Sample collection and observations

We have investigated the diversity of commensal galeommatoideans in mud flats in Hakatajima, Ehime Prefecture and Takehara, Hiroshima Prefecture, Japan since 2009 by digging up various invertebrates' burrows (Goto et al., 2011, 2012, 2014). Sipunculus nudus (Sipunculidae) is one of the major burrowing invertebrates in these study sites. This sipunculan lives in temporal burrows, which are horizontal, not branched, and weakly lined with mucus, in bottom sediments. We collected P. rugata from the burrows of S. nudus in the middle or lower intertidal zone of mud flats at spring low tide during 2011 and 2012. The bivalves were preserved in 70–100% ethanol and brought back to the laboratory. The detailed observations of shell morphology and anatomy were performed under a dissecting microscope. We used a specimen collected from Takehara for molecular analysis (Table 1).

DNA extraction, PCR, and sequencing

Total genomic DNA was isolated from the bivalves following a previously described method (Goto et al., 2012). A small piece of soft tissue was homogenized in 800-µl lysis buffer and incubated at 55°C overnight, after which 80 µl of saturated potassium chloride was added to the lysate. This solution was incubated for 5 min on ice and then centrifuged for 10 min. The supernatant (700 µl) was transferred to a new tube, cleaned once with a phenol/chloroform solution, and precipitated with an equal volume of 2- propanol. The DNA pellet was rinsed with 70% ethanol, vacuum-dried, and dissolved in 100-µl TE buffer.

Table 1.

GenBank accession numbers of the specimens used in this study. The sequences obtained in this study are shown in bold with asterisk.

We sequenced fragments of the 18S and 28S genes. Polymerase chain reactions (PCRs) were used to amplify ∼1700 bp of 18S and ∼1000 bp of 28S. Amplifications were performed in 20-µl mixtures consisting of 0.4 µl of forward and reverse primers (20 µM each; Table 2), 2.0 µl of ExTaq buffer, 1.6 µl of dNTPs (2.5 µM each), 0.1 µl of ExTaq polymerase (TaKaRa, Otsu, Japan), and 15.1 µl of distilled water. Thermal cycling was performed with an initial denaturation for 3 min at 94°C, followed by 30 cycles of 30 s at 94°C, 30 s at a gene-specific annealing temperature (50–55°C), and 2 min at 72°C, with a final 3 min extension at 72°C. The sequencing reaction was performed using PCR primers and internal primers (Table 2) and a BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA); the reaction products were electrophoresed on an ABI 3130 sequencer (Applied Biosystems). The obtained sequences were deposited in the DDBJ/ EMBL/GenBank databases with accession numbers LC126832-LC126833 (Table 1).

Phylogenetic analyses

In addition to the sequences of P. rugata obtained in this study, we also collected sequence data of other galeommatoideans and outgroups from GenBank (Table 1). Sequences of the 18S and 28S genes were aligned using the Muscle program (Edgar, 2004) with default settings in the software Seaview (Galtier et al., 1996; Gouy et al., 2010). We employed Gblocks v0.91b (Castresana, 2000; Talavera and Castresana, 2007) to eliminate the ambiguously aligned regions in 18S and 28S alignments. Phylogenetic trees were constructed using the Bayesian and maximum likelihood (ML) methods. Bayesian analyses were performed using MrBayes 3.1.2 (Ronquist and Huelsenbeck, 2003) with substitution models chosen by Kakusan 4 (Tanabe, 2011). In the combined data set, substitution parameters were estimated separately for each gene partition (18S: SYM Gamma, 28S: GTR Gamma). Two independent runs of Metropolis-coupled Markov chain Monte Carlo were carried out simultaneously, sampling trees every 100 generations and calculating the average standard deviation of split frequencies (ASDSFs) every 1000 generations. Using the ‘stoprule’ option, analyses were continued until ASDSF dropped below 0.01, at which point the two chains were considered to have achieved convergence. As ASDSF was calculated based on the last 75% of the samples, we discarded the initial 25% of the sampled trees as burn-in. We confirmed that analyses reached stationarity well before the burn-in period by plotting the ln-likelihood of the sampled trees against generation time. Maximum likelihood analyses were performed using RAxML (Stamatakis, 2006) as implemented in raxmlGUI 1.31 (Silvestro and Michalak, 2012). The robustness of the ML tree was evaluated by 1000 bootstrap replications. Datasets were partitioned by gene and the GTRGAMMA model was implemented.

Table 2.

Primers used in this study.

Fig. 1.

Platomysia rugata and its host sipunculan Sipunculus nudus. (A) Left side of P. rugata. (B) Right side of P. rugata. (C) Dorsal view of P. rugata. (A-C) are different individu-als. (D) Platomysia rugata and its host S. nudus. (E, F) Platomysia rugata in the burrows of S. nudus. Each arrow indicates P. rugata.

Host association

We surveyed approximately 200 and 45 burrows of S. nudus in Hakatajima and Takehara, respectively, resulting that eight and three specimens of P. rugata were collected in each study site. All the bivalves were found as a solitary in the burrows of S. nudus (Fig. 1D-F). The occurrence rate of P. rugata in S. nudus burrows was ∼4 and ∼6.7% in Hakatajima and Takehara, respectively. The bivalves were attached to the burrow-wall surface mainly near the host body (Fig. 1E, F), whereas those directly attaching to the hosts were not found. We observed the bivalve behavior in an aquaria with its host for a while. The bivalve hided under the host body, but did not attach to the host. The other sipunculan Siphonosoma cumanense commonly inhabits mud flats in these study sites (Goto et al., 2011). We surveyed at least 20 burrows of S. cumanense in each study site. However, P. rugata was not collected from the S. cumanense burrows. We also surveyed the burrows of echiurans (Ikedosoma gogoshimense and Arhynchite hayaoi), holothurians (Protankyra bidentata), and echinoids (Echinocardium cordatum), but did not find any P. rugata.

Shell morphology and anatomy

Shell: The shell is very thin and fragile, its shape elongate-ovate and equivalve and nearly equilateral with beak in the middle (Figs. 1, 2). The anterior is slightly more expanded than posterior. Shell is covered by whitish, thin periostracum (Figs. 1, 2). Sculpture is prominent, consisting of distinct and evenly spaced commarginal ribs and close-set concentric striae at shell ventral margin. Radial ribs are absent. The pallial line is entire. Hinge of each valve consists of a single stout cardinal tooth in front of the umbo and well-developed oblique internal ligament posterior to the umbo (Fig. 2F, G).

Anatomy: A single large inner demibranch is present and covers the frequently branched ovary (Fig. 2B, C). Small labial palps are attached to the anterior end of the demibranch. Anterior and posterior adductor muscles are ovate, anterior is more elongate and larger than posterior (Fig. 2B). Foot is moderate-sized, tongue-shaped, with a small, prominent heel and a large rounded toe (Fig. 1A, B). The mantle edges narrowly extend beyond the margin of the shell and bear numerous regularly arranged very short papillae (Fig. 1A, B). No prominent tentacles were observed.

Fig. 2.

Platomysia rugata. Left valve (A-C). Right valve (D, E). The hinge structure of left valve (F) and right valve (G). Shell length = 5.5 mm. Abbreviations: aa, anterior adductor muscle; ac, anterior cardinal tooth; f, foot; id, inner demibranch; il, internal ligament; ov, ovary; pa, posterior adductor muscle.

Phylogenetic position

To determine the phylogenetic position of P. rugata in Galeommatoidea, we performed molecular phylogenetic analyses using 41 galeommatoideans and eight outgroup species (including five heterodont and three non-heterodont bivalve species) (Table 1). In the resulting tree (Fig. 3), P. rugata was monophyletic with Pseudopythina, Peregrinamor, and Byssobornia (Bayesian posterior probability = 1.00, Bootstrap support value = 100). Sipunculan-associated species was not recovered as monophyletic (Fig. 3). Platomysia rugata is relatively closely related with another sipunculan-associated species Pseudopythina aff. nodosa (Fig. 3), although they were not monophyletic (Fig. 3).

Fig. 3.

Phylogenetic position of Platomysia rugata within Galeommatoidea. The Bayesian tree was reconstructed based on combined sequence data (18S + 28S). Numbers above branches indicate Bayesian posterior probabilities followed by supporting maximum likelihood bootstrap values. White circles indicate free-living species; black circles, commensal species associated with non-sipunculan hosts; blue cir-cles, commensal species associated with sipunculan hosts. Information on host taxa is based on Goto et al. (2012) and Goto et al. (2014).