We describe a new species of polyclad flatworm, Paraplehnia seisuiae sp. nov., from 298–310 m depths in the Sea of Kumano, West Pacific, Japan. Paraplehnia seisuiae sp. nov. is characterized by i) a developed muscular wall proximally occupying about one-third of the prostatic vesicle, ii) no common duct between the spermiducal bulbs and the prostatic vesicle, and iii) a genital pit between the male and female gonopores. We provide a partial sequence (712 bp) of the mitochondrial cytochrome c oxidase subunit I gene as a DNA barcode for the species. Our phylogenetic analyses based on 603-bp 28S rDNA sequences indicate that P. seisuiae sp. nov. is nested in a clade consisting of stylochoid species along with unidentified species of Stylochus. It suggests that Plehniidae belongs to Stylochoidea, although this should be confirmed by future studies that contain Plehnia arctica (Plehn, 1896), the type species of the type genus of the family. The interfamily relationship among the superfamily Stylochoidea remains poorly resolved.

28S rDNA, Acotylea, Bayesian inference, COI, marine flatworm, maximum likelihood, taxonomy

Polyclad flatworms in the family Plehniidae Bock, 1913 are characterized by possessing sperm ducts or a common sperm duct entering the neck of a prostatic vesicle, the latter lacks an ejaculatory duct (Hyman 1953). The majority of plehniids have been reported from sublittoral zones by dredging (e.g. Bock 1913, 1923; Kato 1939; Hyman 1953; Hagiya 1993); some species were described from more than 200 m depths (Bock 1913; Hyman 1953).

The superfamilial affinity of Plehniidae has not been molecularly tested, while both Faubel (1983) and Prudhoe (1985) placed the family in Stylochoidea Poche, 1926 based on morphological characters. No plehniid has been represented in recent molecular phylogenetic analyses (Aguado et al. 2017; Bahia et al. 2017; Tsunashima et al. 2017; Litvaitis et al. 2019), although Bahia et al. (2017) indicated that Plehniidae possibly belongs to Cryptoceloidea Laidlaw, 1903. There has been a conflict between Faubel (1983) and Prudhoe (1985) as to the genus-level classification in Plehniidae. Faubel (1983) divided Plehniidae into three genera: Diplehnia Faubel, 1983, Discocelides Bergendal, 1893, and Plehnia Bock, 1913. Prudhoe (1985) separated this family into four genera: Discocelides, Nephtheaplana Prudhoe, 1985, Paraplehnia Hyman, 1953, and Plehnia. Later, Newman and Cannon (1997) established a new genus, Myoramyxa, within Plehniidae in the sense of Prudhoe (1985).

In this paper, we describe a new species of plehniid flatworm from Japan. We provide a partial sequence of the cytochrome c oxidase subunit I (COI) gene as a DNA barcode for the new species. We estimate the phylogenetic position of Plehniidae, represented by the new species, among other acotylean polyclads by molecular analyses using partial 28S rDNA sequences.

A single polyclad specimen was collected by dredging during the research cruise No. 1722 by Training/Research Vessel (TRV) Seisui-maru. The worm was anesthetized in a MgCl₂ solution prepared with tap water so that it had the same refractive index (or “salinity”) as the seawater, using an IS/Mill-E refractometer (AS ONE, Japan), and then photographed with a Nikon D5300 digital camera with external strobe lighting provided by a pair of Morris Hikaru Komachi Di flash units. For DNA extraction, a piece of the body margin was cut away from the specimen and fixed in 100% ethanol. The rest of the body was fixed in Bouin’s solution for 24 h and preserved in 70% ethanol. It was then cut into two (anterior and posterior) pieces. Both pieces were dehydrated in an ethanol series and cleared in xylene, then embedded in paraffin wax, sectioned at 7 µm thickness, stained with hematoxylin and eosin (HE), and embedded in Entellan New (Merck, Germany). They were observed under an Olympus BX51 compound microscope and photographed with a Nikon D5300 digital camera.

Sections containing part of copulatory apparatus, mounted on one of the slides, were re-stained by Mallory’s trichrome method to yield clear contrast between the muscular and connective tissues. The cover glass was removed by steeping the preparation in xylene for 24 h. The sections on the slide were hydrated in an ethanol series. HE staining was then removed by washing in 50% ethanol containing 0.5% HCl for 2 h. After Mallory’s staining, the sections were likewise embedded in Entellan New.

Total DNA was extracted by using Boom et al.’s (1990) silica method. A fragment of the cytochrome c oxidase subunit I (COI) (712 bp) was amplified with primers Acotylea_COI_F and Acotylea_COI_R (Oya and Kajihara 2017) as a reference for DNA barcoding. A 1004-bp fragment of 28S rDNA was amplified with primers fw1 and rev2 (Sonnenberg et al. 2007) for molecular phylogenetic analyses; the primer pairs have been used in other phylogenetic studies of polyclads (e.g. Bahia et al. 2017; Litvaitis et al. 2019). Polymerase chain reaction (PCR) amplification conditions were 94 °C for 5 min; 35 cycles of 94 °C for 30 s, 50 °C (COI) or 52.5 °C (28S rDNA) for 30 s, and 72 °C for 1.5 min (COI) or 2 min (28S rDNA); and 72 °C for 7 min. All nucleotide sequences were determined by direct sequencing with a BigDye Terminator Kit ver. 3.1 and 3730 Genetic Analyzer (Life Technologies, California, USA); two internal primers, fw2 and rev4 (Sonnenberg et al. 2007), were used in sequencing 28S rDNA. Sequences were checked and edited using MEGA ver. 5.2 (Tamura et al. 2011).

Additional 28S rDNA sequences of Acotylea were downloaded from GenBank. Two cotylean species were chosen as outgroups (Table 1): Cestoplana rubrocincta (Grube, 1840) and Pericelis orbicularis (Schmarda, 1859), the former was transferred to Cotylea by Bahia et al. (2017). Sequences were aligned using MAFFT ver. 7 (Katoh and Standley 2013), with the FFT-NS-i strategy selected by the “Auto” option. Ambiguous sites were removed with Gblocks ver. 0.91b (Castresana 2002) using a less stringent option. The optimal substitution models selected with Kakusan4 (Tanabe 2011) under the Akaike Information Criterion (AIC) (Akaike 1974) were GTR+G.

Phylogenetic analyses were performed with maximum-likelihood (ML) methods and Bayesian Inference (BI). The ML analysis was performed with RAxML ver. 8.2.3 (Stamatakis 2014). Nodal support within the ML tree was assessed by analyses of 1,000 bootstrap pseudoreplicates (Felsenstein 1985). BI was performed with MrBayes ver. 3.2.2 (Ronquist and Huelsenbeck 2003). The Markov chain Monte Carlo (MCMC) process used random starting trees and involved four chains for 1,000,000 generations. The first 25% of the trees were discarded as burn-in.

Type slides have been deposited in the Invertebrate Collection of the Hokkaido University Museum, Sapporo, Japan (ICHUM). The sequences determined in this study have been deposited in DDBJ/EMBL/GenBank databases with the accession numbers LC466999 (COI) and LC467000 (28S rDNA).

In this paper, we adopted Prudhoe’s (1985) – instead of Faubel’s (1983) – classification system as to the infrafamilial classification of Plehniidae because this system was followed by some of the subsequent researchers (e.g. Hagiya 1993; Newman and Cannon 1997). Faubel (1983) did not accept Paraplehnia because he considered that “the presence (P. japonica) or the absence (P. pacifica) of Lang’s vesicle demands a separation of both these species” (Faubel 1983: 54–55) and classified Paraplehnia pacifica as a Diplehnia, which was characterized by lacking a Lang’s vesicle (Faubel 1983). However, Kato (1939: 68, fig. 3) clearly stated that the “Lang’s vesicle is small and irregularly elongated, disposed immediately behind the vagina bulbosa in the ventral part of the body” in the original description of Paraplehnia pacifica and also included a line drawing of the Lang’s vesicle as a schematic figure of the copulatory apparatus of Paraplehnia pacifica. The validity of Diplehnia should be tested by future molecular studies along with Diplehnia caeca (Hyman, 1953), the type species of the genus.

Our 28S rDNA analyses corroborate the taxonomic views by Faubel (1983) and Prudhoe (1985) in that Plehniidae, as represented by Paraplehnia seisuiae sp. nov. in this study, should be placed in Stylochoidea (Fig. 3), rather than in Cryptoceloidea as Bahia et al. (2017) suggested. Faubel (1983) and Prudhoe (1985) placed Plehniidae in Stylochoidea based primarily on the reproductive-system morphology and the arrangement of eyespots, respectively. Bahia et al. (2017) carried out a 28S-rDNA-based molecular phylogenetic analysis covering 19 families and 32 genera of polyclads. Based on the analysis, Bahia et al. (2017: 674) circumscribed Cryptoceloidea as having “oval to elongated body, without tentacles, and with cerebral, nuchal, and marginal eyespots” and Stylochoidea as having “rounded body, nuchal tentacles, and cerebral, nuchal, and sometimes marginal eyespots”, among other super-familial redefinitions. Bahia et al. (2017: 675) stated that “Polyposthiidae and Plehniidae possibly belong to Cryptoceloidea”, probably because Plehniidae had been defined as having no tentacles (Bock 1913); our new species, Paraplehnia seisuiae sp. nov., also lacks tentacles. Litvaitis et al. (2019) inferred the internal relationships of Polycladida using 28S rDNA sequences representing 22 families and 37 genera, and identified morphological characters for each clade recovered. Although Stylochoidea was found to be monophyletic, Litvaitis et al. (2019) concluded that this superfamily cannot be defined by any morphological or developmental synapomorphy. It was because in Litvaitis et al.’s (2019) analyses, Stylochoidea turned out to contain members that lack tentacles (e.g., Latocestidae) and have elongated body (e.g., Latocestidae and Leptostylochus), which do not fit to Bahia et al.’s (2017) circumscription for this superfamily. In our analysis, Stylochoidea was “split” into two clades (Fig. 3), and our new species Paraplehnia seisuiae sp. nov., having no tentacles, appeared in one of the two stylochoid clades along with Stylochus sp. of Bahia et al. (2017). If we suppose that Paraplehnia seisuiae sp. nov. is more closely related to Plehnia arctica (Plehn, 1896) (originally in Acelis; type species of Plehnia, which in turn is the type genus for Plehniidae) than any other type species of the type genera of all the nominal families potentially belonging to Stylochoidea, our new species should belong to Plehniidae. If so, and given that Bahia et al.’s (2017) identification of Stylochus sp. (see Molecular phylogeny above) was correct, Plehniidae should belong to Stylochoidea. Our study corroborates the opinion of Litvaitis et al. (2019) in that at least the presence/absence of tentacles is not appropriate to circumscribe Stylochoidea.

The inter-familial relation of Plehniidae was not resolved in this study. It is probably due to the shortness of the 28S rDNA sequence (603 bp) employed in the analyses. Future studies should be done with additional molecular makers and more extensive taxon sampling.

We thank the captain and crew of TRV Seisui-maru, and all the participants of the research cruise No. 1722 for help in collecting the specimen. YO is thankful to Dr Naoto Jimi (National Institute of Polar Research) and Mr Akito Ogawa (The University of Tokyo) for inviting YO to this cruise; Dr Juliana Bahia (Ludwig-Maximilians-Universität München) for providing the sequence data. This study was funded by Research Institute of Marine Invertebrates (No. 2017 IKU-3) for YO.