A new troglobite species of Habeastrum Simone, 2019 from Brazil, and support for classification in Diplommatinidae (Mollusca, Caenogastropoda)

The genus Habeastrum Simone, 2019 was recently described based on empty shells, counting with two troglobite species. Conchological features allowed a preliminary classification in the caenogastropod family Diplommatinidae, but this family allocation was left open to future studies. Herein, we present a detailed anatomical study of newly acquired specimens, confirming the classification in Diplommatinidae. These new specimens, from Minas Gerais state, SE Brazil, belong to a new troglobite species described herein, Habeastrum strangei sp. nov. The present records extend the genus distribution ca. 1,100 km east-northeast.


Introduction
The genus Habeastrum Simone, 2019, has been recently described to allocate two unique and previously unknown species collected in caves in Mato Grosso do Sul state, central Brazil. It presently includes two terrestrial troglobite species: the type species H. parafusum Simone, 2019 and H. omphalium Simone, 2019. The original description of the genus was based solely on empty shells (the operculum was also unknown) and the following conchological features allowed a tentative classification in Diplommatinidae (Simone 2019): the minute size, the delicate axial sculpture, and the rounded profile of the shell whorls. However, unlike most New World diplommatinids, the shell of Habeastrum is dextral, with a simple non-deflected aperture, and uniform shell growth (i.e., the whorls increase regularly and slowly in size during growth).
The latter features are unusual for the family (Thiele 1929;Wenz 1938) and Simone (2019) recognized that the familiar attribution of the new genus should be left open for future studies counting with specimens with preserved soft parts.
Such specimens were recently sent to us by two different sources: Prof. Dr. Maria Elina Bichuette, a researcher from the Universidade Federal de São Carlos (UFSCar, Brazil) specialized in Brazilian troglofauna, and the environmental consulting company Carste Ciência e Meio Ambiente (CARSTE, Brazil). These new specimens were collected in caves in various different municipalities in Minas Gerais state, southeastern Brazil. We present here an anatomical study of the specimens, confirming the diplommatinid affinity of the genus. Furthermore, they are deemed to belong to a new species, described herein as Habeastrum strangei sp. nov.

Material and methods
The new material from Minas Gerais contained both whole specimens (preserved in 70% ethanol) and empty shells. It was donated by Dr. Bichuette to the malacological collection of the Museu de Zoologia da Universidade de São Paulo (MZSP, São Paulo, Brazil). Selected specimens (four females and three males) were dissected in ethanol following standard techniques (Simone 2011), aided by a stereomicroscope equipped with a camera lucida and a digital camera. The shell was lightly crushed with the aid of forceps and the soft parts were extracted without being damaged. All dissection steps were photographed and all drawings were made with the help of the camera lucida. The radula was extracted using standard techniques and examined under SEM in the Laboratory of Electronic Microscopy of the MZSP. SEM images of the shell were captured in the Staatliches Museum für Naturkunde Stuttgart (SMNS, Stuttgart, Germany) and of the radula in the MZSP. The list of examined material, as well as collection details, can be found below.
Two specimens were randomly selected for DNA extraction, with two samples acquired from each. Unfortunately, we could not acquire good quality DNA for amplification and sequencing. Even though the specimens seemed fresh, we suspect they might have stayed too long in the field sample before being fixed in ethanol and/or not have been thoroughly fixed: the snails were deeply retracted into their shells and thus, the ethanol might not have reached them properly given the presence of the operculum. After these four trials, using two different extraction kits and protocols, we opted for not spending more specimens.
Diagnosis. Spire conical and tall. Teleoconch sculpture consisting of strongly prosocline, markedly raised ribs, more widely spaced than in congeners.
Radular sac very long, stored as 4-5 compact whorls located to the right of esophagus in the region preceding buccal mass (Fig. 3C, D: rs). Radular nucleus (rn) simple, faintly bulging. Radula (Fig. 2B-E) taenioglossate, with two marginal teeth and one lateral tooth on each side of central rachidian tooth. Each tooth is composed of a central large claw-like denticle and two smaller denticles on each side (the lateral-most being slightly smaller).
Buccal cavity with pair of wide and low dorsal folds as continuation from jaw plates, gradually diminishing towards esophagus. Pair of buccal ganglia (Fig. 3E: bg) of considerable size, rounded, each one located laterally, in transition between buccal mass and esophagus. Esophagus (es) as simple continuation of buccal cavity, relatively narrow; pair of dorsal folds from buccal cavity gradually disappear in anterior esophagus, keeping a simple and smooth surface. Pair of lateral esophageal glands (Fig. 3C-E: eg); right gland ~1/3 haemocoel's length, slightly broader than esophagus, bulging relatively uniformly along right esophageal wall; left esophageal gland ~1/2 of right gland's length, but circa twice its width; both esophageal glands hollow inside, walls and inner surface thick and glandular. Posterior esophagus narrow, simple (lacking glands or chambers) (Figs 3B-E: es), about as long as anterior esophagus; esophageal insertion in stomach small, ventral, ~1/3 of main gastric chamber (Fig. 3B: es). Stomach (Fig. 3B: st) size and location described above; its main chamber very wide and long, occupying most of the whorl posterior to kidney. Main gastric chamber lacking internal fold or subchamber; its posterior end suddenly narrows, running posteriorly as a narrow tube along 1.5 whorl (Fig.  3B: st at right); its posterior end is rounded, unclear if it connects to digestive gland, as it is lying along the gland's middle region. Anterior region of stomach bluntly tapering; intestine originating in rounded anterior gastric end. Intestine (Fig. 3B: in) performing a tight loop anteriorly to stomach, suddenly turning left and posteriorly, performing a wide loop dorsally to anterior gastric region; afterwards intestine runs through renal tissue towards anterior region of pallial cavity. In the pallial cavity, rectum (rt) runs through prostate in males (Fig. 3B) or to the left of visceral oviduct in females (Fig. 4A). Anus (an) long, siphoned (mainly in males), ending posteriorly to mantle border.
Central nervous system (Fig. 4B). Nerve ring hypoathroid, relatively large, occupying ~1/6 of haemocoel ( Fig. 3C: pn, ce). Located just posteriorly to buccal mass. Each cerebral ganglion (ce) spherical, with ~1/4 of nerve ring volume. Cerebral commissure narrow, about as wide as each cerebral ganglion. Each pleural (pl) and pedal (pn) ganglion fused with each other, separated only by shallow ventral furrow; each pleuro-pedal ganglion slightly larger than each cerebral ganglion. Both commissures very narrow, of similar length on both sides, ~1.5 times longer than the width of each ganglion.
Habitat. In the caves in Presidente Olegário municipality, this species was found in epigeal environments and in all cave zones (entrance, twilight, and dark zones). Most specimens collected consisted of empty shells, but the good preservation considering their fragility suggests they were not subjected to transport. No precise information was recorded from other municipalities.

Discussion
As discussed in the original description of Habeastrum (Simone 2019), the following shell features pointed towards a diplommatinid affinity: the minute size, the rounded profile of the whorls, and the delicate axial sculpture ( Fig. 1F-H). On the other hand, a series of charac-ters is uncommon for a diplommatinid (Thiele 1929;Wenz 1938), such as the uniform shell growth (most diplommatinids have uneven growth of the body whorl, sometimes resulting in bizarre shapes), the dextral coiling (most New World diplommatinids are sinistral), the presence of umbilicus (most seal the umbilicus with the last whorl), and the simple peristome (most have expanded, sometimes trumpet-like, peristomes) (Tielecke 1940;Webster et al. 2012). Some genera however, like Nicida W.T. Blanford, 1868 from the Indian subcontinent and Pugnellia Oppenheim, 1895 from the Italian Eocene, contain dextral species with simple peristomes (e.g., Wenz 1938;Raheem et al. 2014). The existence of these features among the conchological variability in Diplommatinidae strengthens the argument for placing Habeastrum within this family. In any event, the set of unique features of Habeastrum amongst the diplommatinids reinforces its generic separation. Webster et al. (2012) suggested that the ancestral state in Diplommatinidae was sinistral coiling, with dextral shells having evolved independently three times. As such, Habeastrum would represent a fourth, South American, instance of reversal to dextrality.
Even though the conchological variation in Diplommatinidae is well described and illustrated in the literature, the same is not true for their anatomical features. Most works have focused on other cyclophoroidean families (Barker 2001, and references therein) and studies on diplommatinid anatomy are very scarce, precluding any extensive comparisons. This gap in the knowledge of anatomy and physiology of operculate land snails, especially when compared to pulmonate snails, has long been recognized (e.g., Hunter 1964;Purchon 1977). To date, the most complete account of Cyclophoroidea remains that of Tielecke (1940), who unfortunately did not have South American diplommatinid specimens available for study. Overall, the shape of the snout, the arrangement of the pallial cavity, reproductive system, and the hypoathroid central nervous system of Habeastrum (Figs 3, 4) are all in line with Cyclophoroidea anatomy (Tielecke 1940;Barker 2011). The classification of Habeastrum in the cyclophoroidean family Diplommatinidae, besides the above-mentioned shell features, can also be based on its similarities to Latin American Adelopoma Doering, 1885. The only other Latin American diplommatinid species with a detailed anatomy is Adelopoma paulistanum Martins & Simone, 2014, from São Paulo state, SE Brazil. Habeastrum strangei sp. nov. shares some features with it, such as: the shape of the snout; the duplication of the eyes; the attributes of the pallial cavity (lacking gill and osphradium, with kidney and pericardium disposed in a similar way); the characters of the jaws and odontophore; and the relatively large and similarly-arranged nerve ring. By contrast, Habeastrum strangei sp. nov. differs from A. paulistanum by: the dextral coiling; the broader width of the pallial genital structures; the presence of esophageal glands; the lack of salivary glands; the stomach bulging posteriorly; and the presence of a penis. The latter is a plesiomorphic state in Diplommatinidae, as some species are aphallic (Tielecke 1940;Baker 2001;Martins and Simone 2014).
There is more information available on the radulae of other Neotropical diplommatinids than any other anatomical feature (Doering 1885;Baker 1923;Bartsch and Morrison 1942;Martins and Simone 2014). The radula of cyclophoroideans is plesiomorphically taenioglossate (Barker 2001;Ponder et al. 2008): two marginal teeth and one lateral tooth on each side of the central rachidian tooth. This condition seems to be kept in Adelopoma (Doering 1885;Baker 1923;Martins and Simone 2014) and also in Habeastrum strangei sp. nov. (Fig. 2B-E).
Besides the now three species of Habeastrum, there are only six other diplommatinids reported from Brazil, belonging to the genera Adelopoma and Habeas Simone, 2013(Simone 2006Birckolz et al. 2016;Salvador 2019). Nevertheless, the family allocation of the latter is yet to be confirmed, given that they are known only from empty shells and present unusual features for the family, such as a large shell size (> 5 mm overall and >10 mm for H. corpus Simone, 2013) and the presence of an anal notch in the aperture (Simone 2013). Most Brazilian diplommatinids have been described in the 2010s, typically from understudied cave environments (Simone 2013(Simone , 2019Martins and Simone 2014;present work). Furthermore, similarities in shell morphology across vast areas might hide a cryptic diversity (Salvador et al. 2018;Salvador 2019). Thus, the true diversity of this group in Brazil and its biogeographical implications are yet to be fully understood.