THE SUPERFAMILIES PUPILLOIDEA AND ENOIDEA (GASTROPODA, EUPULMONATA) IN BHUTAN

The species of two gastropod superfamilies, i.e. Pupilloidea and Enoidea, that have been recorded in Bhutan, are described and illustrated. Five families with ten species in total are dealt with. Three species are described as new to science, viz. Pupisoma (P.) paroense Gittenberger et Leda, n. sp., Pseudonapaeus occibhutanus Gittenberger, Gyeltshen et Sherub, n. sp., and Laevozebrinus parvus Gittenberger, Gyeltshen et Leda, n. sp. Distribution maps are presented for all the species. Some biogeographical considerations are added. kEy words: taxonomy; Pupilloidea; Enoidea; distribution; Bhutan Publication LSID urn:lsid:zoobank.org:pub:D0FFDEEA-E9DD-49D3-A7F6-3966DE97765C Pupisoma paroense LSID urn:lsid:zoobank.org:act:3548B71D-F10F-4F5E-89D1-9CB4F8FA9D75 Pseudonapaeus occibhutanus LSID urn:lsid:zoobank.org:act:A0D0A357-6619-4149-9FC5-91D8FDAC4F20 Laevozebrinus parvus LSID urn:lsid:zoobank.org:act:2562A8E7-02CD-47A7-A2D1-B7B44D10879E


INTRODUCTION
The description of Truncatellina bhutanensis by GittEnbErGEr et al. (2013) marks the beginning of the malacological inventory of the Kingdom of Bhutan, which took place with fieldwork from 2012 on. This geographically very diverse country is cut by high mountain chains with corresponding deep valleys. It belongs to the SE. Himalayan border zone. It has a surface of 38,394 km 2 , with an altitude of ca. 100 m a.s.l. in the south and over 7,500 m a.s.l. in the north. For a more detailed account, see GittEnbErGEr et al. (2018a). Meanwhile there has been a gradual increase in our knowledge of the gastropods and bivalves of that country. In a preliminary fieldguide (GittEnbErGEr et al. 2017) a first draft of the diverse molluscan fauna was published, inevitably with many uncertainties and far from complete in many respects. With an ongoing, stepwise approach we aim at an improved summary for the Bhutanese molluscs, their taxonomy and distribution. Here we present the first results of our inventory for the species of two superfamilies, viz. Pupilloidea and Enoidea. We are aware of the fact that this article offers an incomplete coverage of the existing malacofauna regarding these taxa since large areas in the country are still unexplored. Waiting for a more complete picture however, could easily result in lengthy silence.

MATERIAL AND METHODS
The specimens were collected without a standardised method, so that the number of specimens in a sample cannot be used as a measure of frequency. Colleting was by eye and by sieving soil samples taken during fieldwork in 2013-2019 within the scope of the Bhutan Evertebrata Inventory Project (see GittEnbErGEr et al. 2017). Most specimens are kept in the National Biodiversity Centre, Serbithang, Thimphu, Bhutan. Some duplicates, including paratypes of the new species, are deposited in Naturalis Biodiversity Center, Leiden, The Netherlands. In the type series, the number of specimens, i.e. shells, unless stated otherwise, is indicated after the slash after the registration number.
For every species, except Bensonella plicidens, there is a description on the basis of the specimens from Bhutan. Referring to budha et al. (2015,2017), in their annotated checklist for the Nepalese malacofauna, we mention for every species whether it is known from Nepal. For the names and authorships of the taxa above the species level, we rely on bouChEt et al. (2017).
The snails were dissected while in ethanol 70%. The genital tract was fixed with needles and shortly stained with cochineal. It was transferred to ethanol 97-100% before the needles were removed. The hardened tract was then transferred stepwise via Euparal essence to Euparal and mounted for permanent use on a microscopic slide. In the descriptions of the genitalia, under the heading 'Genital tract', we use proximal and distal in relation to the body wall.
Photographs were taken with a Canon EOS 7D, using a Canon Macro Photo Lens Mp-E65mm with a Macro Ring Lite MR-14EXII. The scanning electron microscopic images were taken with a Jeol JSM-Description. The brown to yellowish brown, rather fragile shell has 5½-6½ moderately convex whorls, separated by an incised suture; it is cylindrical with a convex-conical apical part, encompassing a third to nearly half of the total shell height. The protoconch is very finely granular. The teleoconch has a characteristic sculpture of prosocline riblets (14-16/mm) that are slightly heightened by periostracal ridges. In between the riblets there is an irregular sculpture of some radial lines on a silky background and a vague pattern of periostracal wrinkles. The umbilicus is less than 3% W broad and not always visible as a hole in strict basal view. The aperture ascends to about half the height of the penultimate whorl; it is little higher than broad (AH/AW = 1.05-1.17) and its height is 28-35% H. Except for the parietal side, there is a broadly reflected apertural lip. About halfway at the parietal side there is a moderately prominent parietal tubercle and near the palatal-basal transition there is an equally prominent second callosity, elongated as a quickly fading lamella inside the aperture. At the outside, shortly behind the apertural lip, there is a distinct rib that may have a lighter colour than the rest of the shell; behind it, the shell is somewhat flattened, sometimes with a faint indentation corresponding with the lamella inside the aperture. Measurements (n = 40). H 2.88-3.55 mm, W 1.66-1.88 mm. Distribution. This is a widespread Asian species, known from NE Iran (Kopet Dagh), northern Afghanistan and the province of Balochistan in Pakistan in the west, and as far eastwards as W China (Tien Shan and Tibet) (Pokryszko et al. (2009: 448). In Bhutan it was recorded only in the western districts of Paro and Thimphu, at altitudes of 2,040-2,450 m a.s.l. Notes. Recent authors incorrectly use the epithet turcmenica. P. turcmenia resembles its sister species P. sterrii (Voith, 1840) (nEkola et al. 2014: 204, fig. 1) in shell shape and sculpture. sChilEyko (1984: 190, fig. 108), mEnG (2008: 224), sysoEv & sChilEyko (2009: 46, fig. 19D), and Pokryszko et al. (2009 et al. (2014: 211) however, the number of apertural 'lamellae' varies from 0 to 3. In all our specimens from Bhutan there is not only a tubercular parietalis but also a slightly elongated, lamella-like palatalis. The shells are clearly P. turcmenia, in particular also by the characteristic sculpture of radial periostracal riblets. P. turcmenia is not known from Nepal. In that country the genus is represented with certainty only by P. eurina (Benson, 1864) and P. triplicata (Studer, 1820) (budha 2015: 10; 2017: 25).

Family Gastrocoptidae Pilsbry, 1918
Genus Gastrocopta Wollaston, 1878 16 km SSW of Thimphu, 3 km N of bridge, altitude 2,100 m a.s.l., 27°20'N, 89°34'E, EGPL 6.04.2013 (NBCB1160); 2 km S of Thimphu, altitude 2,350 m a.s.l., barren S-exposed slope, 27°26.35'N, 89°39.15'E, EGCGPL 20.10.2018(NBCB 1194. Description. The greyish white, slightly transparent shell is cylindro-ovoid to subcylindrical; it has 4¾ convex whorls with irregular growth lines. The last whorl is obliquely flattened and has a slight indentation, corresponding with the position of the palatalis inferior inside the aperture; the umbilicus is minute. The aperture is roundish, with a continuous, broadly reflected peristome, not protruding at the parietal side; there are (1) two columellar denticles, viz. a prominent columellaris and a smaller infracolumellaris, which is positioned near the transition to the basal side, (2) two or three palatal denticles, viz. a prominent palatalis inferior, a moderately prominent palatalis superior, and a rudimentary suprapalatalis, which may be lacking completely, and (3) at the parietal side, a parietalis that is fused with the little smaller angularis into a parieto-angularis, and often a small infraparietalis. Measurements (n = 18  Pilsbry, 1917 are considered synonyms (see below). . The light brown, conical shell has irregular, radial, vague riblets and growth lines; the umbilicus is narrow. The aperture is little higher than broad; it is broadly rounded basally. Inside the aperture there are 14-16 'lamellae', most of which are hooked; they vary considerably in length. The peristome is more or less clearly interrupted at the parietal side. Measurements (after budha & baCkEljau 2017: 198). H 2.0-2.5 mm, W 1.5-1.9 mm. Distribution. This easily identifiable species (but see notes) is reported with two subspecies from a limited number of scattered localities, ranging from NW. India in the west to Japan and Taiwan in the east (Fig.  9). This geographic distribution is unusual because of the combination of localities that belong in different biogeographic areas, like Taiwan and Bhutan. We do not know a similar pattern from other molluscs. The only record for Bhutan is at an altitude of 1,700 m a.s.l. SW of Pemagatshel in the district of that name. Notes. Only one fragile shell was found. This vulnerable specimen was not used for photographs. It has the conspicuous claw-like apertural 'lamellae', which are considered diagnostic for this species. However, budha & baCkEljau (2017: 197) suggested that the hooked lamellae might not even be a species specific character state since in many samples labelled as this species the lamellae are simple, not hooked. They further hypothesized that there has been a "mixture of hooked and non-hooked shells in Benson's collection". It remains to investigate then whether there are additional differences between these two categories of shells. Sub-adult specimens could be studied to learn how the apertural lamellae are formed. The simple lamellae might represent an intermediate stage that is followed occasionally by partial dissolving of the lamellae, eventually resulting in the apertural claws. Secondary dissolving apertural lamellae are known from Pupilloidea like Orcula species (zilCh 1959: 155, GittEnbErGEr 19831996: 198  Description. The fragile, whitish shell is discoid, with a low conical spire that is hardly visible in frontal view. The protoconch of 1⅛ whorls is smooth. The following, little more than 2 teleoconch whorls, have thin radial ribs that are heightened by periostracal lamellae; in between the ribs, the periostracum forms an irregular pattern of interconnected, roughly radially oriented line fragments and even smaller wrinkled ridges. There are 55-72 (n = 15) main ribs on the last whorl. The edge of the roundish aperture is narrowly reflected but not thickened and shortly interrupted without a callous thickening; its upper palatal side descends to about the periphery of the shell or little lower. The elliptical umbilicus measures nearly 40% W.  . 8) is also inconclusive for that taxon. Pupa seriola was superficially described after two shells by bEnson (1863: 427) from (p. 428) "regione Orissae (Cuttack)", India. The measurements are indicated as 2½ × 1⅓ mm and for one shell the presence of a parietal lamella in mentioned. These data are sufficient to conclude that the Pupisoma from Bhutan is a different species, with smaller shells without any denticle or lamella in the aperture. Etymology. The neutral Greek noun 'soma' in Pupisoma goes with the epithet paroense, to refer to the occurrence in the district of Paro.   Diagnosis. Shell thin, ovoid-conical, with spiral striae and a colour pattern of oblique irregular streaks. Description. The shell is relatively thin and ovoid-conical, with slightly convex whorls, separated by an incised suture. The apical part is light yellowish brown; most of the shell has brownish, very irregular streaks in the direction of the growth lines and spots, alternating with lighter parts. The protoconch has 1½ whorls and is smooth. The 5-5½ teleoconch whorls have dense irregular growth lines, crossed by relatively prominent dense spiral lirae. The aperture is higher than broad, measuring ca. 40% H; its columellar edge is straight and runs parallel to the middle part of the palatal edge that gradually passes into the basal edge. The whitish apertural lip is somewhat thickened and expanded but not reflected, except for the columellar part in front of the narrowly open umbilicus. There is hardly any parietal callus.

Vallonia costohimala Gerber et Bössneck, 2009
Measurements (n = 35). H 13.6-17.5 mm; W 6.3-6.8 mm Genital tract (Figs 26-28). The most proximal part of the penis is little narrower than the basal part (A1) of the penial appendix which inserts close to the genital atrium. The penis gradually narrows considerably towards the attachment of the retractor muscle shortly before the transition to the proximal part (e1) of the epiphallus, which is 2-3 times as broad and contains an elongated papilla. The short following part (e2) is nearly as narrow as the distal ending of the penis and connects to the relatively long and broad part (e3) that ends distally with a small caecum after which an equally broad part (e4) follows, ending with a short but prominent flagellum (f). The vas deferens (vd) inserts laterally to the epiphallus. The penial retractor muscle (rp) and the appendix retractor muscle (ra) insert separately on the diaphragm.
The basal part of the penial appendix (a1) has a lumen with a dense transverse structure, it is as broad as the longest segment (e3-4) of the epiphallus and half as long. A short globular part (a2), near the attachment of the separate retractor muscle, is followed by a twice as long cylindrical part with a   Description. The shining, fragile shell is slender conical with a blunt apex; it has 5½-6¼ moderately convex whorls, separated by a deeply incised suture. There is an irregular, vague, colour pattern of light and darker brown, or yellowish brown blotches. The protoconch is fine silky with some spiral lines near the suture; the teleoconch has coarse growth lines and parts with very faint spiral striae. The more or less regularly oval aperture is higher than broad, measuring 35-40% H; its straight or curved columellar edge is about as long as the parietal interruption, whereas the palatal edge gradually passes into the basal edge. The apertural lip is poorly thickened and little expanded; its uppermost columellar part is curved in front of the very narrow umbilicus. There is a noticeable periostracum. Measurements (n = 20). H 7.6-9.6 mm, W 3.9-4.6 mm. Genital tract (Figs 38-40). The entire penis (p) is about as broad as and little narrower than the epiphallus (e). The penial appendix (a) inserts about halfway on the penis; it has a proximal part (a1), with irregular longitudinal lamellae in the lumen which is about as broad as the penis. Its short, globular part (a2), with a transverse luminal structure, is followed by a narrower, short segment (a3), with a meandering structure in the narrow lumen; this is followed by the narrowest segment (a4) with a relatively wide meandering structure and a thin wall, and the elongated bursa (a5) that has a wide lumen and a moderately thick wall. The epiphallus (e) is nearly twice as long as the penis; it has a short blunt flagellum and its most proximal and broadest part (e1) is shorter than half the length of the penis and broader, with an irregular structure in the lumen. After a short narrowing (e2), a longer segment with luminal septa follows (e3); an inconspicuous thickening, the caecum (c) borders the distal part of the epiphallus (e4) with a more irregular lumen and the flagellum (f) with a longitudinal fold. The separate retractor muscles (ra, rp) run from the diaphragm to the distal third of a1 and close to the distal end of the penis. The vagina (v) is clearly broader and shorter than both the penis and the oviductus (o). The pedunculus (pd) is about as long as the diverticulum (d); the globular bursa (b) is connected to the pedunculus with a narrow duct that is as long as its diameter.
In one of the three dissected snails, the genital tract looked different, most probably showing a phase of sexual activity (Fig. 40). Here, most of the male parts are enveloped by a strongly inflated part of the tract, with a conspicuous projection. The epiphallus is easily recognisable because of its flagellum and caecum. Its position in relation to the genital atrium has reversed. The same applies to a4-a5 of the penial appendix. We refrain from speculations about morphology and functionality here.
Differentiation. The geographically closest congeneric subspecies Laevozebrinus nepalensis nepalensis Schileyko et Frank, 1994 (Figs 35-36) and L. nepalensis myagdiensis Kuznetsov et Schileyko, 1997, both from Nepal, differ in stronger, on average smaller shells with a thickened and reflected apertural lip. Their genital tracts differ in the bursa copulatrix that has no diverticulum, in the position of a more prominent caecum about halfway the epiphallus and in the retractor muscles that are united shortly before the attachment to the diaphragma. Laevozebrinus mustangensis Kuznetsov et Schileyko, 1997 from western Nepal is larger; its genital tract differs in a longer, slender flagellum and in the insertion of the retractor muscle near the distal end of a1 of the penial appendix. Laevozebrinus guttula (Muratov, 1992), as figured by sysoEv & sChilEyko (2009: fig. 24C), is equally small and only distinguishable by the more oblique attachment of the palatal edge of the aperture and the very narrow columellar margin. Since that species is known from only the SW. Kopet Dagh in Turkmenistan, 3,500 km WNW of Bhutan, we consider a close taxonomic relationship unlikely.  fig. 37C), whereas it is much more gradual in the original drawings for the species (kobElt 1902: pl. 131 figs 3, 4). Since the basal angularity could be an individual aberration, we studied both the lectotype and the only known paralectotype of this species. Photos of these specimens were made available by courtesy of Ms. Sigrid Hof (Forschungsinstitut und Naturmuseum Senckenberg, Frankfurt am Main, Germany) and are reproduced here. In conformity with the original drawings, the basal part of the ap-erture is regularly rounded in the paralectotype and angular in the lectotype, that has no growth lines indicative of former breakage and repair. The relative size of the aperture is also quite different in both shells. In that respect, the two shells from Bhutan correspond with the lectotype of M. hanleyana. For the time being, we here follow GudE (1914: 233), who described a mixed sample of M. nilagiricus and M. hanleyana, emphasising differences that are recognisable very well in our specimens. The large disjunction between the type locality in southern India and our record for Bhutan strongly reminds of the biogeographically curious records for M. nilagiricus.

DISCUSSION
The largest species diversity of Pupilloidea and Enoidea was found in the western districts of Haa, Paro and Thimphu, which can be regarded as a biogeographical entity. In their distribution, both the genera and the species in that area are quite diverse. The genera Pupilla and Vallonia are Holarctic (zilCh 1959: 166, GErbEr 1996, and contribute with a widespread Asian and a local SE Himalayan species, viz. P. turcmenia and V. costohimala, respectively. Truncatellina may be considered a Palaearctic genus (zilCh 1959: 148), despite the occurrence of T. sykesii (Melvill et Ponsonby, 1893) in S. Africa (van bruGGEn 1978: 920); it is represented by the endemic T. bhutanensis. The genus Gastrocopta, which cannot be classified unequivocally in a biogeographic category (see zilCh 1959: 160-162), was recorded with G. huttoniana, that has a relatively wide Asian range. The poorly known genus Pupisoma, that was represented by the endemic P. paroense, also does not fit clearly in any of the biogeographical categories (see zilCh 1959: 174). The genera Pseudonapaeus and Laevozebrinus, each with an endemic Bhutanese species in the western districts of Bhutan, viz. P. occibhutanus and L. parvus, are both restricted to the Asian part of the western and central Palaearctis (sChilEyko 1998: 196, 198). Only the two Mirus species, both with disjunct Bhutanese-Indian ranges (rahEEm et al. 2014: 65), are Indo-Malayan; the few records suggest that they are not restricted to a particular part of southern Bhutan. The distributional pattern of Bensonella plicidens s. lat., a species which is reported from a zone of 6,000 km between 22°N and 35°N, with only one record in eastern Bhutan, is exceptional.
While considering the distributional patterns, it should be taken into account that large parts of western Bhutan are still unexplored. There are no records, for example, from the western district of Samtse. Ongoing research in India and Nepal may show whether the species that are referred to as endemic here, are endemic indeed. Despite this fact, we may already conclude however, that there is a considerable percentage of endemism in the malacofauna of Bhutan. That conclusion is based on the present and various previous taxonomic analyses. Three of the four species of Rahula Godwin-Austen, 1907(Euconulidae H. B. Baker, 1928