Land snails of Leptopoma Pfeiffer , 1847 in Sabah , Northern Borneo ( Caenogastropoda : Cyclophoridae ) : an analysis of molecular phylogeny and variations in shell form due to geography

The Leptopoma is a species rich genus with approximately 100 species documented 9 according to shell morphology and animal anatomy. Many. The identification of the 10 Leptopoma species are described in terms ofwas mainly dependent on shell morphology (e.g. 11 shell size, shape, sculpture and colour patterns) and animal anatomy of a small number of 12 examined materials. However, the implications of the interand intra-species variations in shell 13 form to the taxonomy of Leptopoma species and the congruency of its current shell based 14 taxonomy with its molecular phylogeny are still unclear. Over the last decade, more than 900 15 collection lots consisting of more than 4000 Leptopoma specimens have been obtainedThere 16 are four Leptopoma species found in Sabah, Borneo, and deposited in BORNEENSIS at 17 Universiti Malaysia Sabah. Accesstheir taxonomyic status remains uncertain due to this 18 collection gave us the opportunity to examine the geographical variationssubstantial variation 19 in shell forms and the . In tThis study, focuses on the phylogenetic relationships and 20 geographical variation in shell form of three Leptopoma species from Sabah were focused in 21 the study of phylogenetic relationship of Leptopoma speciesand geographical variation in shell 22 form based on more than 4000 Leptopoma specimens collected between 2000 and 2016 in 23 Sabah. The phylogenetic relationship of three Leptopomathese species was first estimated by 24 performing maximum likelihoodMaximum Likelihood and Bayesian analysis based on 25 mitochondrial genes (16S and COI) and nuclear gene (ITS-1). After thisThen, a total of six 26 quantitative shell characters (i.e. shell height, shell width, aperture height, aperture width, shell 27 spire height, and ratio of shell height andto width) and three qualitative shell characters (i.e. 28 shell colour patterns, spiral ridges, and dark ringapertural band in aperture) of the specimens 29 were mapped across the phylogenetic tree and tested for phylogenetic signals. Data on shell 30 characters of Leptopoma sericatum and Leptopoma pellucidum from two different locations 31 (i.e. Balambangan Island and Kinabatangan) where both species occurred sympatrically were 32 then obtained to examine the geographical variations in shell form. The molecular phylogenetic 33 analyses suggested that each of the three Leptopoma species was monophyletic and indicated 34 congruence with only one of the shell characters (i.e. shell spiral ridges) in the current 35 morphological-based classification. Other qualitative and quantitative shell characters were 36 incongruent with the Leptopoma species phylogeny. Although the geographical variation 37 analyses suggested some of the shell characters indicating inter-species differences between 38 the two Leptopoma species, these also pointed to intra-species differences between populations 39 from different locations. This study provides an initiation to resolve the taxonomy conundrum 40 for the remaining 100 little known Leptopoma species from other regionsThis is the first study 41 on Leptopoma species is based on small sample sizes and the findings appear only applicable 42 Formatted: Font: Not Italic Commented [R1]: It is not quite clear whether you describe the past situation or your current study. How about something like “Species-level identification in this group has been based on shell morphology and colouration, as well as some anatomical features based on small sample sizes.”

two Leptopoma species, these also pointed to intra-species differences between populations from different locations. This study provides an initiation to resolve the taxonomy conundrum for the remaining 100 little known Leptopoma species from other regions and highlights a need to assess variations in shell characters before they could be used in species classification. 235 Two-way ANOVA showed that all shell quantitative characters (except aperture height) differed 236 between the two locations (Table 2). In addition, all shell quantitative characters except shell width 237 and aperture height also differed between the two species. There was interactive effect of species 238 and location on the aperture height, shell spire height and ratio between shell height and width.

240
Chi-square analyses indicated significant association between the frequencies of shell 241 colour patterns and both the factors of species identity (Fisher's exact test: p =0.0000) and location 242 (Fisher's exact test: p=0.0000). Similarly, there was significant association between the 243 frequencies of the presence of the dark ring band in the aperture and both the factors of species 244 identity (Pearson's Chi-Squared with Yates' continuity correction: X 2 (1, N=249) = 4.88, 245 p=0.0271) and location (Pearson's Chi-Squared with Yates' continuity correction: X 2 (1, N=249) 246 = 12.910, p=0.0003). Both contingency tables are available in Table S5 & S6 in Supplementary 247 File 2. Overall, the shell characters did not show consistent differences between L. pellucidum and 248 L. sericatum since the differences in shell form were coupled with geographical variations and 249 interaction effects between geography and species. 250 251 DISCUSSION 252 Although Cyclophoridae represents the most diverse family, it is also one of the less taxonomically 253 resolved Caenogastropoda families. The current classification of many Cyclophorids was based 254 solely on morphology characteristics where its reliability remains doubtful as this taxa possesses 255 exceptionally diverse variations in morphology. Past research has shown that molecular 256 phylogenetic analyses could provide insights into the taxonomy of morphologically-ill land snails 257 in this region (Nantarat et al., , 2014bLiew et al., 2009;Liew et al., 2014). This study 258 presents the first molecular phylogeny investigation on genus Leptopoma in Sabah, one of the 259 understudied taxa within Cyclophoridae, and examines the concordance between morphology and 260 phylogeny as well as geographical variations in shell form.

262
All the phylogenetic trees based on different genes were congruent and provided significant 263 support for the monophyly of three morphologically similar Leptopoma species in Sabah namely 264 L. pellucidum, L. sericatum and L. atricapillum. The phylogenetic placement of genus Leptopoma 265 in Sabah was in concordance with its traditional morphology-based classification. For example, 266 the placement of Leptopoma pellucidum 6014 (Fig. 3) in this study, previously assumed as L. 267 vitreum due to its white colour shell which differed from other L. pellucidum, was revealed as 268 within the L. pellucidum clade which supported Vermeulen (1999)'s decision to assign L. vitreum 269 as synonymous to L. pellucidum. In the case of L. pellucidum and L. sericatum, Vermeulen 270 separated them into two species provisionally due to the existence of intermediate forms between 271 the two species. In this study, results suggested that the two species could be unequivocally 272 regarded as separate. In short, the findings of this study are in line with past research which 273 proposed that a combination of morphology and molecular approaches could improve taxonomy 274 of land snails.

276
A morphological character is assumed to have strong phylogenetic signal when the same 277 character clusters together within closely-related species . This could be a 278 useful diagnostic indicator for species delimitation. The phylogenetic signal tests showed that 279 spiral ridges had a significant phylogenetic signal (λ=1, K>1). Distinct spiral ridges were present 280 in L. sericatum and L. atricapillum while L. pellucidum had weak spiral ridges. This indicated that 281 weak spiral ridges might be an automorphy character for L. pellucidum which could be useful in 282 discriminating L. pellucidum from L. sericatum and L. atricapillum. This result was in agreement 283 with  where spiral ridges were also used as a key to delimitate between L. 284 pellucidum and L. sericatum.

286
The presence of a dark ring band in the aperture of land snails has not been observed in 287 other Cyclophorids and was not mentioned in other revision works of Leptopoma species. A 288 phylogenetic signal test showed that the presence of a dark ring band exhibited a significant 289 phylogenetic signal. However, this character was found to be strongly affected by geographical 290 variations when two species from two different locations were compared. All shells with a dark 291 ring band located in the shell aperture were collected from a single location in Kinabatangan, i.e. 292 the Tabin Wildlife Reserve area. The dark ring band was presented in both species with L. 293 pellucidum showing more instances than L. sericatum. The underlying causes of the presence of 294 this shell character remain yet unknown. Compared to results from phylogenetic signal test, the 295 presence of a dark ring band in the shell aperture would not be a reliable character to distinguish 296 between Leptopoma species due to geographically-induced morphology variations.

298
Shell colour patterns are usually used as key determinants to discriminate between species 299 in traditional morphology classification. One of the sister taxa of Leptopoma, the species in genus 300 Cyclophorus, was distinguished unambiguously based on shell patterns that were also supported 301 by molecular data (Nantarat et al., 2014b). Compared to genus Cyclophorus, shell colour patterns 302 of the genus Leptopoma, particularly in L. sericatum and L. pellucidum, exhibited a weak 303 phylogenetic signal. This case of shell colour pattern polymorphisms of the two Leptopoma species 304 is similar to other well-known land snails namely Cepaea nemoralis and C. hortensis (Owen & 305 Bengtson, 1972; Ozgo & Schilthuizen, 2012; Cameron & Cook, 2012; Cameran, 2013). However, 306 unlike Cepaea land snails that have been studied extensively, the causal mechanism for the 307 Leptopoma land snail's diverse shell colour patterns is still unknown. This study also revealed that 308 the Leptopoma species exhibits idiosyncratic differences between locations in the degree of shell 309 polymorphisms. For example, the Leptopoma population at Balambangan Island has more shell 310 colour patterns as compared to the population at Kinabatangan. As a result, the geographically-311 induced variations in shell colour patterns and weak phylogenetic signal strongly suggest that shell 312 patterns should not be used as a diagnostic character for the genus Leptopoma. . From the phylogenetic signal 317 test, only shell height produced a significant signal. In the Vermeulen (1999) description of L. 318 sericatum and L. pellucidum, the ratio between shell height and width of L. sericatum is slightly 319 smaller than L. pellucidum. This study revealed a high degree of geographical variations in the 320 quantitative shell characters; for example, both Leptopoma species from Balambangan Island were 321 larger than the same species found in Kinabatangan. Previous studies suggested that land snails 322 found on islands tend to undergo extensive morphological diversification (Johnson & Black, 2000;323 Stankowski, 2011). In view of this, quantitative shell characters are thus not advisable as a 324 diagnostic indicator for species delimitation due to the strong influence of geographical variations.

326
This study has only revealed partial information on the phylogenetic and morphology 327 variations of all Leptopoma species in their entire distribution range. However, despite its small 328 geographical scale, the study has resolved taxonomic uncertainties of three Leptopoma species and 329 revealed notable variations in both the quantitative and qualitative shell characters for the species. 330 From the findings it is clear that any attempt of the taxonomy works on the rest of ca. 100 331 Leptopoma species in the future should consider the possible caveats in using the shell characters 332 as the sole evidences and should instead include molecular phylogeny in the study. 333 CONCLUSION 334 This study represents the first attempt to conduct phylogenetic investigation into the genus 335 Leptopoma and provides phylogenetic assessment of the genus in Sabah. The results 336 unambiguously separate L. pellucidum, L. sericatum and L. atricapillum into three distinct 337 monophyletic groups, and highlight substantial congruence among the traditional morphological 338 classifications based on spiral ridges and molecular phylogeny of the Leptopoma species in Sabah. 339 After performing the phylogenetic signal tests, it can be stated that all quantitative and many 340 qualitative shell characters are not reliable diagnostic indicators for discriminating between the 341 Leptopoma species due to the considerable geographical variations in shell form. This study 342 represents an attempt to resolve the taxonomy conundrum for the remaining 100 little known 343 Leptopoma species from other distribution regions. Further studies that include more samples from 344 a wider geographical reach are recommended.    Tree as in Figure 3, which only 14 adult of the three Leptopoma species were included whereas the juvenile specimens together with other outgroup taxa were dropped from the tree. Different categories of the three qualitative shell characters: spiral ridges, shell colour patterns and dark ring band in aperture (referred to Figure 1A, 1B, 1C respectively) were represented by different colour of the squares; and the six shell quantitative measurements: