Snails associated with the coral-killing sponge Terpios hoshinota in Okinawa Island, Japan

Terpios hoshinota is a thin encrusting sponge that overgrows live scleractinian corals and it is linked to coral loss in many reefs. However, our knowledge of the species associated with this sponge species is poor. During a periodical survey of T. hoshinota in 2020, we found tiny snails crawling on the sponge in the subtropical waters around Okinawa Island, Japan. We observed egg capsules inside the sponge tissue and veliger larvae released from the egg capsules. Molecular analyses of both the snails and veliger larvae (cytochrome oxidase I, COI) showed that they were identical and belonged to Joculator sp. (family Cerithiopsidae). There was no direct observation of predation on the sponge by this snail; however, to the best of our knowledge, this is the first report on a close association between a snail and the sponge T. hoshinota.


Scientific Reports
| (2021) 11:20709 | https://doi.org/10.1038/s41598-021-00185-x www.nature.com/scientificreports/ attributed to their small size (< 2.5 mm in shell height, Fig. 3a-c) and dark coloration. Egg capsules with veliger larvae were found in the histological sections of the specimens from Sesoko Island (January 22, July 4, 2020) and from Onna (August 4, 2020). Cerithiopsidae (Fig. 3a-c) and Triphoridae (d: Coriophora fusca, e: Euthymella elegans) snails on T. hoshinotae were collected from August to November 2020. The mating behavior of the two snails was observed twice in September (Suppl. Movie 1), from the snails on the Nakijin sponge's surface and it continued even when the snails were moved to a Petri dish (Fig. 5a). Live egg capsules were found for the first time on July 24, 2020, from the Nakijin sample. Egg capsules at the stage of nearly releasing veliger larvae were visible as swollen bumps near the sponge surface (Fig. 6a, suppl. movie 2), and their size was similar to that of sponge larvae. The position of egg capsules was consistent with that of the coral calice. On the day of hatching, the egg capsules became swollen, and larvae became visible through the capsule membrane with decreasing density of sand particles trapped by the sponge. The larvae swam actively inside the capsule and then hatched, swimming out of the capsule (Fig. 6b-f). The exact time of release was observed only once in the aquarium around 8 pm on December 10 (Fig. 6e). The mean number of   www.nature.com/scientificreports/ veliger larvae per egg capsule was 111.7 ± 17.3 (mean ± SD; range 83-132, n = 6), calculated using ethanol-fixed egg capsules. The shell length of veliger was 138.6 ± 6.0 μm (mean ± SD; range 127.3-151.5 μm, n = 51). After hatching from the egg capsule, veliger larvae started to swim and showed strong positive phototaxis toward light (Fig. 5b, suppl. movie 3). We attempted to culture the larvae in a Petri dish with filtered seawater (< 0.45 µm), but they survived only a few days.   www.nature.com/scientificreports/ Egg capsules were found in the histological sections initially prepared for observing sponge reproduction. The sponges containing egg capsules were observed in the samples obtained from Sesoko Is. on July 4 and from Onna on August 4, 2020. Figure 7 shows many egg capsules laid deep into the tissue of Terpios hoshinota, and the sizes of the egg capsules (1.2 mm in diameter) were close to that of sponge larvae (Fig. 7b).
Molecular analyses based on COI gene sequences indicated that each snail of a and b types ( Fig. 3) and the veliger larvae are the same species, Joculator sp. In the phylogenetic tree ( Fig. 8), veliger larvae were included in a monophyletic clade with Joculator sp. supported by high bootstrap values (100%). In addition, low levels of genetic divergence, ranging from 0.6 to 1.5%, were observed between the two snail specimens identified as Joculator sp. and the veliger larvae. These values for the COI sequences of Joculator sp. were similar to the range of intraspecific divergences for the other cerithiopsid species (0.0 to 2.8% 40 ).

Discussion
In this study, we examined the snails associated with the sponge Terpios hoshinota for the first time. The number of snails observed in this study was small (< 6 individuals per species, Fig. 3); however, sponge-associated snails may be distributed widely, because snails, egg capsules, and veliger larvae were found at four Okinawa Island sites. Terpios-affected islands are abundant along the Ryukyu Archipelago 15 . Egg capsules and veliger larvae were observed between July and December in the present study, indicating that their reproductive season lasts for at least six months, from summer to fall.
Spongivores (sponge-eating organisms) include various animals, such as nudibranchs, snails, echinoids, fish, and turtles 9,10,26,28 . Relatively large (5-20 cm in length) dorid nudibranchs consume Terpios sp. in the northeastern Pacific 27,29 . Terpios hoshinota is a spiculate demosponge 13 and has a cytotoxic compound 30 ; therefore, this sponge is not palatable for predators. In addition, the sponge spicules (ca. 200 µm long) and particles on the surface of T. hoshinota tissues act as barriers to predators. However, this sponge armored with spicules, particles, and toxic substances would be a relatively safe place for snail larvae to lay their egg capsules. This study did not determine the direct evidence of the snails feeding on sponge tissues; however, there is a possibility that, like other cerithiopsids, these snails use sponges as a food source via excavation of soft tissue using their proboscises 31,32 .
In this study, we collected three different snail species from the surface of Terpios hoshinota. The number of snails was small; however, more intensive and quantitative surveys could find more sponge-associated snails, from the widely distributed Terpios in southern Japan. Therefore, survey of areas containing sponge-affected reefs along the Ryukyu Archipelago is required. It is possible that even if the sponge-associated snails consume www.nature.com/scientificreports/ sponges, they are unable to alter the growth of the sponge significantly, owing to their small size. Therefore, the snails are unlikely to be candidate biological control agents for inhibiting the spread of the coral-killing sponge Terpios hoshinota. However, studies on the species composition, geological distribution, and abundance of associates, including snails, would reveal a new view of the coral-killing sponge Terpios hoshinota as a host organism.    10″ N, 127°42′31.07″ E), all around Okinawa Island, Japan (Fig. 1). At all sites, dense aggregations of branching Montipora corals had developed in a shallow moat (maximum depth 2 m) together with massive Porites spp., foliose M. aequituberculata, leafy Pavona frondifera, corymbose Acropora spp., and other scleractinians. Some of these hard corals were fully or partly covered by Terpios hoshinota (Fig. 2). We collected the snails during the regular monthly sampling in Sesoko Island and Nakijin, during reproductive studies of the sponge, as well as from other sites where snail or veliger larvae were observed.
Collection of snails and veliger larvae. The small size and dark color of the snails made it difficult to find them on the black sponge in the field. Most snails were found during close observation using a dissecting microscope. Veliger larvae released from the sponge were trapped in a filter cup (100 µm nylon mesh filter, cell strainer, BD Biosciences Discovery Labware) together with sponge larvae. The histological observations for the reproductive studies in sponge tissues were conducted as follows: the tissues were fixed with 10% formalin solution, dehydrated with a graded series of ethanol, embedded in paraffin, and stained with hematoxylin/eosin dyes. Presence/absence of snail egg capsules in the sponge tissue were recorded. The snails, egg capsules, and veliger larvae were observed using a light microscope (Eclipse Ci, Nikon Co.), a dissecting microscope (SMZ-1000, Nikon Co.), and a digital microscope (Dino-Lite Premier, AnMo Elec. Co.) to obtain time-lapse images. The snails were observed in the field on the collection day; other observations and culture experiments were performed in the marine laboratory at Sesoko Station, Tropical Biosphere Research Center, University of the Ryukyus.

Molecular identification of snails and veliger.
We could collect multiple samples of only two morphological types of snails. The shell of a type was brown (Fig. 3a), and that of b type was sandy-yellow with a dark red-brown suture (Fig. 3b,c). Two cerithiopsid snails (one specimen each of a and b types; Fig. 4a,b for suture, c and d for protoconch, respectively), collected from Odo in October 2020, and approximately 60 individuals of unidentified veliger larvae were fixed and preserved in pure ethanol for morphological and molecular identification. Cerithiopsid snails were identified to the genus level based on shell morphology, as described previously 31,[33][34][35][36] . In addition to the visual morphological identification, molecular identification was performed using cytochrome c oxidase subunit I (COI) sequences. The total DNA of snails and veliger larvae was extracted from foot tissue and 20 whole veligers, respectively, using the DNeasy Tissue Extraction Kit (Qiagen). The mitochondrial COI sequences (658 bp) were amplified through polymerase chain reaction (PCR) using the primer pairs LCO1490 and HCO2198 37 , following the conditions described earlier 38 . The PCR products were visualized through electrophoresis on a 1.5% Tris-Borate-EDTA agarose gel and purified with ExoSAP-IT (Thermo Fisher Scientific). The purified products were Sanger sequenced in both directions using an ABI 3730xl Genetic Analyzer (Applied Biosystems) at Eurofins Genomics (Tokyo, Japan). The COI sequences were manually aligned using Mesquite version 3.61 38,39 and compared with previously reported sequences of cerithiopsid species 40,41 . Genetic divergences among the sequences were quantified using the Kimura 2-Parameter (K2P) distance model 42 using MEGA X version 10.1.7 43 . Phylogenetic relationships of cerithiopsid species were reconstructed from COI sequences (621 bp) using the maximum-likelihood (ML) methods. The ML tree reconstruction was performed under GTR + G model in RAxML v.7.4.2 44 with a bootstrap analysis of 1,000 pseudoreplicates. Nucleotide sequences were deposited in the DNA Data Bank of Japan (DDBJ) under the accession numbers LC598716-LC598717 for snails and LC598718 for veliger larvae. The sequenced specimens were deposited as a voucher (specimen number: 20210831-HF010-12) in the Atmosphere and Ocean Research Institute (AORI), The University of Tokyo (https:// www. aori.u-tokyo. ac. jp, contact person: Hiroaki Fukumori, fukumori@aori.utokyo.ac.jp).
Sampling and field studies. All necessary permits for sampling and observational field studies were obtained from the concerned authorities. Coral sampling was performed with approval from the authorities of Okinawa Prefecture, Japan.

Data availability
The datasets generated or analyzed during the current study are available from the corresponding author upon reasonable request.