Diversity and biogeography of land snails (Mollusca, Gastropoda) in the limestone hills of Perak, Peninsular Malaysia

Abstract Limestone hills are now gaining global conservation attention as hotspots for short-range endemic species. Levels of land snail endemism can be high at limestone hills, especially at hill clusters that are geographically isolated. In the State of Perak, Peninsular Malaysia, limestone hills have been opportunistically surveyed for land snails in the past, but the majority have yet to be surveyed. To address this knowledge gap, we systematically surveyed the terrestrial malacofauna of 12 limestone hills that, based on our opinion, are a representation of the limestone land snail assemblages within the State. Our inventory yielded high sampling completeness (>85%). We found 122 species of land snails, of which 34 species were unique to one of the surveyed hills. We identified 30 species that are potentially new to science. The number of land snail species recorded at each hill ranged between 39 and 63 species. Four of the sampled limestone hills namely, Prk 01 G. Tempurung, Prk 55 G. Pondok, Prk 47 Kanthan, and Prk 64 Bt Kepala Gajah, have high levels of species richness and unique species, representing 91% of the total species recorded in this study. We identified two clusters of limestone hills in central Perak with distinct differences in land snail species composition – a northern hill cluster on elevated granite bedrock and southern hill cluster in a low-lying valley surrounded by alluvial soils. As limestone hills continue to be quarried to meet the cement demand, the four identified limestone hills, along with other hills from the two clusters, warrant urgent conservation attention in order to maintain high species diversity within Perak’s terrestrial malacofauna.


Introduction
Limestone hills are popularly known as "arks of biodiversity" because they contain high levels of species endemism, especially in Peninsular Malaysia where these hills only cover 0.2 % of the total land area (Schilthuizen 2004;Clements et al. 2006;Chua et al. 2009;Liew et al. 2014).
To date, at least 445 limestone hills have been documented in Peninsular Malaysia, with the majority located in the States of Kelantan (149 hills), Pahang (124 hills) and Perak (93 hills) (Liew et al. 2016). The State of Perak has the third largest number of limestone hills, but it has the largest number of operating quarries (Liew et al. 2016). The majority of these hills can be found within the Kinta Valley, with some other hills scattered around the northern part of the Kinta Valley and Bintang Range (Figure 1). Over the past decade, the endemism of different plant and animal at the limestone hills of Perak has drawn attention from evolutionary biologists (Clements et al. 2008a), conservationists (Clements et al. 2008b) and concerned taxonomists (Kiew et al. 2014;Vermeulen and Marzuki 2014). Given the large number of limestone hills in Perak, it is not practical to spare every hill from quarrying. Thus, conservation prioritisation needs to be conducted, ideally based on the biogeographical patterns of endemic taxa such as land snails (Clements et al. 2008b).
The limestone hills in Perak form an important region because that was where the foundation for land snail knowledge in Peninsular Malaysia was laid. More than 50 land snail species have been described from this State alone; these species descriptions were the outputs of three main faunistic studies by separate groups of naturalists and malacologists. The first was an expedition to central Perak by de Morgan (1885a), who described 33 new land snail species from the area. Second was the work by von Möllendorff, who published a species checklist of around 60 species and described 15 new species based on the collection by R. Hungerford at Gunung Larut (von Möllendorff 1886(von Möllendorff , 1891. Third was the survey done by E. Townsend at Gunung Pondok, where around 13 new species were described (Godwin-Austen and Nevill 1879; Crosse 1879a, 1879b). Besides these faunistic surveys in Perak, at least 20 new species were described based on taxonomic studies of particular land snail taxa or random sampling (van Benthem Jutting 1952, 1954, 1961a, 1961bTomlin 1938Tomlin , 1939Tomlin , 1941Fulton 1901Fulton , 1902Sykes 1903;Godwin-Austen 1909). More recent studies suggest that about one-third on the known species from Peninsular Malaysia can be found in Perak, particularly around the Kinta Valley (Davison 1991;Chan 1998aChan , 1998bClements et al. 2008b;Maassen 2001). In Perak, land snail species richness on a hill can reach as high as 53 species (Clements et al. 2008b).
Despite intensive land snail research conducted in the State of Perak, the land snail inventory is far from complete. Most of the comprehensive surveys only report the name of the species, without illustrations of the species, or with provisional working species names, such as sp. 1, sp. 2. (e.g. Davison 1991;Clements et al. 2008b). This has made it difficult to update current checklists because species identities in previous reports cannot be used to compare with recently collected specimens. Furthermore, many of these unnamed species in previous reports are probably new species yet to be described or are doubtful species that require taxonomic revision. Hence, good annotated checklists of land snails with traceable specimens and high quality illustrations are key to improve the current knowledge of land snails in Perak and Peninsular Malaysia in general.
The biogeography of organisms on limestone hills in Perak is also poorly known. In Peninsular Malaysia, land snail communities on limestone hills can generally be divided into two groups, one on the hills along the east coast and the other on the hills along the west coast, both of which are separated by Main Range in the centre of the peninsula (van Benthem Jutting 1960;Clements et al. 2008b). Within the State of Perak, Davison (1991) noticed land snails on limestone hills can be further divided into two groups, one on the hills within Kinta Valley and the other one on the hills scattered in the area north of Kinta Valley, though based on an incomplete land snail inventory (Davison 1991: 6 -35 species from 12 hills, averaging 17 species per hill, compared to Clements et al. 2008b: 32 -53 species from 6 hills, averaging 42 species per hill). As such, further quantitative analysis on the biogeographical patterns of land snails with more comprehensive species inventories from limestone hills that are representative of Perak's terrestrial malacofauna can contribute to improved limestone conservation planning within the State.
Limestone hills have been recently highlighted as one of Malaysia's vulnerable ecosystems due to surrounding forest degradation and quarrying activities (Ministry of Natural Resources and Environment Malaysia 2016). Many land snail species that are endemic to limestone hills in Peninsular Malaysia are already extinct or on the brink of extinction (http://www.iucnredlist.org/search/link/57e5b1fd-2bc5b54c). In order to prevent further species extinctions, limestone hills that are currently being quarried or intact must be urgently assessed for land snail diversity.
Here, we conduct a land snail inventory in and around the Kinta Valley of Perak to: (1) provide an annotated checklist of land snail species for the State, along with photographic images for each species; and (2) elucidate land snail diversity and biogeographical patterns across 12 limestone hills within the State. For the second objective, we specifically examined relationships between limestone hill parameters (size and isolation) and species richness and biogeographical patterns of land snails. Finally, we discuss the conservation implications of our study for limestone hills in Perak.

Sampling design and sites
In addition to six limestone hills previously surveyed by Clements et al. (2008b), six other limestone hills were included in this study. Among the six previously sampled hills, only one (i.e. Gunung Bercham) was resampled because of the low sampling Table 1. The geographical coordinates, size, isolation of 12 limestone hills sampled in this study in and around the Kinta Valley of Perak. The names of the limestone hills follow a standardized national code developed by Liew et al. (2016).

No.
Limestone hills Longitude Latitude Size (km 2 ) Degree of Isolation @

Species richness
completeness reported by Clements et al. 2008b. Sampling was conducted between 16/8/2016 and 24/8/2016. Our sampling approach consisting of 12 limestone hills will (1) result in the sampling of more hills in clusters that have a larger number of limestone hills; and (2) provide wider geographical coverage of limestone hills across the Kinta Valley (Table 1; Figure 1).

Land snails sampling and processing
In each of the seven limestone hills, four 2 m × 4 m plots were established (Clements et al. 2008b;Liew et al. 2008;Suppl. material 1). In each plot, a total of five litres of top soil and leaf litter were collected for extraction of micro-snails (< 5 mm) and the plot was searched for macro-snails. Upon returning to the laboratory, all macro-snails were cleaned with running water and then dried in an oven. After that, shells were extracted from soil samples by manually picking up the shells under a stereomicroscope (Liew et al. 2008). For species identification, a complete literature of Malay Peninsular terrestrial mollusca was consulted, with emphasis on the most recent species compilation and overview by Maassen (2001). All specimens from recent sampling and previous studies (Clements et al. 2008b) were identified to morphospecies based on a combination of photographs, illustrations and description of types. Morphospecies that could not be assigned to an available name were given working morphospecies names (for example, Acmella 'Kanthan 1'). Species reported to be present in our study sites in the literature, but not found during our sampling, were not included in our results. All specimens were catalogued in the BORNEENSIS Mollusca collection database and were deposited in the BORNEENSIS collection of Institute for Tropical Biology and Conservation, Universiti Malaysia Sabah.

Data analysis
A species checklist was compiled for the 12 limestone hills and was arranged according to the classification of Maassen (2001). Under species or subspecies, four subsections were provided: (1) Reference to figures; (2) Materials examined, which includes only species sampled from this study (with the exception of some species where poor shell preservation warranted a substitute specimen from other studies for photographic purposes); (3) Distribution, which is separated into distribution within Peninsular Malaysia and distribution elsewhere; (4) Remarks, which includes brief diagnosis, comparison with conspecifics and note on new records. All species are considered native unless noted as synantropic in Remarks.
For our analysis of land snail diversity patterns, we first assessed species diversity in terms of (1) species richness -total number of species for each hill; and (2) unique species -species found only in one of the twelve limestone hills in this study. Next, we assessed whether there was a correlation between the species richness and the number of unique species for the twelve hills. We also examined relationships between species diversity and limestone hill parameters (i.e. size and degree of isolation). Depending on the normality test of data for each parameter, Spearman or Pearson correlation coefficients were used to test for correlations. Analysis was conducted using built-in function in the R statistical environment v. 3.3.1 (R Core Team 2016).
For our analysis of land snail biogeographical patterns, land snail data were tabulated into a data matrix of 122 species × 12 limestone hills that consisted of absence (0)/presence (1) data for each species on each of the twelve hills. We evaluated sampling completeness for all 12 hills and compared their species composition to determine their degree of similarity. A cluster analysis was performed to objectively assign limestone hills into clusters so that hills within each cluster are similar to one another with respect to overall land snail composition. In our analysis, a dissimilarity distance matrix was calculated using the Jaccard similarity coefficient (i.e. Jaccard index) based on the data matrix. Next, hierarchical clustering on the 12 sites based on the dissimilarity matrix was performed using the method of complete linkage. Analysis was conducted using package 'vegan' (Dixon 2003) and 'iNEXT' (Hsieh et al. 2016) in the R statistical environment v. 3.3.1 (R Core Team 2016).

Diversity and biogeography of land snails
We achieved high sampling completeness (> 85 %) for all seven limestone hills sampled in our study (Suppl. material 3). The species richness and number of unique species that can be found only in a single hill is listed in Table 1. The number of land snail species recorded in each hill ranges between 39 species and 63 species. Gunung Kanthan has the highest number of species -63 species with 6 unique species, and followed by Gunung Tempurung -54 species with 9 unique species. The list of 34 unique species found in each limestone hills is listed in Table 2. The species richness and number of unique species of each hill were also mapped to understand land snail diversity patterns among the limestone hills in and around the Kinta Valley of Perak (Table 1) Species richness and the number of unique species were not correlated among the 12 hills (p = 0.052, Figure 2A). There appears to be no relationship between the species richness and limestone hill size (p = 0.12) and isolation (p = 0.5) ( Figure 2B, C). On the other hand, the number of unique species was strongly correlated with limestone hill size (r s = 0.91, p = 0.0005), but not isolation (p=0.43) ( Figure 2D, E).
Cluster analysis identified two clusters of limestone hills based on land snail species composition: a northern cluster and southern cluster (  A Relationship between species richness and number of unique species B Relationship between species richness and limestone hills size (km 2 ) C Relationship between species richness and degree of isolation of the limestone hills D Relationship between number of unique species and limestone hills size (km 2 ) E Relationship between number of unique species and degree of isolation of the limestone hills.

Figure 3.
Cluster analysis of land snail species composition of 12 limestone hills in and around the Kinta Valley of Perak, Peninsular Malaysia. Two clusters were identified: a northern (blue dots) and southern cluster (yellow dots). Green polygons with red boundary refer to major clusters of limestone hills in Kinta Valley identified based on the 2.5 km buffer analysis (i.e. at least 2.5 km from the nearest cluster/hills). Red dots indicate limestone hills not sampled in this study.

Checklist
In total, we recorded 122 species from the 12 limestone hills sampled in and around the Kinta Valley of Perak (Suppl. material 2). This checklist consists of 23 families and 47 genera. There are 30 out of 122 species identified from this project which could not be assigned to scientific names that are currently published. Some of these unnamed species are potentially new to science but systematic taxonomic revisions need to be done as many of these groups have not been critically revised. The most diverse genera were Opisthostoma (13 species), Microcystina (10 species), and Diplommatina (9 species).  Distribution. In Peninsular Malaysia, known from Perak, Selangor and Kelantan (Maassen 2001). Elsewhere, in Tenasserim (=Tanintharyi), Myanmar and Salang (=Phuket), Thailand (Maassen 2001).
Remarks. Distinguished from congeners by its larger, low spired shell. Glossy surface. Periphery prominently keeled. Radial ribs prominent along the section of the whorl parallel to the pneumatophore, indistinct elsewhere. Shape similar but shell larger than Chamalycaeus oligopleuris.   Figure 5C Materials examined. mykarst-027: BOR/MOL 9039, BOR/MOL 9127.
Remarks. Shell shares similar prominent, widely spaced-out radial and spiral ridges with Chamalycaeus oligopleuris but differs in having taller spire and rounder periphery.
Distribution. Known from Gunung Kanthan only, but surrounding hills have yet to be adequately surveyed.
Remarks. Distinguished from congeners by its arrangement of radial ribs, taller spire and the position of the pneumatophore.  Remarks. Distinguished from all congeners by its tall spire, conical shell, penultimate whorl of constant width and the cross-hatching of radial and spiral ridges in post-apical whorls. Figure 5F Materials examined. Remarks. Shell similar size to Chamalycaeus microdiscus but whorls differ in being rounder and radial ribs denser and less pronounced. New record for Kinta Valley.
Remarks. Shell distinguished from all other discoid cyclophorids by the presence of a deep, tunnel-like sutural canal which begins as a tube-like peristomal folding at the suture part of the aperture.

Cyclophorus semisulcatus (Sowerby, 1843)
Remarks. Distinguished from congeners by its small to large shell, pronounced spiral ridges and somewhat keeled periphery. Colour pattern highly variable.
Remarks. Shell small, conical. Spire tall. Dorsal whorl angled at mid-distance between periphery and suture, in a region termed major spiral rib by Solem (1966). Differ from Lagocheilus townsendi in having a narrower penultimate whorl. New record for Perak.
Distribution. Known from Gunung Pondok only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell small, conical. Spire tall. Whorls convex, smooth with fine radial growth lines. Periphery keeled. Differ from congeners in the absence of spiral ridges.   Distribution. Known from Perlis, Perak and Selangor only (Maassen 2001). Remarks. Shell small, conical. Spire tall. Similar to Lagocheilus kobelti in the presence of angled dorsal whorls at major spiral rib but penultimate whorl is more expanded and shell larger (Sykes 1903).
Remarks. Shell small, conical. Dorsal whorls with widely spaced, indistinct brown spiral ridges. Distinguished from Leptopoma perlucidum (de Grateloup, 1840) by its taller spire relative to shell width. This is an arboreal species.
Remarks. Distinguished from Opisthoporus rostellatus (Pfeiffer, 1851) by the pneumatophore being nearer to the aperture and the absence of a wing-like sutural extension at the peristome.   Distribution. In Peninsular Malaysia, known from Perak (Maassen 2001). Elsewhere, in Singapore (Maassen 2001).
Remarks. Distinguished from Opisthoporus penangensis (Stoliczka, 1872) by the pneumatophore being longer and further from the aperture and the presence of a wing-like sutural extension at the peristome. Figure   Distribution. Known from Batu Kebelah and Bukit Kepala Gajah, Perak only, but surrounding hills have yet to be adequately surveyed.

Opisthoporus solutus (Stoliczka, 1872)
Remarks. Very small shell at maturity compared to all congeners. Shell discoid. Medium spire. Fine spiral ridges and radial ridges interspersed with major radial ridges at regular intervals. Often encrusted with mud and plant debris.   Remarks. Shell small to medium-sized. Final whorl occasionally detached, with large upturned pneumatophore formed through peristomal folding. Shell patterns vary from dense zigzags and peripheral band over brown whorls to fugitive brown zigzags over white or purple whorls. Rhiostoma macalpinewoodsi (Laidlaw 1939) is synonymous with Rhiostoma jousseaumei, as intermediates were found in our study sites. Distribution. Known from Gunung Rapat only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell dextral, very small. Cylindrical, tall spire. Radial ribs dense, spiral ridges very fine. Differ from Arinia (Notharinia) micro in that the final whorl is completely detached and coiled downwards.
Distribution. Known from mykarst-027 only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell sinistral. Shell very similar to Diplommatina diminuta von Möllendorff, 1891 but differ in having denser radial ribs.

Distribution. Known from upper Kinta Valley and Perak River valley, Perak only (Maassen 2001).
Remarks. Shell dextral. Smallest shell of all congeners. Also differ from congeners by the peristome that is reflected and thickened multiple times, sharply angular at the base of the columella and slightly angular at the upper outer lip (Laidlaw 1949). Davison (1991) considers D. crosseana as possibly related to Diplommatina maduana Laidlaw, 1949. Figure 10A Materials examined.   (Maassen 2001). Elsewhere, in Singapore, Thailand and Indonesia (Ho 1995, Panha and Burch 2005, van Benthem Jutting 1941.

Distribution. Found across Peninsular Malaysia
Remarks. Shell dextral, rarely sinistral. Shell vary greatly in height, rib density, number of whorls (7 to 8) and aperture thickness within and between populations. After examining large samples from our study sites, we concur with Tweedie in Laidlaw (1949) that it is not possible to separate Diplommatina canaliculata von Möllendorff, 1886 from Diplommatina nevilli. The specimens labelled as Diplommatina mirabilis in Davison (1991) (Maassen 2001). Remarks. Shell dextral, rarely sinistral. Distinct from all sympatric congeners by its rather distended ultimate whorl and reflected aperture which extends to the right of the body whorl. Sinistral specimens of this species were labelled as Diplommatina sp. in Davison (1991). Figure 11B Materials examined. Remarks. Similar to Opisthostoma gittenbergeri in the manner of coiling for the penultimate and ultimate whorls. Differ from O. gittenbergeri in wider whorl coiling and position of apical and antepenultimate whorls. Radial rib density is variable.
Distribution. Known from Gunung Kanthan only, but surrounding hills have yet to be adequately surveyed.
Remarks. Similar to Opisthostoma subconicum Vermeulen, 1994, in shell shape, differ in having a more depressed apical whorl, less expanded peristome, more pronounced parietalis at the lower aperture and tooth at the upper aperture.
Distribution. Known from Gunung Tempurung only, but surrounding hills have yet to be adequately surveyed.  Distribution. Known from Gua Datok and Gunung Bercham only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell shape distinctive among congeners. Differ from Opisthostoma cf. gittenbergeri in whorl coiling being tighter and the position of apical and antepenultimate whorls.
Opisthostoma 'Kanthan 1' Figure 13C Materials examined. Remarks. Spire height, radial rib density and whorl shape variable. Differ from Opisthostoma cf. subconicum in having rounder whorls, finer but distinct radial ribs and the direction the aperture is facing. Distribution. Known from Kramat Pulai (van Benthem Jutting 1952) and Gunung Rapat only, but surrounding hills have yet to be adequately surveyed.
Remarks. Differ from other flat-shelled congeners by the convexity of the whorl, larger umbilicus, the position of apical and antepenultimate whorls and the direction the aperture is facing (van Benthem Jutting 1961a).
Distribution. Known from mykarst-025 only, but surrounding hills have yet to be adequately surveyed.
Remarks. Differ from other flat-shelled congeners by its much smaller shell and most distinctly, the coiling of the antepenultimate whorl.   Distribution. Known from Gunung Tempurung, but surrounding hills have yet to be adequately surveyed.
Remarks. Distinct from congeners in shell shape. Radial ribbing dense. Opisthostoma 'tempurung 1 detached' has a detached ultimate whorl similar to Notharinia 'tasek 1' but differ in having larger and less number of whorls. Figure 14C Materials examined. Remarks. Distinguished from congeners by its box-like shell shape, pronounced but widely spaced radial ribs and the direction the aperture is facing is parallel to the coiling axis of the apical whorl. Previously known from Gunung Kanthan only (Maassen 2001). Figure 14D Materials examined. Prk 23 G. Rapat: BOR/MOL 10214, BOR/MOL 10242, BOR/ MOL 10267.

Opisthostoma vermiculum Clements & Vermeulen, in Clements et al. 2008
Distribution. Known from Gunung Rapat only ), but surrounding hills have yet to be adequately surveyed.
Remarks. Distinguished from congeners by its shell shape and the presence of four coiling axes. Differ from Opisthostoma cf. vermiculum only in the coiling axis of the apical and antepenultimate whorls.  Remarks. Shell large. Operculum hard, calcified, multilamella. Distinct from any sympatric land snail species in the structure and coiling of the whorls. Distribution. In Peninsular Malaysia, found in Perak and Selangor (Maassen 2001). Elsewhere, in Moulmein (=Mawlamyine), Myanmar and Ko Samui, Thailand (Maassen 2001).
Remarks. Shell colour varies from white to brown, glossy. Whorl convexity varies slightly but is always less expanded than congeners. Differ from all sympatric Pupina species in being small, whorls less expanded as well as the formation and curvature of the aperture and its narrow canals.  Remarks. Shell colour brown, glossy when alive. Whorl more convex and expanded than most congeners. Shell larger than most sympatric congeners. Distinguished from congeners by the formation and curvature of the aperture and its canals, as well as the peristome extension right of the penultimate whorl. Figure 15D Materials examined.  Distribution. In Peninsular Malaysia, known from Penang, Perak and Pahang (Maassen 2001, Tumpeesuwan andPanha 2008). Elsewhere, in Petchaburi, Thailand as well as Tavoy (=Dawei) and Tenasserim (=Tanintharyi), Myanmar (Tumpeesuwan and Panha 2008).
Remarks. Shell small. Differ from the sympatric Georissa semisculpta by its taller spire, pronounced and crenulated spiral ridges as well as less obtuse whorls. Shell colour yellow to red. Figure 16D Materials examined. Remarks. Shell small. Differ from the sympatric Georissa monterosatiana by its lower spire, indistinct spiral sculpture over radial growth lines and very obtuse whorls. Shell colour pink to red. New record for Kinta Valley. Prior to this study, G. semisculpta was thought to be extinct because it was presumed endemic to Gunung Pondok, where it has not been detected ever since its original description (Davison 1991).

Clade Heterobranchia Informal group Pulmonata Cuvier, in de Blainville 1814 Family Achatinellidae Gulick, 1873 Genus Elasmias Pilsbry, 1910
Elasmias terrestris (Brazier, 1876) Figure  Distribution. In Peninsular Malaysia, found in Kedah and Perak (Maassen 2001). Elsewhere, New Guinea and Australia (Maassen 2001, Solem 1988. Remarks. Small shell. Distinguished from others by its oviform shell and truncated columella with two lamella teeth (one more pronounced than the other). This is a new record for Perak.
Remarks. Large shell. Perak shells differ from Penang shells only by their flatter spire. Whorls brown with darker periphery. Peripheral keel pronounced. Shell surface has numerous short spiral crenulations.

Distribution.
Known from Gunung Tempurung only, but surrounding hills have yet to be adequately surveyed.

Macrochlamys 'tempurung 2'
Distribution. Known from Gunung Tempurung, but surrounding hills have yet to be adequately surveyed.

Microcystina clarkae Maassen, 2000
Remarks. Very small shell. Easily distinguished from congeners by its white shell, tight whorls, fine radial ribs and finer spiral pits. This is a new record for Peninsular Malaysia.  Distribution. Known from Bukit Kepala Gajah, Perak only, but surrounding hills have yet to be adequately surveyed.

Microcystina 'guatokgiring 2'
Distribution. Known from Bukit Kepala Gajah, Perak only, but surrounding hills have yet to be adequately surveyed.

Microcystina 'tempurung 2'
Distribution. Known from Gunung Tempurung, Perak only, but surrounding hills have yet to be adequately surveyed.

Microcystina 'tempurung 3'
Distribution. Known from Gunung Tempurung, Perak only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell somewhat similar to Microcystina townsendiana but differ in having wider whorls and the different manner of spiral and radial sculpture. Figure 20E Materials examined.  Distribution. Widespread in Peninsular Malaysia (Maassen 2001). Native to East and Southeast Asia but now pantropical (de Winter et al. 2009).

Microcystina townsendiana Godwin-Austen & Nevill, 1879
Remarks. Medium-sized shell. Globular shell. Spire height low to medium. Umbilicus present. Periphery reflected when mature. Shell white to brown, occasionally with dark brown peripheral band. This is a synanthropic species.

Distribution. Known from upper Kinta Valley only.
Remarks. Small shell. Spire tall. Periphery rounded, almost keeled. Radial ribs pronounced, spaced out and equidistant from each other. Spiral ridges distinct but not as pronounced as radial ribs. This species was previously presumed endemic to Gunung Kanthan .  Distribution. In Peninsular Malaysia, known from Penang and Perak (Maassen 2001). Elsewhere, in Salanga (=Phuket), Thailand (von Martens, 1883).
Distribution. Known from Bukit Kepala Gajah, Perak only, but surrounding hills have yet to be adequately surveyed.
Distribution. Known only from Prk 53 Hill KF only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell medium-sized, brown. Periphery strongly keeled. Dorsal whorls have dense and pronounced sradial ribbing while the umbilical side is glossy with fine radial growth lines.
Remarks. Shell medium-sized. Brown and non-glossy at dorsal whorls, white and glossy at umbilicus. Periphery keeled but less pronounced than Pseudoplecta bijuga. Whorls have dense and fine radial growth lines. This is a synanthropic species.
Distribution. Known from Gunung Tempurung only, but surrounding hills have yet to be adequately surveyed.
Remarks. Shell small. Spire low, shell discoid. Distinguished from other discoid snails by the presence of strong spiral ridges of equal width. In between these spiral ridges, fine round pits are deposited in a neat matrix. Differs from Glyptaulax artificiosa (Benson, 1856)   Distribution. In Peninsular Malaysia, known from Pahang (Maassen 2001) and Perak. Elsewhere, ranges from Sumatra to Maluku, in Indonesia (Vermeulen and Whitten 1998).
Remarks. Distinguished from Philalanka pusilla by the white shell, taller spire, two to three major spiral ridges with many minor spiral lines and less changes in the angularity of whorls. New record for Perak.    (Maassen 2000).

Philalanka pusilla Maassen, 2000
Remarks. Shell smaller and flatter than Philalanka kusana, spire low. Shell yellow to light brown. Five or more major spiral ridges and more changes in angularity of whorls. New record for Peninsular Malaysia.

Family Euconulidae Baker, 1928 Genus Kaliella Blanford, 1863
Kaliella barrakporensis (Pfeiffer, 1852) Figure 26C Materials examined. Distribution. In Peninsular Malaysia, known from Perak only although it is likely widespread. Elsewhere, distributed from Africa in the west to Flores, Indonesia in the east (Vermeulen et al. 2015).
Remarks. Shell small, conical. Differ from congeners in its fine but distinct radial sculpture, less dense coiling of whorls and slightly convex whorls. Figure 26D Materials examined. Distribution. In Peninsular Malaysia, currently known from Penang, Perak and Selangor although it is likely widespread. Elsewhere, distributed from India in the west to Australia and Tahiti in the east (Vermeulen et al. 2015).

Kaliella calculosa (Gould, 1852)
Remarks. Shell slightly larger than Kaliella barrakporensis. Differs from sympatric congeners in its fewer but more convex whorls, especially obvious in the ultimate whorl. Radial sculpture fine but distinct. Spiral sculpture usually present. Figure 27A Materials examined. Distribution. In Peninsular Malaysia, known from Perak, Johor (Pulau Aur) and Kelantan although it is likely widespread (Maassen 2001). Elsewhere, from Sundaland to Australia, Fiji and Samoa (Vermeulen et al. 2015).

Kaliella scandens (Cox, 1871)
Remarks. Shell shape similar to Kaliella doliolum (Pfeiffer, 1846). Radial rib density highly variable, from very fine growth lines to coarse ribs. Distinguished from K. doliolum by the presence of fine spiral lines and absent or indistinct radial ribs at the umbilical section of the whorls. This is a synanthropic species.
Distribution. Known from Gunung Tempurung only, but surrounding hills have yet to be adequately surveyed.
Remarks. Distinguished from other Southeast Asian congeners by its convex whorls at both dorsal and umbilical sides, sharp peripheral keel, dense and pronounced radial ribs as well as glossy, spiral lined umbilical whorls. First record of Rahula in Peninsular Malaysia.
Remarks. A distinctive species with a thin, bullet-shaped shell. Spire tall, suture indistinct, whorls smooth and glossy. Aperture elongated and narrow.
Remarks. Shell size varies. Differs from Discatemon leptoglyphus in the prominence of lamella, the absence of a peripheral keel and having more bulbous whorls. Prior to this, Davison (1991) reported D. plussensis from Gunung Tchehel and a few hills near Sungai Siput North only.
Remarks. Shell small, spire tall. Whorls slender than Allopeas clavulinum. Spiral growth lines fine. This is a synanthropic species.
Remarks. Distinguished from all confamilials by its large shell, rounded apical whorls, fine but pronounced radial ribs and angular lower periphery.
Remarks. Distinguished from Peninsular Malaysian congeners by its distinct spiral lines at apical whorls, becoming less at post-apical whorls. Radial sculpture more prominent at post-apical whorls.   Distribution. Found on limestone karsts across Peninsular Malaysia (Maassen 2001). Elsewhere, in southern Thailand (Maassen 2003).
Remarks. Distinct among congeners in its tall spire, keeled periphery, very long trumpet-shaped ultimate whorl and apertural teeth arrangement. Shell colour varies from dark to light brown.  Remarks. Spire tall but fewer than Gyliotrachela hungerfordiana. Penultimate whorl very expanded and periphery keeled, giving a depressed shell shape. Ultimate whorl trumpet-shaped, with aperture face at 45 degrees of the plane perpendicular to the coiling axis. Apertural teeth arrangement unique.
Distribution. Known from Gunung Tempurung, Perak only, but surrounding hills have yet to be adequately surveyed.
Remarks. Differ from Paraboysidia kelantanensis tenuidentata van Benthem Jutting, 1950, of Kramat Pulai by its gradually flatter and more expanded whorls, periphery rounded, shallow suture, short trumpet-like extension of ultimate whorl and apertural teeth arrangement.    Distribution. In Peninsular Malaysia, known from Perlis and Perak (Maassen 2001). Elsewhere, in the tropical regions of the Indian Ocean and the Pacific Islands (Seddon 2000).
Remarks. Distinguished from sympatric congener Ptychopatula orcula by its smaller shell, tighter whorls, narrower umbilicus and tall spire.

Discussion
Apart from providing a comprehensive assessment of land snail diversity for the limestone hills in and around the Kinta Valley of Perak, our study has also elucidated interesting biogeographical patterns to assist in conservation planning. Our study has also paved the way for more in-depth land snail taxonomy studies to be conducted. For instance, many of the land snail species recorded by Davison (1991) and Clements et al. (2008b) could not be assigned to scientific names that have been published, until now. Speciose genera such as Microcystina, Diplommatina and Opisthostoma, require critical taxonomic revisions that need considerable effort. We hope unnamed species in these genera, now with traceable reference specimens collected from our study, can be described in the near future.

Land snail species diversity and biogeographical patterns
Correlation tests initially suggested no statistically significant relationship between the number of unique species and the degree of limestone hill isolation. However, when examined closely within a geographical context, the number of unique species for each hill could also be associated with the degree of isolation of major limestone clusters of each hill (Table 1). For example, Prk 55 G. Pondok, Prk 64 Bt Kepala Gajah, and Prk 01 G. Tempurung all located in the limestone clusters at the periphery of Kinta Valley and at least 50 km away from other limestone clusters outside of Kinta Valley (see Figure 6 in Liew et al. 2016).
At this point in time, our assignment of unique species should be considered "unique" to a particular hill in the context of hills that have been surveyed systematically in our study. The degree of endemism of a species depends on the geographical context, e.g. endemic to Malaysia, Perak, Kinta Valley, hill clusters or a single hill. For example, Charopa lafargei which previously presumed as endemic to Gunung Kanthan , is shown in our study to also occur on the limestone hills at the north of Kinta Valley. Also, Hydrocena semisculpta which was thought be endemic to Gunung Pondok (Davison 1991), is a relatively widespread species -can be found in 11 out of the 12 hills surveyed in our study.
Our comprehensive inventory has corroborated Davison's (1991) hypothesis that land snail assemblages on limestone hills in Perak can be divided into two groups, one comprising of hills within Kinta Valley and the other comprising of hills scattered in the area north of Kinta Valley. This biogeographical pattern cannot be explained by geographical distance between different limestone hills alone. For example, although Bat Cave Hill and Batu Kebelah Hill (northern cluster) are closest to Prk 53 Hill KF, the land snail assemblage of Prk 53 Hill KF is more similar to the land snail assemblages in the southern cluster hills.
It seems that biogeographical patterns of land snail assemblages at limestone hills in and around the Kinta Valley of Perak were influenced by the geology and topography of the area. For instance, all the hills in the southern cluster are located in the lowlying floodplains of the Kinta valley, while the hills of the northern cluster are located on hilly granitic areas north of the Kinta valley floodplain. As such, this biogeographical pattern could be a result of geological isolation (de Morgan 1885a; van Benthem Jutting 1960;Davison 1991).
The rocks in Perak (including limestone) first emerged in the Mesozoic -the Late Jurassic period (van Benthem Jutting 1960;Metcalfe 2013). After that, limestone hills in Perak became isolated due to two different processes that produced the two biogeographical regions (i.e. limestone hill clusters) (Paton 1961). For the northern hill cluster, the isolation of hills was likely to be caused by structural deformity with intrusion of granite and these hills were probably never contiguous from the start (Paton 1961). For the southern hill cluster, the isolation of hills was probably caused by limestone erosion, recent marine incursion and alluvial deposition (Paton 1961). However, our hypothesis of geological isolation being the main determinant of biogeographical patterns observed in our study requires further tests using phylogeographic approaches, where the divergence time of a particular land snail species can be estimated (by sampling populations from different limestone hills) and matched with historical geological events, including the more recent local geological events or climate changes in the Pleistocene.

Conservation implications
Despite containing the second highest number of limestone hills in Peninsular Malaysia after the State of Kelantan (see Liew et al. 2016), the State of Perak appears to be experiencing the greatest threat from limestone quarrying with 45 out of 78 limestone quarries in the peninsula being located in this State as of 2014 (Minerals and Geoscience Department Malaysia 2014). As such, there is an urgent need to identify limestone hills with high levels of species endemism for conservation prioritisation within this State. Four hills, namely Prk 01 G. Tempurung, Prk 55 G. Pondok, Prk 47 Kanthan, and Prk 64 Bt Kepala Gajah, were found to collectively contain 91% of the total land snail species occurring at 12 hills sampled in our study. In order to prevent the extinction of unique land snail species, these hills require immediate attention from conservation practitioners to mitigate ongoing threats from quarrying and surrounding forest loss.

Conclusion
Our study has updated the state of knowledge of land snail diversity and biogeography on limestone hills in Perak, Peninsular Malaysia. Larger limestone hills seem to have more unique species, of which several could be site endemic. In addition, we have shown that land snail assemblages can be divided into two groups that occupy limestone hills at two different areas in Perak -this was likely caused by different geological processes that separated the limestone hills. Most importantly, our findings can be used to assist in limestone conservation planning, especially within the Kinta Valley of Perak.