A new species of Spirobranchus (Annelida: Serpulidae) is described based on specimens collected at the coastal Shonan area of Sagami Bay and the adjacent areas of Honshu, Japan. Spirobranchus akitsushima sp. nov. forms large aggregations in the intertidal rocky zone of warm-temperate Japanese shores. This species was referred to as Pomatoleios kraussii (Baird, 1864) until the monotypic genus Pomatoleios was synonymized with Spirobranchus. This new species is formally described based on morphologically distinct Japanese specimens with supporting DNA sequence data. The calcareous opercular endplate of Spirobranchus akitsushima sp. nov. lacks a distinct talon, but some specimens have a slight rounded swelling on the endplate underside, while in other species of the S. kraussii complex a talon is present, usually extended, and with bulges. We examined sub-fossil tube aggregations of the new species and suggest that such aggregation stranded ashore is a good indicator of vertical land movements (uplift and subsidence) resulting from past events, such as earthquakes, in Honshu, Japan.

Cosmopolitan species, paleo-aggregation, sea level indicator

The family Serpulidae Rafinesque, 1815 is a unique and distinct group of marine annelids that inhabits self-secreted calcareous tubes and is recorded in all habitats of the world oceans, from the intertidal zone, shallow-water coral reefs to abyssal and hadal depths, as well as in brackish and freshwater habitats. Currently, the family comprises 562 valid species in 69 genera (Capa et al. 2021). The most speciose genus of the family is Hydroides Gunnerus, 1768, with more than 100 species; Spirobranchus de Blainville, 1818 is the second largest genus with 36 nominal species (Capa et al. 2021; Tables 3, 4).

Approximately 70 serpulid species have been recorded in Japanese waters (Nishi et al. 2017). Among them, 11 are species of the genus Spirobranchus, while the morphospecies S. cruciger (Grube, 1862) and S. gaymardi Quatrefages, 1866 are considered synonyms of S. corniculatus (Grube, 1862) based on a recent genetic study (Willette et al. 2015). The group of species commonly known as Christmas Tree Worms is the most conspicuous in the genus Spirobranchus because of its brilliantly colored spiral radiolar crowns. These large-bodied species (e.g., S. corniculatus and S. gardineri Pixell, 1913 in the Pacific) are associated with hermatypic corals in warm temperate to tropical waters of Japan.

Another well-known species of Spirobranchus is distributed in temperate to sub-tropical Japanese coastal areas from Honshu to Kyushu, and in the vicinity of the Nansei Archipelago. This species is known in Japan under the common name “Yakko-kanzashi Gokai” because the ventral side of its opercular peduncle has two dark lateral bands on a white background, which makes it look like “Yakko”: this Japanese word describes a unique hairstyle (or a person with such a hairstyle) with a shaved top of the head and hair around the ears cut in the shape of a plectrum (pick) used for Samisen, a traditional Japanese stringed instrument (Otsuki 1935). As “Kanzashi” is a Japanese word for an ornamental hairpin and “Gokai” means a polychaete worm, then “Kanzashi-Gokai” is a Japanese common name for serpulid polychaetes.

“Yakko-kanzashi” is a gregarious species commonly forming distinct intertidal belts along with barnacles and bivalves. Morphologically, the specimens of “Yakko-kanzashi” are characterized by opercula covered with simple endplates, arrangement of radioles in two semi-circles, absence of collar chaetae in adults, and tough thick-walled blue or purple tubes with sharp or flattened keels. This species has been recorded under a number of scientific names. Initially it was referred to (e.g., Okuda 1937, 1940; Utinomi 1956) as Pomatoleios crosslandi Pixell, 1913, a species originally described from off Tanzania. After P. crosslandi was synonymized with Pomatoleios kraussii (Baird, 1864), the attribution of the Japanese population changed accordingly (e.g., Imajima and Hartman 1964; Okuda and Imajima 1965; Uchida 1992). Most recently it was referred to as Spirobranchus kraussii (e.g., Nishi et al. 2017) because the genus Pomatoleios was synonymized with Spirobranchus by Pillai (2009). The reported range of the nominal species spans in Japan from northern Honshu to tropical Okinawa (e.g., Onagawa Bay: Okuda 1937; Okinawa: Okuda 1940; Nishi 1993; Sagami Bay: Imajima 1968; Wakayama, Izu, Kochi: Uchida 1978). Some studies examined its distribution (Nishi 1993), early development (Sawada 1988), and life history (Miura and Kajihara 1984).

The assignment of the Japanese Spirobranchus “Yakko-kanzashi” to the morphologically similar intertidal belt-forming Spirobranchus kraussii was based on the wide distribution attributed to S. kraussii. After its original description from warm-temperate coasts of South Africa, the taxon was subsequently reported from numerous tropical and subtropical localities (Persian (Arabian) Gulf, Pakistan, Sri Lanka, Philippines, Hawaii, Australia, China (including Hong Kong), Japan, Korea, Singapore, Suez Canal, and eastern Mediterranean, see Simon et al. 2019). However, such wide, nearly cosmopolitan distributions were recently questioned (Hutchings and Kupriyanova 2018). Genetic studies revealed that this warm temperate intertidal species is restricted to South African coasts and that taxa under this name from other areas belong to a large complex of regionally distributed species (Simon et al. 2019; Pazoki et al. 2020; Sivananthan et al. 2021).

Two specimens collected in Japan from Manazuru, Sagami Bay, Honshu and deposited in the Australian Museum (AM W.49980 and AM W.49981) were sequenced and used in the study of Simon et al. (2019). The sequences formed a distinct genetic lineage denoted as Spirobranchus sp. 1 by Simon et al. (2019) providing evidence supporting the presence of an undescribed species of the S. kraussii complex in Japan. The most recent genetic study by Kobayashi and Goto (2021) recovered three genetic lineages within the S. kraussii complex in Japan, which suggests that there are at least three unnamed species in Japan: S. sp. 1 from warm temperate localities (Seto, Wakayama), and two from tropical Okinawa (Spirobranchus spp. 5 and 6).

Serpulids forming intertidal belts and relics of such assemblages are useful fixed biological indicators (FBIs) as they provide data on short-term fluctuations in sea-level (Baker et al. 2001b). A belt-forming Australian serpulid Galeolaria caespitosa Lamarck, 1818 was used as a marker species in relative sea-level height analyses of past environmental changes (Bird 1988; Baker et al 2001a, b). The height differential of fossil to living encrustations is a simple and reliable measure of changes on tectonically stable coasts of eastern Australia (Baker and Haworth 1997). Japanese “Yakko-kanzashi”, occupying intertidal habitats similar to those of G. caespitosa, is a useful paleoindicator of sea-level changes caused by tectonic events, such as earthquakes. While current aggregations are always found at the sea level, paleo-aggregations are stranded ashore far above it. In Tanabe Bay, Kii Peninsula, current aggregations had the upper limit of +0.1 – +0.2 m from the mean sea level (MSL) (Nishimura 1972). Kayanne et al. (1987) defined dense aggregation of tubes as “almost 100% of areas of 10 cm⁻² were covered by serpulid tubes”, and they reported a similar upper limit (+0.1 to ± 0.1 m from MSL) of dense aggregations found on Boso Peninsula, Chiba. Comparisons of Nishibata et al. (1988) revealed the upper limit of the current population (= dense aggregation) as +12–± 2 cm from MSL, while that of the fossil ones ranged from +68 to +235 cm. Nishibata et al. (1988) showed that paleo-aggregations at the site of Taisho-Kanto great earthquake in A.D. 1923 were located 1.2–1.4 m above MSL, while those found in the vicinity of Genroku-Kanto great earthquake in A.D. 1703 raised to 2.3 m above MSL. Similarly, Maemoku and Tsubono (1990), Shishikura (2003a) and Shishikura et al. (2007) used uplifted paleo-aggregations to reconstruct the earthquake history along Miura, Boso and Kii Peninsula, Honshu. In Muroto, Kochi, Maemoku (2001) estimated that the older tube aggregations uplifted to 8.3–9.1 m between 2800 and 4500 years ago as a result of an earthquake.

The main aim of this study is to formally describe and name the common intertidal gregarious species of Sagami Bay and adjacent areas previously referred to as S. kraussii, using a combination of morphological and molecular data. We also examine and describe in detail paleo-aggregations (stranded ashore and rarely overlapping with the current tube aggregations) of this species.

Specimens were collected around Sagami and Suruga Bay (Fig. 1A, B) and specimens from Chichijima Island, Ogasawara were added for a comparison. The specimens designated as types were collected in Wakaejima, Kamakura, Sagami Bay (Fig. 1C, D). Current and paleo- tube aggregations of the species were photographed (and some tubes were collected) at Tsurugizaki and Jogashima (Fig. 1F–I) and altitudes of their aggregations were compared to current MSL.

Figure 1. 

Map of collection sites A Japan and adjacent seas B Sagami Bay, Suruga Bay, and Pacific side of Honshu, and collection sites on Miura Peninsula and Yokohama C collection sites of Miura Peninsula and Yokohama D Wakaejima, Kamakura, type locality of Spirobranchus akitsushima sp. nov. E Hayama F Western part of Miura Peninsula, showing Tsurugizaki and Jogashima G Jogashima H close-up view of collection sites of Jogashima I close-up view of Tsurugizaki. Key: ○: paleo-aggregation, ●(red): current distribution. Arrow in H indicates (A) in Fig. 3; P1–P3 and C1 in I indicates site of (I) and (J) in Fig. 3.

The holotype, paratypes and additional specimens were deposited in the Natural History Museum and Institute, Chiba (CBM-ZW), Japan, the Coastal Branch of Natural History Museum and Institute, Chiba (CMNH-ZW), Katsuura, Chiba, and Marine Science Museum, Tokai University (MSM-INV), Shimizu, Shizuoka, Japan. Two specimens are deposited in the Australian Museum (AM) (AM W.49980 and AM W.49981).

Terminology for voucher specimens used to produce molecular samples was used following Pleijel et al. (2008). Hologenophore is a specimen voucher from which the molecular sample is derived, paragenophore is a putatively conspecific voucher specimen collected together with the ‘molecular specimen’, and syngenophore is a voucher collected at another place or time.

A total of 14 worms for which DNA has been sequenced (hologenophores sensu Pleijel et al. 2008) were preserved in 75% ethanol. Some paratypes and non-type specimens were anesthetized with magnesium hydroxide and photographed alive or after being fixed in 10% formalin seawater. In order to examine the morphology of the lower endplate surface (presence of the talon and its shape), endplates were taken out from the opercular tissue using scalpel and forceps.

For scanning electron microscopy (SEM) observation specimens were dehydrated through gradual series of ethanol for 10 min in each and finally washed with 100% ethanol for 10 min. The samples then were washed with 1:1 and 1.5:0.5 mixture of 100% ethanol and hexamethyldisalazane (HMDS) for 10 minutes in each, and finally washed with 100% HMDS for 10 min following Nation (1983) and Wang et al. (2018). Specimens were left overnight to ensure HMDS evaporation, then coated with platinum and viewed under a JEOL JF7001FM at the Instrumental Analysis Center of Yokohama National University.

The partial sequences of the mitochondrial cytochrome b (cytb) gene, nuclear internal transcribed spacer-2 (ITS2) region, and 18S and 28S rRNA genes were used for comparisons with congeneric species. Genomic DNA was extracted from posterior abdomens of ethanol-fixed worms collected from the Shonan area (Sagami Bay) and from Omaezaki (Suruga Bay) (Table 1) by heating at 96 °C for 20 min in 50 μl of TE buffer with 10% Chelex 100 (Bio-Rad Laboratories, Hercules, CA) according to Richlen and Barber (2005). Undiluted or 10-fold diluted DNA extract was used as a template for polymerase chain reaction (PCR). The 10 μL reaction mix contained 7.05 μL of sterilized water, 0.05 μL of TaKaRa Ex Taq Hot Start Version (TaKaRa Bio, Kusatsu, Japan), 1.0 μL of 10× Ex Taq Buffer, 0.8 μL of 2.5 μM dNTP mixture, 0.05 μL of 50 μM of each forward and reverse primers, and 1.0 μL of template DNA for the mitochondrial cytb gene and the nuclear ITS2 region. The 25 μL reaction mix contained 11.3 μL of sterilized water, 12.5 μL of 2 × KOD One PCR Master Mix (TOYOBO, Osaka, Japan), 0.1 μL of 50 μM each of forward and reverse primers, and 1.0 μL of template DNA for nuclear 18S rRNA gene. The 10 μL reaction mix contained 4 μL of sterilized water, 5 μL of 2 × KOD One PCR Master Mix (TOYOBO, Osaka, Japan), 0.05 μL of each 50 μM forward and reverse primers, and 1.0 μL of template DNA for nuclear 28S rRNA gene.

The primer pairs used for PCR amplifications and sequencing are listed in Table 2. The PCR cycling conditions were (1) initial denaturation at 94 °C for 120 s followed by 35–45 cycles of denaturation at 94 °C for 30 s, annealing at 45 (for cytb) or 50 °C (for ITS2) for 40 s, and extension at 72 °C for 20 s, and a final extension at 72 °C for 300 s (TaKaRa Ex Taq), (2) 36 cycles of 98 °C for 10 s, 58 °C for 5 s, and 68 °C for 2 s for 18S rRNA gene (KOD One PCR Master Mix), and (3) 32 or 36 cycles of 98 °C for 10 s, 62 °C for 5 s, and 68 °C for 1 s for 28S rRNA gene (KOD One PCR Master Mix). The PCR products were purified using EnzSAP PCRClean-up Reagent (EdgeBio, San Jose, CA) and sequenced by Eurofins Genomics (Tokyo, Japan). The forward and reverse complementary sequences and contigs were assembled using GeneStudio ver. 2.2.0.0 (GeneStudio, Inc., Suwanee, GA). The obtained sequences have been deposited in the DDBJ/ENA/GenBank databases with accession numbers LC661622–LC661671 (Table 1). Intra-specific pairwise genetic distances (p-distance) for cytb sequences of Spirobranchus species were determined using MEGA version 11 software under default settings (Tamura et al. 2021).

Phylogenetic analyses based on concatenated gene sequences (cytb + ITS2 + 18S + 28S) and sequences of each gene/region were conducted using the sequences obtained in the present study supplemented with those sourced from DDBJ/ENA/ GenBank databases (Table 1). The sequences of Galeolaria hystrix Mörch, 1863 and G. gemineoa Halt, Kupriyanova, Cooper & Rouse, 2009 were used as outgroups. The sequences of each gene/region were aligned using the MAFFT online service ver. 7 with the L‐INS‐i algorithm (Katoh et al. 2019). Ambiguously aligned regions of alignments were eliminated by employing Gblocks server ver. 0.91b (Castresana 2000) with the following less stringent settings: minimum number of sequences for a conserved/flank position were half the number of sequences + 1, maximum number of contiguous non-conserved positions was eight, minimum length of a block was five, and with half of the allowed gap positions. The final lengths of the alignments were 359 (cytb), 528 (ITS2), 1717 (18S), and 774 (28S) bp for the multiple sequence alignment.

Maximum likelihood (ML) analyses performed using IQ-TREE (Nguyen et al. 2015) implemented in PhyloSuite under Edge-linked partition model. For the concatenated dataset, the HKY+F+I+G4, K2P+I, TNe+I and TIM3+F+G4 models were selected for the cytb, ITS2, 18S and 28S rRNA gene/regions, respectively as the best-fit substitution model by ModelFinder (Kalyaanamoorthy et al. 2017) as implemented in IQ-TREE under the Bayesian information criterion (BIC). For the single gene/region data, the K3Pu+F+I+G4, K2P+G4, TNe+I and TIM3+F+G4 models were selected for the cytb, ITS2, 18S and 28S rRNA gene/region respectively. The robustness of the ML trees was evaluated by the Shimodaira-Hasegawa-like approximate likelihood-ratio test (SH-aLRT) with 5,000 replicates (Guindon et al. 2010), approximate Bayes (aBayes) test (Anisimova et al. 2011), and ultrafast bootstraps (UFBoot) with 5000 replicates (Hoang et al. 2018).

In addition to Spirobranchus kraussii and S. cariniferus, five new species, one from Arabian (Persian) Gulf, one from Brazil, two from South Asia, and the last one from Japan, identifiable mainly by the opercular characters, were recently formally described and named in the Spirobranchus kraussii complex (e.g., Brandão and dos Santos Brasil 2020; Pazoki et al. 2020; Sivananthan et al. 2021; this study). Spirobranchus lirianeae from Brazil was described without molecular data, and is identifiable by its opercular morphology as well as by its non-gregarious populations inhabiting subtidal habitats. Spirobranchus manilensis from oyster beds of South-East Asia was also described without molecular data, but it is identifiable by opercular morphology (see Table 3).

Live aggregations of Spirobranchus akitsushima sp. nov. are common on the shorelines of Sagami Bay and Miura Peninsula, while sub-fossil tube aggregations have also been recorded in Jogashima and Tsurugizaki along Miura Peninsula. The blue- or purple-colored subfossil tubes with prominent characteristic keels and lateral transversal ridges were well preserved (Fig. 3E, G) and stranded ashore well above MSL. The lower one of Tsurugizaki might be a result of the Taisho-Kanto great earthquake in 1923, and the upper one possibly resulted from the Genroku-Kanto earthquake in 1703, as suggested by Nishibata et al. (1988) and Shishikura (2003a, b). It means that we have records of S. akitsushima sp. nov. dating from at least 300 years.

Fouling serpulids forming aggregations on artificial substrates are commonly reported as introduced or cryptogenic species (possible introductions) (see Ruiz et al. 2000). Vectors of serpulid introductions are shipping, including hull fouling, and fisheries, including fouling on commercial mollusks such as oysters, scallops, turban shells, and abalones (Ruiz et al. 2000). Highly successful invasive serpulids, such as Hydroides elegans (Haswell, 1883) and H. ezoensis Okuda, 1934, have been found in large aggregations on ship hulls, a prominent vector of species translocation, and in communities on experimental fouling panels suspended in harbors. These Hydroides species have also been frequently recorded on oysters, scallops, and other molluscan shells, another vector of introduction. In contrast, Spirobranchus akitsushima sp. nov., although very common on natural substrates, was only reported from unspecified artificial substrates in coastal areas (e.g., Horikoshi and Okamoto 2007) and on concrete blocks of wave breakers and harbor walls in Yokohama harbor (Fig. 2H–J). Specimens of Spirobranchus akitsushima sp. nov. have been rarely found on experimental panels (Miura and Kajihara 1983; Raveendran and Harada 2001) and are not found on shells of commercial mollusks. Their distributions are limited to intertidal areas, and Miura and Kajihara (1983) reported that the species appeared 30–80 cm above the mean high water spring tide in Aburatsubo Bay and that the settlement of larvae was not observed on submerged experimental plates. Thus, anthropogenic translocation to other oceans is unlikely to occur. We argue that Spirobranchus akitsushima sp. nov. is a species native to Japan, not a non-indigenous species or invader. The species is likely to have regionally restricted distributions around Japan as supported by DNA sequence data and presence of fossilized tube aggregations.

Our molecular phylogenetic analysis using four molecular markers (cytb, ITS2, 18S, and 28S rDNA) has led us to distinguish species among morphologically very similar taxa of S. kraussii complex in Japan. The present study showed that the specimens from Manazuru as mentioned by Simon et al. (2019) and other newly sequenced specimens from eastern Sagami Bay and Omaezaki, western-most part of Suruga Bay, belong to the same species described here as S. akitsushima sp. nov. This new species is distributed along the Pacific coastline of Honshu from Sagami Bay in the north to Shirahama in the south. The results of molecular analysis suggest that the S. akitsushima sp. nov. is genetically distinct from the other S. kraussii-complex species described from outside Japan. Interspecific p-distance between the cytb sequences of S. akitsushima sp. nov. and the other described S. kraussii-complex species were found to be 19.4 to 24.5% (Table 3), which is comparable to that observed within the available members of S. kraussii-complex species (14.6–26.9%) and other serpulid genera such as Ficopomatus (19.2%, Styan et al. 2017), Galeolaria (22.8–24.5%, Halt et al. 2009), and Hydroides (15.8–23.1%, Sun et al. 2016).

Kobayashi and Goto (2021) reported three unnamed genetic lineages of S. kraussii complex in Japan: S. sp. 1 (= S. akitsushima sp. nov.) from Seto, Wakayama, southern Honshu and two from Okinawa, S. sp. 5 from Yagachi and S. sp. 6 from Oura Bay. The presence of two distinct species of the complex in Japan was expected because of the boundary between Osumi Islands and Ryukyu Islands, known as Tokara Tectonic Straight or Tokara Gap, where the Kuroshio current crosses the Ryukyu Islands chain from the west to the east (see Motokawa 2017). As expected, Spirobranchus sp. 5 showed a 20.7–22.2% differences in cytb gene sequences with S. akitsushima sp. nov. and S. sp. 6. Such distance is commonly found between morphologically distinct congeneric species (e.g., Willette et al. 2015; Pazoki et al. 2020) leading Kobayashi and Goto (2021) to the conclusion that Spirobranchus sp. 5 is a genetically and ecologically distinct undescribed species.

The status of Spirobranchus sp. 6 sensu Kobayashi and Goto (2021) is less certain. Unexpectedly, it is genetically closer (3.7–4.1% only in cytb) to S. akitsushima sp. nov. from Honshu than to S. sp. 5 also from Okinawa (21.7% in cytb). Kobayashi and Goto (2021) suggested that the genetic differences between Honshu and Oura Bay are quite large, considering the lack of genetic differentiation for specimens within Honshu Island or low genetic diversity at each studied locality. They also noted that “either interbreeding still exist between the lineages in Shirahama and Oura Bay, or that the sorting of the two lineages is incomplete” (Kobayashi and Goto 2021: 13). Clearly, we need to study the population structures of Amami Archipelago and Kyushu situated between Honshu and Okinawa Islands before we can determine whether or not specimens of S. akitsushima sp. nov. and S. sp. 6 belong to the same species.

Future genetic studies of these Japanese and other Asian populations (e.g., Paik 1989: Korea; Sun and Yang 2014; Huang et al. 1992: China; Sun et al. 2012: Hong Kong) might reveal other distinct species from S. kraussii complex.

We are grateful to Dr. G. Kobayashi for his unpublished observations and sequence data, and for samples from Seto. We thank the staff of the Instrumental Analyses Center of Yokohama National University for the use of scanning electron microscope. We thank R. Bastida-Zavala and an anonymous reviewer for their useful suggestions and corrections. This study was partly supported by the grant (JPMEERF20204R01) from the Environment Research and Technology Development Fund of the Environmental Restoration and Conservation Agency of Japan to HA, and the Australian Biological Resources Study (ABRS) grant RG18-21 to EK.