, light beige and/or pale red in coloration, oval protrusion has a slit-like mouth when expanded. There was no noteworthy variation between holotype and paratypes.

Internal morphology. Zooxanthellae absent. Mesenteries approximately 28–32, in macrocnemic arrangement (fifth mesentery complete). Mesoglea thickness 250–300 μm at the actinopharynx region. Encircling sinus consisting of oval and flattened lacunae present. Large lacunae in mesoglea and ectoderm resulting from dissolution of encrustations consist of sand and silica particles by hydrofluoric acid. Mesoglea thicker than ectoderm. Reticulate mesogleal muscle. Reticulate marginal muscle bends at a right angle (Fig. 3a, b). Complete mesenteries fertile (Fig. 3c).

Fig. 3
figure 3

Internal morphology of Epizoanthus protoporos sp. nov. (holotype: NSMT-Co 1797). a Longitudinal section of polyp. b Closed-up image of reticulate marginal muscle. c Cross-section of polyp at level of mesenterial filaments. Abbreviations: CM: complete mesentery, IM: incomplete mesentery, O: oral disk, RMM: reticulate marginal muscle, T: tentacle, TT: testis. Scale bars: 3 mm (a), 200 μm (b), 500 μm (c)

Full size image

Cnidae. Basitrichs and microbasic b-mastigophores, microbasic p-mastigophores, and spirocysts (Fig. 2, Table 1).

Distribution. Epizoanthus protoporos sp. nov. has been only found in Japanese waters around the Sea of Kumano, Mie, at depths of 250–300 m.

Notes. Although Granulifusus niponicus was not often active in the aquarium, we observed some behaviors. Occasionally, the front end of its foot stretched and stroked the polyps of Epizoanthus protoporos sp. nov. (Fig. 4). At this time, the tip of the foot was bifurcated so as to pinch the polyps of Epizoanthus protoporos sp. nov., and the foot twisted strongly against the distal polyps to rotate the shell and bring the tip of the foot closer to the polyps.

Fig. 4
figure 4

Images of Granulifusus niponicus acting on Epizoanthus protoporos sp. nov. a–c The front, side, backside of image that the front end of G. niponicus’s foot stretched and stroked the polyps of Epizoanthus protoporos sp. nov. d–f Closed-up image of the front end of G. niponicus’s foot acting on Epizoanthus protoporos sp. nov

Full size image

Molecular phylogeny. The results of the phylogenetic analyses using the concatenated dataset are shown in Fig. 5. Sequences of Antipathozoanthus remengesaui, B. puetoricense, Paraozoanthus darwini, Savalia savaglia, and Umimayanthus chanpuru were used as outgroup. Epizoanthus formed a monophyletic clade with complete support (ML = 100%, BI = 1.00). Within the Epizoanthus clade, Epizoanthus protoporos sp. nov. was sister to E. rinbou Kise and Reimer, 2019 with moderate support (ML = 59%, BI = 0.99). Another gastropod-associated species, Epizoanthus sp. S02, was located within another clade consisting of Epizoanthus ramosus, and is known to have an association with hermit crabs within the families Diogenidae and Paguridae (Ates, 2003). Thus, this study recovered at least two independent origins for symbioses between zoantharians and gastropods.

Fig. 5
figure 5

Maximum likelihood tree based on combined dataset of COI, mt 12S-rDNA, mt 16S-rDNA, 18S-rDNA, ITS-rDNA, and 28S-rDNA sequences. Number at nodes represent ML bootstrap values (> 50% are shown). White circles on nodes indicate high support of Bayesian posterior probabilities (> 0.95). Red squares indicate species associated with gastropod mollusks

Full size image

Remarks. Including the current study, there are two zoantharian species associated with Granulifusus gastropods; Paleozoanthus reticulatus and Epizoanthus protoporos sp. nov. However, there are distinct morphological differences separating these species. The polyp size of Epizoanthus protoporos sp. nov. is larger than that of P. reticulatus (2.7–4.7 mm in height and 5.3–9.1 mm in diameter vs 1.7–3.1 mm in height and 3.5–5.5 mm in diameter). Additionally, Epizoanthus protoporos sp. nov. has 28–32 tentacles, while P. reticulatus has 20–24 tentacles. Epizoanthus protoporos sp. nov. is also distinguished from other Indo-Pacific gastropod–associated Epizoanthus species, E. thalamophilus Hertwig, 1888, E. egeriae, E. indicus, and E. rinbou, by combinations of morphological characteristics and molecular differences (Table 2). Epizoanthus protoporos sp. nov. can be easily distinguished from both E. thalamophilus and E. indicus as the coenenchyme of the two latter species are continuous and cover associated gastropod shells completely, while the coenenchyme of Epizoanthus protoporos sp. nov. is not continuous and only partially covers associated Granulifusus gastropod shells. Additionally, E. indicus is generally larger than Epizoanthus protoporos sp. nov. in polyp size (E. indicus polyps: 4.0 mm in height and 10.0 mm in diameter). Additionally, associated gastropod species are different between E. thalamophilus and Epizoanthus protoporos sp. nov.; the former species is associated with Borsonia symbiotes, while the latter is associated with Granulifusus niponicus. E. rinbou, and Epizoanthus sp. S sensu Reimer et al. (2010) are smaller than Epizoanthus protoporos sp. nov. in polyp size. Moreover, holotrichs were not found in any tissues of Epizonthus protoporos sp. nov., while holotrichs are present in the tissues of E. rinbou. Finally, the molecular phylogeny clearly showed that Epizoanthus protoporos sp. nov., E. rinbou, and Epizoanthus sp. S are placed in different groups (Fig. 5).

Table 2 Summary of morphological data from Epizoanthidae species associated with gastropods
Full size table

Etymology. The specific name is derived from the Greek word protopóros meaning “explorer”, as this species colonizes on rocket-like gastropod shells to gain mobility.

Japanese common name. Naganishi-yadori-sunaginchaku.

Discussion

In hexacorallian groups, actiniarian-gastropod associations have been the subject of some research attention, and results have demonstrated that such associations provide benefits to hosts and associates as both partners receive protection and can increase their survival rates (Pastorino, 1993; Riemann-Zürneck, 1994; Mercier & Hamel, 2008). On the other hand, research on zoantharian-gastropod associations has lagged considerably compared to other cnidarian-gastropod associations due largely to the difficulty of specimen collection, as zoantharian-gastropod associations are exclusively found in mesophotic to deeper waters. Among the few existing zoantharian-gastropod studies, a recent taxonomic study suggested that Epizoanthus rinbou has an obligate association with the gastropod Guildfordia triumphans based on molecular and morphological datasets (Kise & Reimer, 2019). In addition, in this study, we theorize that an obligate association between Epizoanthus protoporos sp. nov. and Granulifusus niponicus exists, as Epizoanthus protoporos sp. nov. was not found in situ on other invertebrates or on rocks, and was always found exclusively on shells of living G. niponicus.

We observed a unique behavior of G. niponicus towards Epizoanthus protoporos sp. nov. in aquaria during this study. Polyps of Epizoanthus protoporos sp. nov. are regularly arranged on the shell margin around the aperture but are not found on the other side of the aperture, and thus, we theorize that this polyp arrangement may depend on the distance that the foot of G. niponicus can reach, with polyps only present out of range of the foot. Epizoanthus protoporos sp. nov. may receive several advantages from G. niponicus such as mobility, substrate, and relatively easy food acquisition. Furthermore, similar advantages may be considered in the association between Paleozoanthus reticulatus and Granulifusus rubrolineatus, as polyp arrangement of P. reticulatus is identical with that of Epizoanthus protoporos sp. nov. However, the advantages to G. niponicus from such a symbiosis are still not clear as Epizoanthus protoporos sp. nov. lacks holotrichs in all tissues, while other gastropod-associated species such as E. rinbou have holotrichs in their tissues (Kise & Reimer, 2019), particularly since holotrichs have been characterized as used in aggression (Rotjan & Dimond, 2010). Therefore, potential protection benefits to G. niponicus via the nematocysts of Epizoanthus protoporos sp. nov. are uncertain. To better understand this symbiotic association, experiments on the influence of predators under controlled laboratory settings are necessary.

Epizoanthus protoporos sp. nov. and P. reticulatus both associate with hosts belong to genus Granulifusus, suggesting that Paleozoanthus and Epizoanthus may be congeneric. This is further supported by the fact that their external morphology resembles each other. However, the presence of fertilized incomplete mesenteries was not confirmed in Epizoanthus protoporos sp. nov. or P. reticulatus in this study. Thus, further studies are needed to confirm if fertilized incomplete mesentery are a diagnostic characteristic to separate the genera Epizoanthus and Paleozoanthus. Confirmation should be achievable via molecular phylogenetic analyses of P. reticulatus. At the same time, it must be noted that based on similarities in sphincter musculature, Low et al. (2016) suggested that Paleozoanthus may correspond to genus Terrazoanthus in the family Hydrozoanthidae.

Based on molecular phylogenetic analyses, we found at least two independent origins for associations with gastropods in Epizoanthidae. In comparison to Epizoanthus protoporos sp. nov. and E. rinbou, Epizoanthus sp. S is located within a subclade consisting of hermit crab-associated E. ramosus. Epizoanthus sp. S is known to associate with the gastropod Unedogemmula unedo (Kiener, 1839). Furthermore, Epizoanthus sp. C sensu Reimer et al. (2010) is found on the empty shells of U. unedo inhabited by hermit crabs, although these two Epizoanthus species are clearly distinct by morphology and molecular phylogeny (Reimer et al., 2010), suggesting that mechanisms for the establishment of the symbiotic associations with gastropods and hermit crabs are different.

The numbers of symbiotic studies on zoantharian-gastropod associations conducted until now are few compared to those on actiniarian-gastropod associations. Thus, continued investigations examining the diversity of gastropod-associated zoantharians have the potential to greatly expand our overall comprehension of anthozoan-gastropod symbioses.