, 1853 ( Gastropoda , Cassidae ) from the Philippines identified by molecular data

The genus Casmaria H. Adams & A. Adams, 1853 (family Cassidae) is widespread in the tropical Indo-Pacific and has been documented from some Atlantic localities as well. Two Casmaria species, C. erinaceus (Linnaeus, 1758) and C. ponderosa (Gmelin, 1791), are common in Indo-Pacific shallow-water sandy bottom communities and are characterized by high morphological variability; both species encompass multiple, often sympatric forms of uncertain status. In the present study we carry out a phylogenetic analysis of some Philippine Casmaria morphs and demonstrate that one of the distinctive morphs earlier assigned to Casmaria ponderosa is in fact a different species, which we describe as Casmaria boblehmani sp. nov. The smooth form of Casmaria ponderosa, C. ponderosa ponderosa, and the solid nodulose form, widely called “form nodulosa” despite being strikingly different in shell morphology, are shown to be conspecific. Studied specimens of these two morphs even from different localities share the same haplotype of the CO1 gene. In light of these new data on the morphological variability of Casmaria species, we discuss criteria of species delimitation in the genus Casmaria and possible affinities of Casmaria boblehmani sp. nov. within the genus.

the complete lack of shell sculpture in many species. The genus is widely distributed in shallow-water marine communities of the tropical zone. Its members are abundant on sandy bottoms of the Indo-Pacifi c from Madagascar to Hawaii and eastern French Polynesia, and are also known from some Atlantic localities, mainly in the Caribbean province (Clench 1944;Abbott 1968).
Despite the fact that the number of species and subspecies currently attributed to the genus Casmaria hardly exceeds a dozen (Kreipl 1997;Buijse et al. 2013;WoRMS (www.marinespecies.org, accessed 10 Feb 2014)), the taxonomy of the genus remains problematic. Traditionally, diverse Indo-Pacifi c Casmaria forms were lumped into two polymorphic and extremely widespread species, C. erinaceus (Linnaeus, 1758) and C. ponderosa (Gmelin, 1791), consistent with the general wide perception of species in the group established by Abbott (1968). In these two species a number of morphologically distinct forms were named that occur in restricted localities, mostly from the periphery of Indo-Pacifi c; recently, most of these were given the status of full species (Buijse et al. 2013).
In the central Indo-Pacifi c both Casmaria erinaceus and C. ponderosa are represented by several sympatric forms, some of them strikingly different in appearance. Firstly, each of the two Casmaria species encompasses a smooth, usually thin-shelled form (C. erinaceus vibex and C. ponderosa ponderosa) and a more robust form, bearing nodules around the periphery (C. erinaceus erinaceus and C. ponderosa nodulosa) (Kreipl 2008;Beu et al. 2012). Moreover, specimens of both C. erinaceus and C. ponderosa demonstrate signifi cant variation in color pattern. The status of these sympatric forms of C. erinaceus and C. ponderosa is uncertain, and the conspecifi city of the smooth and nodulose forms in both species remains unproven.
In the present study we carry out a molecular-phylogenetic analysis of a set of sympatric Casmaria forms from the central Indo-Pacifi c, mainly from the Central Philippines, in order to establish species limits in Indo-Pacifi c Casmaria species.

Material
Material for molecular studies was collected by night diving in mid-June 2012 at Danajon Banks, off Olango Island, Cebu Province, in the Central Philippines (seven specimens) and in September 2012 at Murray Beach, off Mun Island, Nha Trang Bay, Vietnam (three specimens). Using conventional taxonomy, nine of the specimens were initially assigned to Casmaria ponderosa and one specimen to Casmaria erinaceus. The nine specimens of Casmaria ponderosa represented both the smooth form, known as C. ponderosa ponderosa, and a solid morphotype widely identifi ed as "C. ponderosa nodulosa". They also demonstrated a considerable variation in color pattern.
In the fi eld a fragment of foot tissue was clipped from each specimen and preserved in 99% ethanol for subsequent DNA extraction.

Preparation of genomic DNA and amplifi cation of selected genetic markers
Genomic DNA was extracted from foot tissue samples using the Gentra PUREGENE DNA Isolation Kit (Gentra Systems, Minneapolis, MN) according to the manufacturer's standard protocol. The 12S, 16S and CO1 fragments were amplifi ed using genomic DNA as a template for polymerase chain reaction (for details of the PCR reactions, see Fedosov et al. (2011)). The resulting PCR products were purifi ed by gel electrophoresis and the DNA fragments recovered from agarose.

Characterization of clones encoding 12S, 16S and CO1 sequences
The purifi ed DNA fragments were annealed to the pNEB206A vector using the USER Friendly Cloning kit (New England BioLabs, Inc., Beverly, MA) following the manufacturer's suggested protocol, and the resulting product transformed into competent cells. The nucleic acid sequences of the resulting 12S, 16S and CO1-encoding clones were determined according to the standard protocol for automated sequencing at the Health Sciences Center Core Sequencing Facility, University of Utah. Sequences of six individual 12S, 16S and CO1-encoding clones from each specimen were compared to ensure sequence consensus.
CO1 sequences of an additional three Casmaria specimens collected during recent expeditions of the Muséum National d'Histoire Naturelle, Paris (MNHN) to the Philippines (PANGLAO 2004 -one specimen) and Vanuatu (SANTO 2006 -two specimens) were contributed by Philippe Bouchet and Nicolas Puillandre (MNHN). For sequence GenBank accession numbers and detailed locality data see Table 1.

Phylogenetic analysis
CO1, 12S and 16S sequences of three caenogastropod species, Distorsio reticularis (Linnaeus, 1758) (Tonnoidea: Personidae), Linatella caudata (Gmelin, 1791) (Tonnoidea: Ranellidae) and Erronea errones (Linnaeus, 1758) (Cypraeidae), accessed in GenBank were used as outgroups. Semicassis pyrum (Lamarck, 1822) (Cassidae) was used as a within family outgroup in CO1 analysis. Sequences were aligned for each gene independently using MUSCLE (Edgar 2004). The accuracy of automatic alignments was then confi rmed by eye using BioEdit version 7.0.0.0 (Hall 1999). Hypervariable regions of the 12S and 16S rRNA genes were excluded from further analyses. Parameters of the substitution model were estimated during the analysis (six substitution categories, a gamma-distributed rate variation across sites approximated in four discrete categories and a proportion of invariable sites) independently for each codon position of the CO1 gene. Bayesian Analysis was performed in two parallel runs in MrBayes (Huelsenbeck et al. 2001), each consisting of six Markov chains of 10,000,000 generations with a sampling frequency of one tree each 10,000 generations and with the chain temperature set to 0.02. When log-likelihood scores were found to stabilize, consensus trees were calculated after omitting the fi rst 25% of trees as burn in.
Abbreviations used in text ANSP = Academy of Natural Sciences, Philadelphia, USA MNHN = Muséum National d'Histoire Naturelle, Paris, France

Phylogenetic analysis
Phylogenetic trees constructed after analysis of three molecular markers ( Fig. 1 A-C) are congruent and are characterized by high support of terminal groupings (PP values indicated in bold). The monophyly of Casmaria is supported in analyses of all three markers, although this result requires further confi rmation after the analysis of a more representative dataset. In all three single-gene trees the studied specimens of the genus Casmaria form three distinct groups; the within-group K2P genetic distance computed for the CO1 gene averaged 0.53%, and was always lower than the minimal between-group K2P genetic distance (6.3%). This difference corresponds to the so-called "barcode gap" and suggests a specifi c status for each of the groups (Puillandre et al. 2012).
The fi rst grouping, represented by two specimens on the CO1-based tree and a single specimen on the 12S and 16S trees, corresponds to Casmaria erinaceus. Specimens preliminarily identifi ed as Casmaria ponderosa formed two separate, well-supported clades on the phylogenetic trees. One of these clades, including four smooth Philippine specimens and three nodulose specimens from Vietnam, corresponds to Casmaria ponderosa. Smooth specimens represent the typical form, Casmaria ponderosa ponderosa and correspond to the type material of this species (Buijse et al. 2013). It is noteworthy that four specimens of this grouping, two smooth Philippine specimens (Casm.4, Casm.8) and two nodulose Vietnamese specimens (Casm.V2, Casm.V3), had identical CO1 sequences. The four specimens forming the second clade, two nodulose Philippine specimens and two additional specimens from Vanuatu on the CO1-based tree, represent a distinct morphotype, morphologically similar to so-called C. ponderosa nodulosa, but differing in details of shell structure and coloration (see below), which were previously widely attributed to Casmaria ponderosa. Our results recognize this morphotype as an undoubted distinct species separate from both C. erinaceus and C. ponderosa; its description is given below.

Note on the synonymy and relevant types of Casmaria ponderosa
Casmaria ponderosa has a long and complex synonymy, which has recently been comprehensively analyzed by Beu et al. (2012). All relevant type specimens were tracked down or proven to be lost. The syntype of Cassis torquata Reeve, 1848, illustrated earlier by Abbott (1968), was designated as the lectotype and also as the neotype of Buccinum ponderosum Gmelin, 1791, B. nodulosum Gmelin, 1791, Phalium quadratum Link, 1807, Buccinum biarmatum Dillwyn, 1817, Buccinum pantherina Dillwyn, 1817, and Cassis tenuilabris Menke, 1828 (Beu et al. 2012: 78). For this reason, none of these names can apply to the new species we describe.

Etymology
This species is named to honor I. Robert (Bob) Lehman, a great scientist, mentor and teacher. One of the authors (BMO) has been the grateful direct benefi ciary of his remarkable generosity.

Description
Adult shell small to medium-sized, solid, glossy, ovate-elongate with very prominent terminal varix, studied specimens 34.8-36.3 mm high. Proportions of shell vary considerably, ratio of shell breadth to height 0.58-0.7. Protoconch of about 4 smooth, evenly convex whorls. Teleoconch of 4-5 whorls; early spire whorls evenly convex, with no sculpture. Last adult whorl and in most specimens penultimate whorl with row of glossy nodules (6 to 12 on last whorl), weakly developed in some specimens ( Fig.  2D-E), prominent in others (Fig. 2 G). Due to nodules, late whorls distinctly shouldered; shoulder situated at mid-height of penultimate whorl and adapical quarter of last adult whorl. Aperture ovate, outer lip strongly thickened, drawn upwards, with wide, shallow sinus in upper side. Interior of outer lip with distinct dentition. Apertural varix very prominent, thick, with ten brownish-black blotches on abapertural face; adapertural face with seven or eight well-developed spines directed transversally to apertural plane. Their positions correspond approximately with black blotches on abapertural face. Columellar shield solid, white, heavily callused, with numerous spiral wrinkles in some specimens. Last adult whorl of pinkish, tan, or pale brown background color, with indistinct brown spiral stripes and a row of brown to blackish-brown spots below suture. Siphonal canal with black blotch. Interior of aperture pale brown. Tip of protoconch brown.  Table 2. Summary of morphological differences among Casmaria species. Abbreviations: OLouter lip; TVterminal varix; n.f.nodulose form; rrarely. Number / direction of spines on OL: fforward; trtransverse to apertural plane. Position of spines on OL: a 1/3anterior third. Number of blotches on TV: N+n -N major blotches + n strokes; N senumber of subequal blotches.

Remarks
Specimens that are most appropriately assigned to Casmaria boblehmani sp. nov. are found at sites in eastern Indonesia, northern Australia (Northern Territory and Queensland), Vanuatu and the South Pacifi c. The largest specimen (51.8 mm) that we can confi dently assign to C. boblehmani sp. nov. is known from the Philippines (Buijse et al. 2013: pl. 6, fi g. 13). In the Philippines C. boblehmani sp. nov. occurs syntopically with C. ponderosa, C. erinaceus and C. turgida. All four species co-occur in sandy patches at depths of 1 to 30-40 meters.
Characters allowing the unequivocal recognition of Casmaria boblehmani sp. nov. and the delimitation of the Casmaria species described to date are summarized in Table 2.
Casmaria ponderosa and C. boblehmani sp. nov. can be distinguished from each other by the number and arrangement of brown blotches on the dorsal side (abapertural face) of the terminal varix: C. ponderosa has fi ve major blotches and two to three intermediate strokes, while C. boblehmani sp. nov. has nine to ten blotches of equal or subequal size, evenly distributed along the varix of most specimens. Another character, although less stable, is a row of spots on the shell base, posterior to the siphonal canal. These are always distinct in C. ponderosa, but are missing in most specimens, or in rare cases present but indistinct, in C. boblehmani sp. nov.
Specimens of Casmaria erinaceus can easily be distinguished from both C. ponderosa and C. boblehmani sp. nov. by the absence of dentition on the inner surface of the outer lip and by the presence of only three or four (very rarely fi ve) spines restricted to the anterior edge of the outer lip, directed forward. The anterior edge of the outer lip is drawn slightly forward and abaxially. Both C. ponderosa and C. boblehmani sp. nov. have distinct dentition on the inner surface of the outer lip and have six or more spines directed transversely to the apertural plane, located along the anterior and medial edges of the outer lip. Most of these characters were mentioned by Abbott (1968) and Buijse et al. (2013) as allowing discrimination between C. erinaceus and C. ponderosa.
Casmaria boblehmani sp. nov. can be readily distinguished from Casmaria turgida (Reeve, 1848). The latter species is rather conservative in morphology throughout its range (Buijse et al. 2013) and has a shell with evenly convex whorls, notable because of the characteristic zigzag color pattern. Besides that, C. turgida exceeds all co-occurring Casmaria forms in size. Unfortunately, C. turgida is missing in our molecular dataset. However, the conservative morphology of this species and its sympatry on a small scale with C. erinaceus, C. ponderosa and C. boblehmani sp. nov. at a number of Indo-Pacifi c localities prove that it is a distinct species.
Despite the fact that C. ponderosa, C. boblehmani and C. turgida in most cases can be distinguished from each other confi dently by the combination of characters discussed above, these characters may intergrade to a certain extent, creating somewhat intermediate forms that may be diffi cult to interpret. Buijse et al. (2013: pl. 6, fi gs 6, 8, pl. 9, fi g. 15) depicted a few specimens of intermediate appearance that we are not certain where to assign. All these specimens have smooth shells intermediate between C. ponderosa and C. turgida in size (48-51 mm).
Among the other Casmaria species not occurring sympatrically with C. boblehmani sp. nov., C. unicolor (Pallary, 1926) is probably the most similar. As in other Casmaria species, C. unicolor varies signifi cantly in sculpture and encompasses both smooth and nodulose forms, and the nodulose form closely resembles C. boblehmani sp. nov. (see Buijse et al. 2013: pl. 22). These two forms seem to differ in the number of spines on the outer lip. C. boblehmani sp. nov. possesses seven to eight spines, while C. unicolor has 10-12, and they are located along the entire length of the lip in many specimens, reaching the adapical angle of the aperture. Apart from that, C. unicolor can be distinguished by its more infl ated upper (adapical) whorl profi le and more clearly distinguished, weakly fl attened sutural ramp, giving it a wider, more stepped spire outline. Casmaria unicolor is restricted to the Red Sea (Fig. 3), and the widely distant distributions of this species and C. boblehmani sp. nov. suggest that these two forms are suffi ciently divergent to be considered as separate species.
The three species Casmaria cernica (G. B. Sowerby III, 1888), C. perryi (Iredale, 1912) and C. atlantica Clench, 1944 are extremely similar to each other. They are characterized by being narrower in outline and with a narrower terminal varix than C. boblehmani sp. nov. Besides that, all three mentioned species share the same arrangement of dark blotches on the dorsal side of the apertural varix as in C. ponderosa: fi ve major blotches with one or two intermediate strokes between each major pair.
In comparison with Casmaria boblehmani sp. nov., C. kalosmodix (Melvill, 1883) has a much larger, far more elongate and more narrowly varicate shell with narrow axial color lines that are not present in other Casmaria species, and with no spines on the outer lip. As in C. boblehmani sp. nov., the blotches on the abapertural face of the varix in C. kalosmodix are narrow, but in the latter species they are grouped in 3-4 clusters, separated by wide interspaces.
Another species, Casmaria kayae Buijse, Dekker & Verbinnen, 2013, in general resembles C. erinaceus. C. boblehmani sp. nov. can easily be distinguished from C. kayae by the number and position of spines on the outer lip and by the color pattern. C. kayae has three or four (fi ve in a few specimens) spines on the anterior portion of the lip, and these are directed forward. Many specimens have a distinctive zigzag color pattern, quite different from the usual coloration of C. boblehmani sp. nov.
Finally, C. beui Buijse, Dekker & Verbinnen, 2013, from Western Indian Ocean, can easily be distinguished from C. boblehmani sp. nov. by its small size, its smooth, thin, relatively very wide shell with distinctively infl ated whorls, and by its characteristic color pattern, resembling that of C. turgida.

Distribution
Casmaria boblehmani sp. nov. is widely distributed across the Pacifi c Ocean: from the Philippines, eastern Indonesia, northern Australia, to Fiji and the Society Islands, and (based on specimens illustrated by Buijse et al. 2013) to Hawaii (Fig. 3). Possibly reaches India (based on the specimen illustrated by Robin 2008).

Discussion
Traditionally, characters of shell shape and sculpture are given more taxonomic weight than coloration, especially in groups with highly variable shells. In particular, in the genus Casmaria the presence or absence of nodules on the shoulder dramatically changes the appearance of the shell; ever since the original description by Linnaeus this has confused the species-level taxonomy of the genus. Our data demonstrate that smooth (typical) and nodulose specimens of C. ponderosa, even from distant localities, may share the same haplotype of the CO1 gene, suggesting that (1) these forms are conspecifi c, and therefore (2) the presence or absence of nodules is not a character of taxonomic signifi cance. This is the fi rst evidence independent of traditional conchological data that confi rms the conspecifi city of distinct morphotypes of one cassid species. The nodulose form of C. erinaceus was missing from our dataset, but we expect that it will prove to be conspecifi c with the smooth form, as with the two forms of C. ponderosa. In general, our results support the hypothesis of species delimitation in Cassidae largely based on the assumption that high within-species morphological variability underlies the rise of multiple morphotypes, while these demonstrate no genetic structure.
Identifi cation of a new species morphologically close to Casmaria ponderosa urged us to review the morphology of Indo-Pacifi c Casmaria species in search of morphological traits that would be consistent with our molecular data, allowing the unmistakable delimitation of Indo-Pacifi c Casmaria species or, at least, of the forms that co-occur in the Philippines. While C. erinaceus and C. kayae can be easily separated based on the number and arrangement of spines on the outer lip, delimitation of the other species is less straightforward and is largely based on shell proportions and coloration. In particular, the number and arrangement of the dark blotches on the abapertural face of the terminal varix is consistent with our hypothesis of species delimitation in Casmaria and seems to be conservative enough to serve as a good diagnostic character. However, although these characters work well in most cases, they vary from individual to individual and can intergrade, giving rise to rare specimens of an intermediate appearance.
Interpretation of these intermediate forms requires further molecular studies.
The morphological characters shown here to be taxonomically signifi cant were largely ignored in prior taxonomic diagnoses of the group, and other overlooked Casmaria species may well exist, distinguishable by more subtle details of shell coloration and morphology and detectable by molecular analysis. Further larger-scale molecular phylogenetic studies on Casmaria and closely related cassid genera are necessary to reveal the hidden diversity of this group.