Many species of the gastropod genus Philine have been named from northeastern Asia but scanty descriptions based predominantly on shells make it difficult to determine which are valid. This, plus the sporadic anatomical and genetic information available for many of these species has led to what may be described as an un-integrated taxonomy. In this situation, it is generally preferable to postpone dissection of rare and unusual specimens until relevant diagnostic characters can be established in broader studies. Micro-CT scanning and DNA sequencing were used to examine such a specimen collected recently from deep waters off northeastern Taiwan. Micro-CT examination of the morphology of the internal shell and gizzard plates suggested that, among named species, the sequenced specimen is most similar to P. otukai. It cannot, however, be definitively referred to P. otukai as that species lacks adequate anatomical description or known DNA sequences. Phylogenetic analyses of newly collected DNA sequences show the specimen to be most closely related to, but distinct from the northern Atlantic Ocean and Mediterranean species, Philine quadripartita. The sequences also confirm genetically that five or more species of Philine occur in northeast Asia, including at least three subject to considerable taxonomic uncertainty.

Gizzard plates, 16S ribosomal RNA, micro-CT scanning, Scaphopoda

Accurate delimitation of species of the highly speciose cephalaspidean genus Philine Ascanius, 1772 is hampered by the brief original descriptions of many of its named taxa. Many such descriptions were made using only the shell, which has limited diagnostic value, and often these have not been subsequently supplemented by studies of internal anatomy, including morphology of the diagnostically important gizzard plates, or DNA sequences.

The difficulty of delimitation is especially marked in northeastern Asia, as exemplified by the situation in Taiwan from which three named Philine species (P. argentata Gould, 1859; P. vitrea Gould, 1859, and P. otukai Habe, 1946) are currently recognised (Lee 1998; Wu 2004; Wu and Lee 2005). We have recently collected a specimen from deep waters off Taiwan that is not readily identifiable owing to the lack of an adequate taxonomic description of comparable species. A fourth species, Philine kurodai Habe, 1946, has been recorded in Taiwan but the relevant specimens were supposed to belong to P. otukai according to Lee (1998) and Wu (2004). Many other species of Philine have been named from nearby regions, and in total, 15 named recent species with medium to large shells have been recorded (Table 1) from the coasts of China, Korea, and Japan (Zhang 2008; Price et al. 2011; Chaban 2014; Chaban and Chernyshev 2014; Chaban et al. 2019; and references in these articles). The anatomical and genetic information available for most of these species is sporadic, resulting in a taxonomy that may be described as un-integrated, with a general lack of diagnostic characters suitable for species discrimination. In this situation, we consider it advisable to minimise destructive sampling of rare specimens until such diagnostic characters can be established from more frequently collected taxa.

We used non-destructive morphological analysis of the collected specimen and DNA sequencing of a small external tissue sample to make initial comparisons with named species. For morphological analysis, we used micro-CT scanning (Shepelenko et al. 2015) which is potentially very useful for studying specimens where dissection would be premature. We focussed on the shape and sculpture of the shell, which is internal in Philine, and the three gizzard plates which have been applied to morphological diagnoses of the genus (e.g., Price et al. 2011).

We determined sequences of 16S ribosomal RNA (16S rRNA), histone H3 (H3), and the D1 expansion region of 28S ribosomal RNA (D1 28S rRNA). We then performed phylogenetic analyses of the sequences, including sequences from previous DNA studies of the Philinidae (Krug et al. 2012; Ohnheiser and Malaquias 2013; Gonzales and Gosliner 2014; Oskars et al. 2015; Chaban et al. 2019). We were unable to sequence the cytochrome c oxidase subunit I bar-coding region, a difficulty also found by Krug et al. (2012), possibly explaining why there are relatively few GenBank accessions for this gene from Philine.

Morphologically, C.559479 has more apparent affinity with P. otukai Habe, 1946, than other Taiwanese species. If the taxa were conspecific, this would represent a very large increase in the depth range of P. otukai which is reported to extend only to 100 metres (Lin 2004). The definite identification of C.559479 as P. otukai would, however, be premature because of the lack of relevant morphological information for that species. The shell of P. otukai is generally larger than that of C.559479, with the type specimen (Otuka 1936: figs 22, 23, as “P. aperta”) of the former having a length of 24 mm and width of 17 mm (Otuka 1936). Although Lee (1998) gives a minimum of 15 mm (in a range of 15–28 mm) most reports from other regions (e.g., Lin 2004) suggest that the shell is much longer (> 25 mm). The gizzard plates of C.559479 appear similar in shape to those of P. otukai illustrated by Lee (1998) depicting the inner surface. However, no pores are visible in the illustrations of the outer surfaces of the plates in Lee (2018).

Specimen C.559479 is readily distinguished from P. argentata, which has two slit-like recesses on the outer surfaces of its gizzard plates (Chaban and Chernyshev 2014), that are absent here (Fig. 3). Philine argentata and P. orientalis were synonymised by (Price et al. 2011). However, they were treated as distinct species by Chaban (2014) so the possibility that C.559479 belongs to P. orientalis may be considered. The specimen C.559479 is distinct externally from P. orientalis sensu (Price et al. 2011) in its smaller size and creamy yellow headshield colour (vs. white in the latter species). C.559479 has a large pair of circular pores on all gizzard plates, whereas the pores are small in P. orientalis Price et al. (2011).

Specimen C.559479 differs from P. vitrea in the small size of the latter (shell dimensions of the type specimen 10 × 8 × 3 mm: Gould 1859). It also differs in the apparently much lower projection of the upper end of the outer lip of P. vitrea, allowing for the damaged condition of this shell section in the specimen studied here.

Sufficient information is available to distinguish C.559479 morphologically from only a few other regional species of Philine. These include P. scalpta in which the gizzard plates are quasi-trapezoidal with elongated pores (Chaban and Chernyshev 2014). The shell of the shallow water P. kinglipini Tchang, 1934 differs by the absence of spiral sculpture (Lin 1997: 198). Philine kinglipini attains a larger size of ~ 40–42 mm length and 18–19 mm width with shell dimensions of 19–20 mm × 14–14.5 mm (Lin 1997). The gizzard plates of P. kinglipini are all dissimilar to each other in size (Tchang 1934), with the second largest being quasi-trapezoidal. Both arms on all three plates of P. kinglipini are broad and nearly equal in length.

Philine paucipapillata Price, Gosliner & Valdés, 2011 was described as a new species from the South China Sea. Specimen C.559479 is readily distinguished morphologically from this species in coloration and size (P. paucipapillata is uniformly white and ~ 4–5 cm in length). The pores in the gizzard plates of P. paucipapillata are minute (Price et al. 2011) in contrast to those of C.559479.

Philine kurodai Habe, 1946, described from Japan, has been reported from Taiwan although the Taiwanese records are now regarded as being based on specimens of P. otukai (Lee 1998; Wu 2004). However, although both P. otukai and P. kurodai were described in Habe (1946) no characters distinguishing them from each other were listed and we have been unable to find any subsequent discussion discriminating these species. Neither P. kurodai nor P. otukai was mentioned by Price et al. (2011) but both are accepted by Chaban (2016).

Three species from northeastern Asia, P. acutangula Adams, 1862, P. crenata Adams, 1862, and P. coreanica Adams, 1855, have received little recent attention, particularly after Herdman (1906) made the latter two synonyms of “Philine aperta” (Table 1). The species all appear to belong to the group of species described by Adams (1855, 1862) categorised by Kurodai and Habe (1954) as “unobvious” and are presently best regarded as species inquirenda. To the best of our knowledge, the locations of their type specimens are unknown.

Analyses including the DNA sequences of C.559479 collected here highlight the taxonomic complexity of Philine in northeastern Asia. The available 16S rRNA data from the region represent at least five species. One of these species is represented by the C.559479 sequence from Taiwan. The other species are P. paucipapillata, P. argentata sensu Chaban et al. (2019), a species uncertainly referred to “P. orientalis”, P. japonica or P. scalpta – see below), and an additional species, denoted “species D” by Oskars et al. (2015), which is represented by histone H3 DNA sequences. The latter may have affinity with Philine habei Valdés, 2008. The Taiwanese specimen of “species D” lacks 16S rDNA sequences but the other specimens referred to it by Oskars et al. (2015) are found in analyses of this gene (Fig. 4) in a strongly supported clade with the P. habei sequence (KJ511780). This clade clearly excludes C.559479. The gizzard plates of P. habei have large elongate pores (Valdés 2008; Price et al. 2011) that appear very different from those of C.559479.

The three sequences of 16S rRNA that were most similar to C.559479 are recorded in GenBank as belonging to Philine aperta. However, the true identity of the specimens from which these sequences were derived appears to be P. quadripartita, a species from the northern Atlantic Ocean and Mediterranean Sea. The sequences form a distinct lineage separate from the two South African specimens from near the type locality of true P. aperta, and were determined from specimens taken within the range of P. quadripartita: AY345016 derives from a specimen from south-eastern Spain (Grande et al. 2004); and both AJ223402 (Thollesen 1999) and DQ093482 (Aktipis et al. 2008) are from Sweden.

The “P. orientalis” 16S rRNA sequence accession JQ691684) from Japan (Krug et al. (2012) is from a specimen identified as P. japonica in the catalogue of the Natural History Museum, London (BMNH:1996409). This is included in a clade with three Californian (non-native range) haplotypes of “P. orientalis” sequenced by Krug et al. (2012) and another from a specimen identified by Chaban et al. (2019) as P. scalpta, a species that was redescribed by Chaban and Chernyshev (2014) who removed it from the synonymy of P. vitrea in which it had been placed by Higo et al. (1999). If P. japonica and P. scalpta are, as maintained by Chaban et al. (2019), distinct from P. orientalis, then sequences truly from the latter species are probably yet to be determined. It is possible that the sequences from the taxon from near Darwin in northern Australia labelled sp. 1 JV 2012 in GenBank (Krug et al. 2012) may belong to P. orientalis or a closely related species. The named species to which these sequences are most similar are the Australasian Philine angasi (Crosse, 1865, in Crosse and Fischer 1865) and P. paucipapillata. Some radular and penial morphology characters distinguish P. paucipapillata and P. orientalis but they are closely related in the morphologically based phylogeny of Price et al. (2011). In this phylogeny, P. paucipapillata is shown as the sister group of the pair composed of P. orientalis and P. angasi.

The difficulty of identifying the unusual deep water Philine specimen studied here emphasises that destructive sampling should be minimised where the taxonomy is un-integrated. We tentatively suggest from examination of the external morphology and micro-CT scanning that it may have affinities with the shallow water P. otukai but any certainty is precluded by the lack of a description of that species, especially one detailing the range of variation it might encompass in characters such as the size of the gizzard plate pores. There is no diagnosis of P. otukai that would guide specimen dissection for relevant anatomical characters. Considerable further research efforts are required to provide a framework in which such characters could be sought. These include DNA studies of a wider range of regional species of Philine, particularly sequences definitely from P. orientalis, P. kurodai, P. otukai, and P. vitrea. Detailed morphological investigations are needed for P. japonica, P. otukai, P. kurodai, and P. vitrea. The status of species such as P. coreanica may remain unknown but it may possibly be resolved if shell fragments from type material ever become available for macromolecular analysis, especially proteomics.

We thank the Australian Museum for funding and Cameron Slatyer for logistical support. We are grateful to Dr W. Rudman, Dr E. Chaban, Dr T. Gosliner, and anonymous reviewers for comments on earlier versions of the manuscript and to the editor, Dr N. Yonow, for very helpful advice. STA gratefully acknowledges support and hospitality of Prof. T.-Y. Chan in Taiwan in 2015. The authors acknowledge the facilities and the scientific and technical assistance of the National Imaging Facility at the Centre for Advanced Imaging, University of Queensland. We thank Professor K.K. Chan for providing the copy of a reference.