Saved by the Shell: Molecular Analysis Detects the Cryptic Sea Hare, Aplysia concava G. B. Sowerby I, 1833 (Mollusca: Heterobranchia: Aplysiidae), from Oceania, with a Redescription

: A recent taxonomic revision split the circumglobal sea hare Aplysia parvula into 10 con-stituent taxa, of which only three are likely to be found in the Southern Paciﬁc. This prompted an investigation of animals previously identiﬁed as A. parvula from Australia. Specimens collected from Eastern Australia and Hunter Island, east of New Caledonia, could not be satisfactorily identiﬁed with any of the currently accepted taxa based on morphological diagnostic features listed in the revision; however, the presence of a highly concave shell is diagnostic. Quantiﬁcation of genetic divergence using Cytochrome Oxidase I (COI) supports the delineation of this species as a distinct taxon, and a phylogenetic reconstruction based on concatenated COI, 16S and H3 markers reveals a sister relationship with the newly described Aplysia ghanimii from the Atlantic and Western Indian Oceans and an undescribed species from Japan. As a result, the name Aplysia concava G. B. Sowerby, I, 1833 is resurrected for this species. As the original description was based solely on a shell, a redescription is provided here with photographs of living animals and microscope images of internal anatomical structures.


Introduction
Many sea slug species previously thought to have wide, almost circumglobal distributions have been revised based on contemporary species delimitation methods using molecular data (e.g., Valdés et al. [1]; Krug et al. [2]). In most cases, these taxa were found to be complexes of cryptic species with allopatric distributions. A recent revision of the putatively circumtropical Aplysia parvula sensu lato Mörch, 1863 by Golestani et al. [3] split this species into ten separate taxa in four distinct genetic lineages. In doing so, the name Aplysia parvula sensu stricto was retained for an Atlantic taxon, four names were resurrected for species from the Pacific and Indian Oceans, two species were described for the first time and two remain undescribed [3]. Some of these taxa are short-range endemics, e.g., Aplysia elongata (Pease, 1860), which is only known from the Hawaiian Islands, or Aplysia japonica G. B. Sowerby II, 1869 from Japan and the Korean Peninsula, whereas Aplysia ghanimii Golestani, Crocetta, Padula, Camacho, Langeneck, Poursanidis, Pola, Yokeş, Cervera, Jung, Gosliner, Araya, Hooker, Schrödl and Valdés, 2019 has a broad but disjunct range in the Atlantic and Western Indian Oceans [3].
Golestani et al. [3] provided a list of taxa inquirenda arising from their review of the literature. Many of the available names were based on descriptions of shells that, in most cases, are inadequate for use in contemporary identifications, a problem common among the sea hares [3], [4]. One exception, among the Aplysia parvula s. l., is Aplysia concava Golestani et al. [3] provided a list of taxa inquirenda arising from their review of the literature. Many of the available names were based on descriptions of shells that, in most cases, are inadequate for use in contemporary identifications, a problem common among the sea hares [3], [4]. One exception, among the Aplysia parvula s. l., is Aplysia concava G. B. Sowerby I, 1833, redrawn later by von Clessin [5] and Pilsbry [6], from "Australia", which has a distinctive, deeply concave shell ( Figure 1A-D). It was synonymised with Aplysia parvula s. str. in the 20th century [7]. Aplysia concava was not recognised amongst the samples studied by Golestani et al. [3]; however, this is probably the result of opportunistic geographic sampling. Indeed, samples from Australia or New Zealand were not included in their analysis, although material from nearby Papua New Guinea, Vanuatu and Indonesia was. Eastern Australia, particularly the Tasman Sea, is a globally important hotspot for sea hare diversity [4], with 16 species reported from Central New South Wales (NSW) alone, including animals previously identified as Aplysia parvula s. l. Of the species described or re-erected by Golestani et al. [3], three are putatively found in the Southern Indo-Pacific: A. atromarginata Bergh, 1905, A. nigrocincta Von Martens, 1880 and an undescribed species, Aplysia sp. 2, in Golestani et al. [3]. Although defining characteristics (molecular, morphological and geographic) were provided for each of these species, some Eastern Australian animals cannot be satisfactorily identified as being any of these three species. Thus, in light of the cryptic diversity now confirmed to exist among A. parvula s. l., it is probable that there remains undiscovered or previously unrecognised diversity among Australian animals. Nevertheless, some specimens of Aplysia parvula s. l. from Eastern Australia, while not able to be reconciled to any of the newly defined taxa, have dis- Eastern Australia, particularly the Tasman Sea, is a globally important hotspot for sea hare diversity [4], with 16 species reported from Central New South Wales (NSW) alone, including animals previously identified as Aplysia parvula s. l. Of the species described or re-erected by Golestani et al. [3], three are putatively found in the Southern Indo-Pacific: A. atromarginata Bergh, 1905, A. nigrocincta Von Martens, 1880 and an undescribed species, Aplysia sp. 2, in Golestani et al. [3]. Although defining characteristics (molecular, morphological and geographic) were provided for each of these species, some Eastern Australian animals cannot be satisfactorily identified as being any of these three species. Thus, in light of the cryptic diversity now confirmed to exist among A. parvula s. l., it is probable that there remains undiscovered or previously unrecognised diversity among Australian animals. Nevertheless, some specimens of Aplysia parvula s. l. from Eastern Australia, while not able to be reconciled to any of the newly defined taxa, have distinctive, highly concave shells, resembling that illustrated for Aplysia concava G. B. Sowerby I, 1833. Although A. concava was reported from Cloudy Bay Lagoon, Bruni (sic) Island, Tasmania by Tenison-Woods [8], who described the shell as "small, horny, very concave and strongly incurved, sub-auriculate on both sides of the apex, not very common" there have been few references to this species in the literature since.
The aim of this study is to examine and redescribe Aplysia concava specimens using modern imaging techniques and to explore its phylogenetic relationship with other species among Aplysia parvula s. l. using data from partial Cytochrome Oxidase I (COI), 16S ribosomal DNA (16S) and Histone 3 (H3) gene sequences. Several specimens exhibiting the species' characteristic shell were sampled from Eastern Australia, along with museum specimens identified using molecular analysis. The present paper provides clarification of the work of G. B. Sowerby I by including details not provided in the original description.

Source of Material
Eleven specimens from Eastern Australian and one from Hunter Island (east of New Caledonia) were used in this study. These comprised tissue from seven specimens sourced from the collection of the Australian Museum, Sydney (AMS), and four whole specimens collected from various locations in northern NSW by MN (Table 1; Figure 6).

Molecular Study
DNA was extracted from~10 mg of foot tissue from each ethanol-preserved specimen using a Qiagen DNeasy 96 Blood and Tissue DNA extraction kit and the supplied protocol. Partial COI sequences were amplified using degenerate primers [9] (jgLCO1490 5 -TITCIACIAAYCAYAARGAYATTGG-3 , jgHCO2198 5 -TAIACYTCIGGRTGICCRAARAAY CA-3 ). Universal primers were used to amplify 16S rDNA [10] (16SarL 5 -CGCCTGTTTAA CAAAAACAT-3 , 16SbrH 5 -CCGGTCTGAACTCAGATCACGT-3 ); and universal primers were used to amplify H3 sequences [11] (H3AF 5 -ATGGCTCGTACCAAGCAGACVGC-3 . H3AR 5 -ATATCCTTRGGCATRATRGTGAC-3 ). For each of these, a master mix was prepared using 5.0 µL of MyTaq X 5 PCR buffer, 0.2 µL MyTaq, 0.8 µL primer 1, 0.8 µL primer 2, 16.8 µL molecular-grade H 2 O and 1.5 µL of extracted DNA. Polymerase chain reaction (PCR) cycling was carried out using a Bio-Rad T100 thermal cycler. Thermo cycler conditions for PCR amplification of the COI marker were as follows: initial denaturation at 95 • C for 3 min, 8 cycles of denaturation at 95 • C for 30 s, annealing at 50 • C for 30 s and extension at 72 • C for 45 s with another 32 cycles of 95 • C for 30 s, annealing at 48 • C for 30 s and extension at 72 • C for 45 s. Finally, elongation was carried out at 72 • C for 5 min. For the 16S and H3 markers, PCR conditions were as follows: initial denaturation at 95 • C for 5 min, 35 cycles of further denaturation at 95 • C for 30 s, annealing at 50 • C for 30 s and extension at 72 • C for 45 s, with a final elongation at 72 • C for 10 min. Visualisation of PCR products was carried out using an Invitrogen E-gel 96, 2% agarose plate to determine the approximate base pair lengths (bp) of COI (658 bp), 16S (422 bp) and H3 (328 bp). PCR products were sequenced at the Australian Genomic Research Facility (AGRF), Perth, WA.
Sequences were de novo assembled using Geneious 11.1.5 [12] and edited by eye. Consensus sequences were then exported and combined with additional sequence data retrieved from GenBank. This includes data from Golestani et al. [3], where COI, 16S and H3 sequences were available for each specimen, and a COI sequence listed as A. parvula but identified using BLAST as a match for Aplysia atromarginata from French Polynesia ( Table 2). Sequences were aligned using the MAFFT plugin [13] using default settings and primers were trimmed from the alignment. Data quality checks were carried out with MEGABLAST [14] and protein translation for COI. An alignment of 16S sequences was uploaded to the online version of GBlocks [15] and processed using relaxed settings to remove ambiguous regions (i.e., containing gaps and/or poorly aligned) [16]. The gene alignments of COI, gblocked 16S and H3 were then concatenated using Geneious.
A phylogenetic tree was produced using W-IQ-Tree [17,18] using default settings and partitioned by gene [19]. W-IQ-Tree incorporates ModelTest [20], which selected HKY + F + G4 as the best-fit model for the first partition, which is scored according to the Bayesian information criterion (BIC) and automatically applied. Trees were visualised using FigTree 1.4.4. Dolabella auricularia (Lightfoot, 1786) was used to root trees for analysis as it is sister to Aplysia [3,21]. Species hypotheses were also explored using COI sequence data. Within-and betweenspecies mean distances were calculated using MEGA V7.0.26 [22] using the Tamura-Nei substitution model [23]. The Automatic Barcode Gap Discovery (ABGD) algorithm was used to examine the potential for new species [24]. This employs a distance-based analysis of the distribution of pairwise distances in the alignment of barcode COI sequences. As recommended by Puillandre et al. [24], a pairwise distance matrix was developed in MEGA 7, which also selects a best-fit evolutionary model, and this matrix was uploaded to the ABGD webserver (without the outgroup) and processed using the default settings. Haplotype networks were constructed using the TCS network algorithm [25] in PopART [26] software.

Morphological Study
Animals were photographed alive either in situ or in aquaria. Preserved specimens were dissected by a ventral or right lateral incision just above the margin of the foot. Internal anatomy was examined using a Leica Zoom 2000 dissecting microscope. The buccal mass was removed and placed in 10% NaOH for four hours to separate the radula. The radula, penis and shell were isolated from the tissues, rinsed in seawater, critical point dried using a graded ethanol series and mounted for imagery using a Hitachi TM4000plus desktop scanning electron microscope (SEM). The reproductive system was examined and photographed using an Olympus S2 × 7 binocular microscope with an Olympus DP26 overhead-mounted digital camera. Field-collected specimens have been deposited with the Australian Museum, Sydney (AMS) and Western Australian Museum (WAM).
Interspecific genetic distances, based on the Tamura-Nei model, delimited Aplysia concava as a distinct taxon from all other known species in the Aplysia parvula complex with a divergence of 2%, whereas the minimum intraspecific divergence values were >4% for all other Aplysia used in this study (Table 3). Delimitation was further supported by a phylogenetic reconstruction based on maximum likelihood using concatenated COI, 16S and H3 markers, which recovered Aplysia concava as a maximally supported, monophyletic clade (approximate Bayes/ultrafast bootstrap value = 1/100), sister to Aplysia ghanimii + Aplysia sp., which comprise Golestani et al. [3] Species Complex 2 ( Figure 2).
Taxonomy 2021, 1, FOR PEER REVIEW 7 and H3 markers, which recovered Aplysia concava as a maximally supported, monophyletic clade (approximate Bayes/ultrafast bootstrap value = 1/100), sister to Aplysia ghanimii + Aplysia sp., which comprise Golestani et al. [3] Species Complex 2 ( Figure 2). A haplotype network reconstructed using COI sequences for A. concava does not show marked geographic structure. Specimens from oceanic islands (Hunter Island, east of New Caledonia and Lord Howe Island, in the Tasman Sea) were separated from other, mainland haplotypes by three and four substitutions, respectively; however, this level of  A haplotype network reconstructed using COI sequences for A. concava does not show marked geographic structure. Specimens from oceanic islands (Hunter Island, east of New Caledonia and Lord Howe Island, in the Tasman Sea) were separated from other, mainland haplotypes by three and four substitutions, respectively; however, this level of divergence is also present among other mainland Eastern Australian specimens (Figure 3).    ( Figures 4 and 5).

External Morphology
Body elongated, with a distinct head and neck. Body widest at visceral region, tapering towards tail and narrowing slightly towards the head and neck. Paired, rolled rhinophores with tapered tips arise on top of the head, posterior to the eye spots. Paired, rolled cephalic tentacles project anterolaterally and may flare along their length, forming a flattened, posteriorly projecting flap. Parapodia do not meet anteriorly, fused high posteriorly, form a fold at their mid-length and do not meet above the mantle. Mantle foramen large, circular, not raised, opening on left side of mantle. Shell visible through foramen. In some, the top of the siphon is level with, in others, it projects above, the parapodia.

Colour Variation
Aplysia concava is highly variable in colour, ranging from dark red to light brown to cream (Figure 4). There is a very fine white parapodial margin and a black submarginal band, which may be discontinuous in light brown to cream animals ( Figure 4B-E). The black band may be broad in dark red animals and the marginal band, grey ( Figure 4A,F). The foot margins, tips of the cephalic tentacles and rhinophores also exhibit a black margin that may also be discontinuous in some specimens ( Figure 4B-E). The mantle foramen has a black margin in dark red animals; however, this is absent in light brown and cream animals. Fine white specks may cluster to form irregular spots all over the body; however, these are absent in dark red animals. The base of the siphon is opaque white in light brown animals.

Internal Morphology
The ovate-auriculate shell is characteristically concave ( Figure 5I-K) and the protoconch is highly inflexed so as to be perpendicular to the rest of the shell ( Figure 5I) and is consistent with that illustrated for A. concava by earlier authors (Figure 1).

Internal Morphology
The ovate-auriculate shell is characteristically concave ( Figure 5I-K) and the protoconch is highly inflexed so as to be perpendicular to the rest of the shell ( Figure 5I) and is consistent with that illustrated for A. concava by earlier authors (Figure 1).

Diagnosis
Aplysia concava can be distinguished from other Aplysia by the presence of a solid or broken black marginal band on the foot, parapodia, rhinophores, cephalic tentacles and by its diminutive adult size. It may be separated from other Southwestern Pacific Aplysia parvula s. l. by the presence of a distinctively concave shell, and, depending on the activity of the living animal, the shell may cause the visceral mass to appear swollen ( Figure 4C). The parapodia are not able to cover the mantle. In brown specimens, there is an opaque white patch at the base of the siphon ( Figure 4B); however, this is absent in dark red specimens. The mantle aperture is large, circular and its margins are not raised.
Although it shares most of the external morphological features of the sister taxa in Golestani et al. [3] Species Complex 2, A. concava is allopatric in its distribution. Aplysia ghanimii is found in the Atlantic and Western Indian Oceans and Aplysia sp. 2 is so far only known in Japan. It can be separated from the potentially sympatric A. parvula s. l. species

Diagnosis
Aplysia concava can be distinguished from other Aplysia by the presence of a solid or broken black marginal band on the foot, parapodia, rhinophores, cephalic tentacles and by its diminutive adult size. It may be separated from other Southwestern Pacific Aplysia parvula s. l. by the presence of a distinctively concave shell, and, depending on the activity of the living animal, the shell may cause the visceral mass to appear swollen ( Figure 4C). The parapodia are not able to cover the mantle. In brown specimens, there is an opaque white patch at the base of the siphon ( Figure 4B); however, this is absent in dark red specimens. The mantle aperture is large, circular and its margins are not raised.
Although it shares most of the external morphological features of the sister taxa in Golestani et al. [3] Species Complex 2, A. concava is allopatric in its distribution. Aplysia ghanimii is found in the Atlantic and Western Indian Oceans and Aplysia sp. 2 is so far only known in Japan. It can be separated from the potentially sympatric A. parvula s. l.
Taxonomy 2021, 1 58 species A. nigrocincta, which has characteristic white or light blue parapodial margins without submarginal black banding, and A. atromarginata, which exhibits a small, raised, "volcano-like" mantle aperture with dark margins and radiating white lines.

Range
Aplysia concava has a confirmed range in the Southwestern Pacific, extending south from New Caledonia to Eastern Australia (from the Queensland and NSW border south to the NSW mid-north coast) and east into the Tasman Sea. Its range may include the northern part of New Zealand (MN pers. obs) and, according to Tennyson Woods [8], it extends as far south as Southern Tasmania (Figure 6).
A. nigrocincta, which has characteristic white or light blue parapodial margins without submarginal black banding, and A. atromarginata, which exhibits a small, raised, "volcanolike" mantle aperture with dark margins and radiating white lines.

Range
Aplysia concava has a confirmed range in the Southwestern Pacific, extending south from New Caledonia to Eastern Australia (from the Queensland and NSW border south to the NSW mid-north coast) and east into the Tasman Sea. Its range may include the northern part of New Zealand (MN pers. obs) and, according to Tennyson Woods [8], it extends as far south as Southern Tasmania ( Figure 6).

Data Availability Statement:
The data presented in this study are openly available in GenBank at https://www.ncbi.nlm.nih.gov/nuccore (accessed on 9 April 2021).