Phylogenetic insights into the terrestrial snails Helicoidei (Gastropoda: Stylommatophora) with special emphasis on the Camaenidae (2024)

Search for other works by this author on:

Article history

Abstract

Introduction

The Helicoidei Rafinesque, 1815 are a widely distributed and diverse infraorder of land snails and slugs distributed in Asia, Australia, North and South America, Europe, and Africa. This infraorder includes two superfamilies, Helicoidea Rafinesque, 1815 and Sagdoidea Pilsbry, 1895 (Bouchet et al. 2017).

Early studies on the taxonomy of the infraorder Helicoidei (known as the superfamily Helicoidea in classifications before Bouchet et al. 2017) were based on morphology. Thiele (1931) divided the Helicoidea into three families, the Bradybaenidae, Camaenidae, and Helicidae. Zilch (1960) followed Thiele’s classification, and Taylor and Sohl (1962) introduced two new families to the Helicoidea, the Oreohelicidae (now Punctoidea) and Helminthoglyptidae. Boss (1982) removed the Oreohelicidae but added Ammonitellidae (now Rhytidoidea) to the Helicoidea. Tillier (1989) used quantitative methods to construct a parsimony tree of the Helicoidea based on morphological characters. He added six new families to the Helicoidea, the Helminthoglyptidae, Vitrinidae (now Limacoidea), Helicarionidae (now Limacoidea), Polygyridae, Sagdidae, and Haplotrematidae (now Haplotrematoidea), and suggested that the Camaenidae might be a polyphyletic group. However, his work was challenged by Emberton and Tillier (1995), who disagreed with his definition of some morphological characters. Schileyko (1991) based his taxonomic system on genital characters. His Helicoidea s.l. included the superfamilies Helicoidea (which included the families Helicidae, Humboldtianidae, and Elonidae) and Xanthonychoidea [which included the families Helicostylidae, Bradybaenidae, Xanthonychoidae, Epiphragmophoridae, Helminthoglyptidae, Ciliellidae, Helicarionidae (now Helicarionoidea), and Hygromiidae]. In 1998, Schileyko (1998) raised the family Sagdidae to the superfamily Sagdoidea, comprising a single family, Sagdidae.

Molecular data have challenged the traditional morphology-based taxonomy of the Helicoidei (formerly Helicoidea in classifications before Bouchet et al. 2017). The first molecular study on the Helicoidea was conducted by Wade et al. (2001), who showed that the superfamily Helicoidea were a monophyletic group. Their phylogeny included seven families: the Helminthoglyptidae, Helicidae, Polygyridae, Camaenidae, Bradybaenidae, Hygromiidae, and Helicellidae. However, the family Camaenidae as traditionally defined (Pilsbry 1985) were not monophyletic (Wade et al. 2001). Following the results of Wade et al. (2001), Bouchet and Rocroi (2005) divided the Helicoidea into 19 families. They still considered the Camaenidae as a family but placed the American camaenids in the Pleurodontidae, following the opinion of Cuezzo (2003). Wade et al. (2006) and (2007) conducted further molecular analyses on the Helicoidea, which confirmed the non-monophyly of the Camaenidae. They also showed that the Camaenidae from America were not a single group but rather split into three distinct lineages: one clustering with the Hygromiidae and the others clustering with the Sagdidae. The Camaenidae from Asia/Australia in the study by Wade et al. (2007) clustered with the Bradybaenidae. Thompson and Naranjo-García (2012) introduced a new family of Mexican slugs, the Echinichidae, which were previously included in Xanthonychidae s.l. (=Xanthonychoidea), to the Helicoidea. Sei et al. (2014) found that the Trichodiscinidae clustered with Pleurodontidae rather than Xanthonychidae, and Trichodiscinidae were subsequently separated from the Xanthonychidae by Bouchet et al. (2017). Razkin et al. (2015) then reorganized the classification of the Hygromiidae, splitting it into three monophyletic families: Canariellidae, Geomitridae, and Hygromiidae. They also placed the Cochlicellidae as a tribe within the Geomitridae. Sei et al. (2017) proposed a major revision of the Helicoidea based on a comprehensive molecular phylogeny. They described a new family, the Labyrinthidae, belonging to the Helicoidea and expanded the Sagdoidea to include the families Sagdidae, Solaropsidae, and Zachrysiidae. Calcutt et al. (2020) also supported the grouping of these three families in the Sagdoidea, but the internal topology differed from that of Sei et al. (2017).

The taxonomy of the Helicoidei has been revised several times in recent years, with the most up-to-date taxonomy and nomenclature provided by Bouchet et al. (2017). One of their most significant changes was the merger of the Bradybaenidae and Camaenidae into a single family, named the Camaenidae, based on the phylogenetic studies by Scott (1996), Wade et al. (2006), and Wade et al. (2007), which showed that the Camaenidae were not a natural group, with the American Camaenidae being more closely related to Sagdidae or Hygromiidae than to other Camaenidae and with the Asian and Australian Camaenidae being more closely related to the Bradybaenidae. Gittenberger et al. (2012) and Bouchet et al. (2017) agreed that the Bradybaenidae should be considered a synonym of the Camaenidae. Another change was to move the Echinichidae, a group of slugs from Mexico, from the superfamily Xanthonychoidea to the Helicoidea (Bouchet et al. 2017). Bouchet et al. (2017) also merged five families that formed a clade [Epiphragmophoridae, Helminthoglyptidae, Humboldtianidae, Monadeniidae, and Xanthonychidae (based on Kohler’s unpublished phylogenomic tree)], treating them as one family, the Xanthonychidae. Bouchet et al. (2017) also placed the Halolimnohelicidae under the Hygromiidae and accepted two new families, Canariellidae and Geomitridae, following Razkin et al. (2015). Finally, Bouchet et al. (2017) created an infraorder, the Helicoidei, which contained two superfamilies, Helicoidea and Sagdoidea. The Sagdoidea included the American camaenid snails that clustered with Sagdidae in the phylogenetic tree of Wade et al. (2007) (Sei et al. 2017). The other American camaenid group, including Theliodomus and Pleurodonte and sister to the Hygromiidae in the tree of Wade et al. (2007), are now treated as Pleurodontidae (Bouchet et al. 2017).

Following dependence on shell characters and before the application of molecular methods, the taxonomy of the Camaenidae was based on the presence or absence of darts and associated morphological features in the genital system. The Camaenidae were based on the lack of darts and associated organs, whereas the Bradybaenidae were recognized by possession of dart sac(s) and related organs, such as mucous glands. However, these criteria do not reflect the relationships of these snails, based on molecular phylogenetic studies. Wade et al. (2007) found that the family Bradybaenidae was not a monophyletic group, and consequently, Gittenberger et al. (2012) treated Bradybaenidae as a synonym of Camaenidae. The current Camaenidae (following revisions by Bouchet et al. in 2017) form a diverse family of snails that live in various habitats in Asia and Australasia (Scott 1996, Bouchet et al. 2017) that consists of four subfamilies, the Camaeninae, Bradybaeninae, Hadrinae, and Helicostylinae. The Camaeninae were the first subfamily to be defined by Pilsbry (1895) as helices without a dart apparatus, with Camaena Pilsbry, 1895 as the type genus. The Camaeninae are distributed in East Asia and Southeast Asia. The subfamily Bradybaeninae were later established for many Chinese bradybaenid species by Pilsbry (1934) as helices that possessed dart structures, with Bradybaena Pilsbry, 1934 as the type genus. The Helicostylinae were introduced by Ihering (1909) for snails that possess a dart sac and are restricted to the Solomon Islands, the Philippines, and New Guinea, with type genus Helicostyla Férussac, 1821. The Hadrinae were proposed by Iredale (1937) for Australian helicid species without dart sacs and were based on the type genus Hadra E. von Martens, 1860. Bouchet et al. (2017) considered both the Rhagadinae and Sinumeloninae, which were previously recognized as subfamilies of Camaenidae (Bouchet and Rocroi 2005), as synonyms of the Hadrinae.

Morphological characters, including genital characters, are considered unstable characters, with several studies revealing incongruences between morphological and molecular approaches (e.g. Hirano et al. 2014). Moreover, some species that were classified based on shell characters have been found to have different genital features than were predicted from their shell characters. For instance, Trichelix diminuta [=Moellendorffia diminuta], which was formerly assigned to the Bradybaenidae, lacks a dart sac (Sutcharit et al. 2020), whereas Traumatophora, which was previously placed in Camaeninae, possesses a dart sac (Wu 2019). These are two examples of revised generic assignments based on genital characters that are consistent with molecular evidence but not with attributions based solely on shell characters.

To date, work on Chinese Camaenidae has been limited. Wu and Guo (2003) and Wu (2004) undertook morphological studies of the Chinese Camaenidae, and Wu (2019) introduced several useful genital characteristics. However, there is a lack of consistency in the use of terminology. For example, the sac opposite the vagin* in the study by Wu (2004) is named the proximal accessory sac by Wu (2019) and Zhang et al. (2021a). Moreover, some new species were described without considering previous literature, as noted by Wu and Qi (2009) and Zhang (2019), with the species range and morphological characters of the type material not being considered adequately. Platypetasus is treated as a synonym of Pseudiberus based only on whorl numbers and on shell and umbilicus measurements (Wu and Qi 2009). Moreover, the type species of Pseudiberus was recorded in the Taihang Mountains (Zhang et al. 2021a), whereas the type species of Platypetasus (Platypetasus innominatus) was recorded from near Kin-cha Kiang (=Jinsha Jiang) (Heude 1885); the geographical distribution of these taxa is completely different, and their distribution ranges do not overlap. Even some morphological characters can be distinguished, e.g. inner thickness near the aperture (Heude 1885, Wu and Qi 2009, Zhang et al. 2021a).

In this study, we: (i) reconstruct a phylogenetic tree based on the large subunit (LSU) ribosomal RNA (rRNA) gene region using Bayesian inference (BI) and maximum likelihood (ML) methods; (ii) examine the taxonomic system of the Helicoidea proposed by Bouchet et al. (2017); and (iii) describe new taxa from China based on morphological characters and the molecular phylogeny.

Materials and methods

Biological material

The specimens used in this study are listed in Table 1.

DNA extraction, PCR amplification, and sequencing

DNA was extracted from foot muscle tissue using the CTAB method (Goodacre and Wade 2001). Approximately 4000bp of the LSU rRNA gene was amplified using the polymerase chain reaction (PCR). The LSU rRNA, embracing part of 5.8S and ITS2, was amplified by nested PCR using the primers LSU-1ii and LSU-12 to produce primary PCR products, then using the primary PCR products as the template for secondary PCR. Secondary PCR was used to amplify six fragments (A–F) using the primers LSU-1iii and LSU-3iii (fragment A), LSU-2ii and LSU-5ii (fragment B), LSU-4ii and LSU-7 (fragment C), LSU-6i and LSU-9i (fragment D), LSU-8ii and LSU-11ii (fragment E), and LSU-10i and LSU-12i (fragment F) (Wade and Mordan 2000, Fontanilla et al. 2017). PCR amplifications for both primary and secondary PCRs were performed using Promega GoTaq^® G2 Master Mix buffer (1 U TAQ, 0.2 µM primers, 200 µM dNTP, and 1.5mM MgCl₂), with 1 µL of DNA template added to 24 µL of 1× Master Mix buffer. The cycling conditions of the primary PCR were as follows: 96°C for 2min, followed by 35 cycles of 94°C for 30s, 50°C for 30s, and 72°C for 3min, then a final extension step at 72°C for 5min. The secondary PCR cycle conditions were as follows: 96°C for 2min, followed by 35 cycles of 94°C for 30s, 45°C for 1min, and 72°C for 2min, then a final extension step at 72°C for 5min. The PCR products were purified using the Qiagen gel extraction kit, and sense and antisense sequencing was performed by Macrogen Inc.

Sequence analysis

DNA sequence trace data for the 12 sequence fragments were assembled using DNASTAR SeqMan v.7.1.0 (Swindell and Plasterer 1997) and the STADEN package (Staden et al. 2000). After assembly, sequences were aligned by MAFFT v.7.505 (Nakamura et al. 2018), with the alignment improved manually in the Genetic Data Environment (GDE) package (Smith et al. 1994). Ambiguously aligned sequence regions were removed from the alignment before phylogenetic analysis using Gblocks v.0.91b (Castresana 2000) (maximum number of contiguous non-conserved positions of five, and minimum length of a block of three), then checked again in GDE. Three thousand, four hundred and twelve sites were deemed to be aligned unambiguously and were used for subsequent tree building.

Phylogenetic trees were constructed using BI and ML methods. The general time reversible (GTR) model incorporating gamma distribution (+Γ) was chosen as the best model by ModelTest-NG v.0.1.7 (Darriba et al. 2019) based on the Bayesian information criterion. Bayesian inference phylogenies were generated using MrBayes v.3.2.7 (Ronquist et al. 2012), with two runs for 5 000 000 generations, sampling every 100 generations, and a temperature of 0.05. The final tree and posterior probabilities (PPs) were determined based on the last 50% of trees (burn-in = 0.5) and majority rule consensus. The split frequencies before burn-in were <0.01. Maximum likelihood phylogenetic trees were generated using RAxML-NG v.1.1.0 (Kozlov et al. 2019), using a heuristic search strategy with 10 random plus 10 parsimony start trees and with subtree pruning and regrafting (SPR) branch swapping. Bootstrap analysis [transfer bootstrap expectation (TBE); Lemoine et al. 2018] was undertaken for the ML tree with 1000 bootstrap replicates and shown as a percentage. Here, ML = 100%, PP = 1.00 is considered as full support, ML > 90%, PP > .9 is considered as strong support, and ML > 70%, PP > .8 is considered good support; anything lower is considered as weak support.

GenBank accession numbers

GenBank accession numbers for the sequences generated in this study are recorded in Table 1.

Results

Molecular phylogenetic analyses

The Bayesian tree of the Helicoidei is shown in Figure 1. Both the Helicoidea (BI PP = 1.00 and 97% ML bootstrap) and the Sagdoidea (PP = 1.00, 100% ML) form strongly supported monophyletic groups in the tree, and the Helicoidea and Sagdoidea cluster as sister groups with full support (PP = 1.00, 100% ML). The Bayesian and ML phylogenetic trees are highly consistent.

Sagdoidea

The Sagdoidea include Solaropsidae, Sagdidae, and Zachrysiidae. The Solaropsidae, represented by Caracolus marginella (Gmelin, 1791) and Caracolus carocolla (L., 1758), form a fully supported monophyletic group in the tree (PP = 1.00, 100% ML). Likewise, the Sagdidae, represented by two species, Hispaniolana crispata (Férussac, 1821) [=Polydontes undulata] and Sagda sp., also form a fully supported monophyletic group (PP = 1.00, 100% ML). The Zachrysiidae contain only a single species, Zachrysia auricoma (Férussac, 1821), in our phylogenetic tree. Sagdidae and Zachrysiidae are sister groups with good support (PP = .99, 86% ML), and Solaropsidae is the sister to the clade comprising Sagdidae + Zachrysidae (PP = 1.00, 100% ML).

Helicoidea

The family Cepolidae is sister to all other Helicoidea with strong support (PP = 1.00, 97% ML). It is represented by a single taxon, Hemitrochus streatori (Pilsbry, 1889) [=Cepolis streatori], in our tree. The remaining Helicoidea can be divided into two sister groups. The first group comprises Hygromiidae s.l., Xanthonychidae, Pleurodontidae, Helicodontidae, and Helicidae with strong support (PP = .94, 91% ML). It is, in turn, divided into two sister groups, the Helicidae with strong support (PP = 1.00, 99% ML) and the Hygromiidae s.l., Xanthonychidae, Pleurodontidae, and Helicodontidae with good support (PP = .94, 75% ML). The second group comprises the Camaenidae + Bradybaenidae + Polygyridae and is strongly supported (PP = 1.00, 96% ML).

Helicidae

The Helicidae consist of a sister-group relationship between the Helicinae and Murellinae + Ariantinae. The Helicinae are strongly supported (PP = .90, 94% ML), but the grouping of Murellinae and Ariantinae has weak support (PP = .66, 65% ML). The Ariantinae are weakly supported in our analyses (PP = .55, 62% ML), and the Murellinae contain only a single species in our phylogenetic tree.

The subfamily Helicinae forms a polytomy, comprising three lineages in the phylogenetic tree. The first lineage exhibits a sister-group relationship between Helix spp. and Caucasotachea vindobonensis (C. Pfeiffer, 1828) [=Cepaea vindobonensis], which has only weak support in BI (PP = .51); ML does not support this topology(Supporting Information, Fig. S1). The second lineage comprises a single species, Eremina desertorum (Forskål, 1775). The final lineage is composed of all remaining Helicinae species and is strongly supported (PP = 1.00, 98% ML). In this lineage, (Macularia sylvatica (Draparnaud, 1801) [=Cepaea sylvatica] + Allognathus minoricensis (Mittre, 1842) + Cepaea spp.) is sister to (Theba pisana (Müller, 1774) + Cornu aspersum (Müller, 1774) [=Helix aspersa] + Cantareus apertus (Born, 1778) + Eobania vermiculata (Müller, 1774) + Otala lactea (Müller, 1774)) (PP = 1.00, 98% ML). Allognathus minoricensis (Mittre, 1842) is sister to Cepaea spp. with full support (PP = 1.00, 100% ML), and Macularia is the sister group of Allognathus and Cepaea spp. with strong support (PP = 1.00, 91% ML). Theba pisana is the sister group of Cornu aspersum [=Helix aspersa] + Cantareus apertus + Eobania vermiculata (Müller, 1774) + Otala lactea with strong support (PP = .96, 97% ML). Cornu is the sister to Cantareus with strong support (PP = 1.00, 99% ML). Eobania is sister to Otala with weak support in BI (PP = .59) but strong support in ML (79% ML). Cornu + Cantareus is the sister group of Eobania + Otala with strong support (PP = 1.00, 97% ML).

Marmorana scabriuscula (Deshayes, 1830) is the only representative of Murellinae in our phylogenetic tree. Arianta arbustorum (L., 1758) is sister to the other Ariantinae (PP = .66, 65% ML), and Isognomostoma isognomostomos (Schröter, 1784) and Helicigona lapicida (L., 1758) are sister groups with strong support in BI (PP = .90) but weak support in ML (55% ML).

Xanthonychidae

The Xanthonychidae contain Monadenia fidelis (Gray, 1834) and Helminthoglypta diabloensis (Cooper, 1869), forming a monophyletic group with good support in BI (PP = .80) but weak support in ML (54%).

Pleurodontidae

Thelidomus aspera (Férussac, 1821) is our only representative of Pleurodontidae, which is sister to Xanthonychidae with good support in BI (PP = .89) but weak support in ML (62%).

Helicodontidae

The Helicodontidae are sister to Hygromiidae s.l. + Xanthonychidae and Pleurodontidae with good support (PP = .94, 75% ML) and have only one species, Helicodonta obvoluta (Müller, 1774).

Hygromiidae s.l.

The Hygromiidae s.l. form a monophyletic group with good support (PP = .99, 78%). Hygromiidae s.l. are sister to Xanthonychidae and Pleurodontidae in the tree, with good support (PP = .95, 78% ML). Among Hygromiidae s.l., the Trochulininae (Hygromiidae) (PP = 1.00, 99% ML) are sister to other Hygromiidae s.l. and contain Trochulus striolatus (Pfeiffer, 1828) [=Trichia striolata], Monacha cantiana (Montagu, 1803), and Euomphalia strigella (Draparnaud, 1801). Geomitridae (PP = 1.00, 98% ML) and Hygromiinae (Hygromiidae) (PP = 1.00, 100% ML) are sister groups with good support (PP = .84, 77% ML). The Hygromiinae contain Monachoides vicinus (Rossmässler, 1842) and Perforatella bidentata (Gmelin, 1791). For Geomitridae, Geomitrinae (PP = 1.00, 80% ML) and Helicellinae are sister groups, and Ponentina subvirescens (Bellamy, 1839) and Cochlicella acuta (Müller, 1774) in Geomitrinae are sister taxa. The Helicellinae contain only one species, Cernuella virgata.

Camaenidae s.l.

The clade of Camaenidae + Bradybaenidae + Polygyridae is strongly supported (PP = 1.00, 96% ML). There is a polytomy at the base of the clade, which comprises: (i) Bradybaeninae (excluding Satsuma and Coniglobus), along with Hadrinae; (ii) a separate branch consisting of Satsuma and Coniglobus; (iii) Helicostylinae; (iv) Polygyridae; and (v) numerous individual lineages of Camaeninae genera, such as Nipponochloritis, Nesiohelix, Trichelix, Beddomea, and Amphidromus. The main Bradybaeninae clade including the Hadrinae (but excluding Satsuma and Coniglobus) is weakly supported in BI (PP = .64) but well supported in the ML analysis (73% ML). Within this, the following sister groups are recognized: Jiaoliaous tchefouensis (Crosse & Debeaux, 1863) and Pseudobuliminus dongyiicus Zhang, n. sp. (PP = 1.00, 93% ML); Euhadra and Mandarina (PP = 1.00, 97% ML); Cathaica pyrrhozona (Philippi, 1845) and Pliocathaica buvigneri (Deshayes, 1873) (PP = .98, 96% ML); and Plectotropis and Bradybaena (PP = .75, 55% ML). Ainohelix, Ezohelix, Manchurohelix, and Paraegista from North East Asia also cluster together in the tree (PP = .87, 64% ML). Satsuma and its previous subgenus, Coniglobus, cluster together, forming a monophyletic clade with full support (PP = 1.00, 100% ML), but fall separately from the main Bradybaeninae clade. Helicostylinae are a monophyletic group with full support (PP = 1.00, 100% ML). Polygyridae are also fully supported (PP = 1.00, 100% ML).

Nomenclature

Helicoidea Rafinesque, 1815
Bradybaenidae Pilsbry, 1898
Bradybaeninae Pilsbry, 1898
Jiaoliaous Zhang gen. nov.

[urn:lsid:zoobank.org:act:83C51892-AFB9-4CDB-B2BF-81E824BAD258]

Type species

Helix tchefouensis Crosse & Debeaux, 1863, original designation.

Jiaoliaous tchefouensis (Crosse & Debeaux, 1863) gen. nov., comb. nov.

The taxonomic description of Jiaoliaous tchefouensis can be found in the Supporting Information, with photographs of Jiaoliaous tchefouensis in the Supporting Information (Fig. S2).

Pseudobuliminus Gredler, 1886

Type species

Helix pseudobuliminus Heude, 1882, absolute tautonymy.

Pseudobuliminus dongyiicus Zhang sp. nov.

[urn:lsid:zoobank.org:act:D44A4F43-6A43-4C3B-95F8-74C4221B5ABD]

The taxonomic description of Pseudobuliminus dongyiicus can be found in the Supporting Information, with photographs of Pseudobuliminus dongyiicus in the Supporting Information (Fig. S3).

Discussion

Helicoidea

Camaenidae s.l.

The Camaenidae, Bradybaenidae, and Polygyridae cluster in our phylogeny with strong support to form the Camenidae s.l. (Fig. 1; Supporting Information Figs. S1 and S4) Relationships within this group are not well resolved, and the Polygyridae form one of the multifurcating branches alongside other camaenid and bradybaenid taxa in a polytomy at the base of the clade. Wade et al. (2007) showed that the Camaenidae from Asia and Australia were a separate group from the American taxa, with these American camaenids subsequently moved to the Sagdoidea and Pleurodontidae by Sei et al. (2017); an arrangement adopted by Bouchet et al. (2017). The remaining Asian and Australian camaenid taxa clustered with the Bradybaenidae in the tree of Wade et al. (2007); subsequently, Gittenberger et al. (2012) and Bouchet et al. (2017) synonymized the Bradybaenidae with the Camaenidae to form the Camaenidae s.l. (Fig. 1). We followed the subfamily classification of Bouchet et al. (2017), including four subfamilies, the Camaeninae Pilsbry, 1895, Bradybaeninae Pilsbry, 1934 (1898), Hadrinae Iredale, 1937, and Helicostylinae Ihering, 1909 (these subfamily relationships are consistent with those established in early taxonomic studies; Pilsbry 1895, 1934, Ihering 1909, Iredale 1937). In our phylogenetic tree, Hadrinae cluster with the Bradybaeninae (excluding Satsuma and Coniglobus), with Satsuma + Coniglobus, the Helicostylinae, the Polygyridae, and several other individual Cameaninae genera forming part of a polytomy at the base of the Camaenidae s.l.

Our phylogenetic tree differs in some respects from the mitochondrial phylogenies of Minton et al. (2016), Zhang et al. (2021b), and Zhao et al. (2023), from which there are three different hypotheses regarding the relationships within the Camaenidae s.l. The first hypothesis, based on the ML phylogenetic trees of Minton et al. (2016) and Zhang et al. (2021b), places the Polygyridae as sister to the Helicidae + Bradybaenidae + Camaenidae, and additionally the Geomitridae for Zhang et al. (2021b), while the Bradybaenidae are sister to the Camaenidae with strong support. The second hypothesis, based on the ML Gene-Order (MLGO) tree of Minton et al. (2016), shows that the Camaeninae, Bradybaeninae, and Polygyridae cluster together but with low support, with Bradybaenidae sister to Polygyridae instead of Camaenidae. The final hypothesis, based on the phylogeny of Zhao et al. (2023), places the Polygyridae as sister to the Camaenidae + Bradybaenidae with strong support, and the Camaenidae are sister to the Bradybaenidae with strong support. Our results for the Camaenidae s.l. are similar to those of Zhao et al. (2023), but the internal topology differs.

The subfamily groupings in our phylogenetic tree are consistent with currently established taxonomic classifications (Bouchet et al. 2017), but they are inconsistent with the phylogeny based on genital characters as proposed by Wu (2004). In Wu’s phylogeny, the Bradybaeninae are considered paraphyletic and cluster with the Helicostylinae under successive weighting, implied weighting, and extended implied weighting (Wu 2004; Supporting Information, Fig. S5B, C), although support is weak, there is no apomorphy, and only a single Camaeninae member was represented (Supporting Information, Fig. S5). Under equal weighting, the grouping of subfamilies was also unresolved, with weak support for branches and with no apomorphy. Bradybaeninae formed a polyphyletic cluster with the single Camaeninae member, Camaena. Helicostylinae remained monophyletic (Supporting Information, Fig. S5A). Examination of our and Wu’s (2004) parsimony phylogenetic techniques indicate that these differences are likely to stem from inaccuracies in the morphological character matrix (Goloboff and Morales 2023).

Based on both morphology (without the dart apparatus) and the molecular phylogenetic results of Wade et al. (2006) and Wade et al. (2007), it has already been noted in earlier literature, such as by Nordsieck (2017), that Satsuma is unique in comparison to other bradybaenid species from East Asia. The Bayesian tree of Sei et al. (2017) suggested that Satsuma is sister to other Camaenidae (excluding Satsuma in the study by Sei et al. 2017), Bradybaenidae, and Polygyridae. In this study, Satsuma s.l. (Satsuma including its subgenus, Coniglobus) is a monophyletic group distinguished from the lineage of Bradybaeninae and Hadrinae. Our conclusion is that Satsuma and its previous subgenera (Coniglobus, Luchuhadra, and Satsuma s.s.) should be recognized as a separate subfamily. Its members are endemic to Taiwan Island, the Ryukyu Islands, Shikoku Island, and Honshu Island (Hoso et al. 2010).

Although the Helicostylinae appear as a monophyletic group in our phylogenetic tree, the phylogenetic relationships within the Helicostylinae, as presented by Batomalaque et al. (2019) based on CO1, exhibit a greater level of complexity, consisting of three polytomic lineages. In the first lineage, Chloraea (Helicostylinae) clusters with Aegista spp. (Bradybaeninae). The second lineage comprises only Helicostylinae, and the final lineage, a subset of Helicostylinae (Phoenicobius and Chysallis), clusters with Nesiohelix (Camaeninae) and Satsuma from Taiwan. Our phylogeny contradicts the topology of Batomalaque et al. (2019), in that Aegista is shown to be sister to Bradybaena, unequivocally falling within the Bradybaeninae. This discrepancy suggests a need for further examination based on 28S of the monophyly of Helicostylinae, with an emphasis on increased sampling.

In our phylogenetic tree, Bradybaeninae and Hadrinae, represented in our phylogeny by a single species, Sphaerospira fraseri from East Australia, cluster together to form a monophyletic group. The Bradybaeninae species within this lineage are mostly distributed in East Asia, with only a single genus, LandouriaGodwin-Austen (1918), ranging across the southern Himalaya. Landouria radleyi has a disjunct occurrence in the Central Highlands of Sri Lanka. However, the placement of eastern species in Landouria has been challenged based on morphological characters (Páll-Gergely et al. 2013).

The Hadrinae group with the Bradybaeninae in both BI and ML analyses. However, we note that Sphaerospira fraseri falls on a relatively long branch in comparison to other Bradybaeninae within the group, and there is therefore a possibility that this placement might be attributable to long branch attraction. Our equal-weighting parsimony phylogeny (Supporting Information, Fig. S4) weakly supports the monophyly of Bradybaeninae based on jackknifing support and suggests Hadrinae as a distinct polytomy branch of the Camaenidae s.l. Our parsimony tree is similar to the phylogeny reconstructed by Sei et al. (2017), in which Bradybaeninae are sister to Polygyridae, and Hadrinae are sister to Polygyridae and Bradybaeninae. Considering the geographical distribution, we encounter a conflict with our phylogenetic results. Wallace’s line remains a relevant barrier for terrestrial Mollusca (Hausdorf, 2019), and Bradybaeninae from East Asia and Hadrinae from Australia are separated by Wallace’s line. According to Razkin et al. (2015) and Sei et al. (2017), the Bradybaenidae separated from the common ancestor of Bradybaenidae + Satsuma ~47.49 Mya. This time point should be earlier than seperation of Bradybaeninae and Hadrinae based on our topology. However, at that time, Asia and Australia were no longer connected, conflicting with the hypothesis that the Bradybaeninae are close relatives of the Hadrinae.

As a whole, the Camaenidae s.l., including the Camaenidae, Bradybaenidae, and Polygyridae, are distributed in different regions of the world: East Asia including Taiwan Island, the Ryuku Islands and Shikoku Island (Bradybaeninae), Taiwan Island, Ryukyu Islands, Shikoku Island (subfamily including Satsuma and its previous subgenera), Australia (Hadrinae), Southeast and South Asia (Camaeninae), Philippines (Helicostylinae), and North America (Polygyridae). Preliminarily, our results match the original distribution definitions of these Bradybaenidae, Camaenidae, and Polygyridae subfamilies, with the exception of the Camaeninae (Pilsbry 1895, 1934, Ihering 1909, Iredale 1937), which was not defined by a specific region by previous authors, but in this study is mainly from Southeastern Asia. When discussing the internal grouping of Camaenidae s.l., the lack of resolution has to be addressed. One plausible explanation is that, following their colonization of a new geographical area and range of habitats, they underwent a phase of rapid diversification.

Hygromiidae s.l.

Our phylogenetic tree resolves the Hygromiidae s.l. as defined by Bouchet and Rocroi (2005) and includes Geomitrinae and Hygromiinae as monophyletic subfamily groups. Bouchet et al. (2017) subsequently accepted the phylogenetic results of Razkin et al. (2015), which split the Hygromiidae into three families, the Canariellidae, Geomitridae, and Hygromiidae. However, the Hygromiidae defined by Razkin et al. (2015) are not monophyletic in our phylogenetic trees, and the topologies in BI and ML are consistent, with good support. This can be resolved by recognizing what is a more stable level of family hierarchy in the broader context of the tree and synonymizing the Geomitridae with the Hygromiidae and recognizing the Helicellinae, Geomitrinae, Hygromiinae, and Trochulininae within the Hygromiidae.

Razkin et al. (2015) and Neiber et al. (2017) both adopted a multi-gene phylogenetic approach based on mitochondrial sequence data for the Hygromiidae s.l. Both studies used the mitochondrial 16S rRNA gene, and Neiber et al. (2017) also incorporated the mitochondrial CO1 section. Notably, previous research by Mardulyn and Whitfield (1999) found that 16S and CO1 phylogenies at the genus level exhibited a lack of signal and were prone to noise, although we note that Fontanilla et al. (2017) successfully used the mitochondrial CO1 gene to look at family-level relationships within the Achatinoidea by restricting analyses to the more conservative first and second codon positions and excluding the saturated third codon positions. Mengual et al. (2008), Zaldivar-Riverón et al. (2008), Klopfstein et al. (2010), and Townsend and Leuenberger (2011) have suggested that the nuclear 28S rRNA gene is a superior marker at ancient time scales when compared with non-conserved genes, such as CO1. Nevertheless, it is important to acknowledge that phylogenies based on the 28S marker still exhibit limited resolution in understanding the relationships within Hygromiidae s.l. In addition, our parsimony tree (Supporting Information, Fig. S4) treated each subfamily as a single branch within the multifurcating lineage of Helicoidea, with the exception of Hygromiinae, which grouped together with Cepolidae. This observation implies that 28S might not be sufficient to elucidate the relationships among these subfamilies.

Xanthonychidae

Although our phylogenetic tree suggests that the Xanthonychidae are monophyletic, it is important to consider the broader context. Previous phylogenetic studies on the Xanthonychidae based on morphological data, including the study by Cuezzo (1998), have raised doubts about their monophyly. These doubts stem from factors such as geographical distribution and morphological phylogeny. Calcutt et al. (2020) have provided support for the hypothesis that the Xanthonychidae are not monophyletic. Their molecular phylogenetic analysis suggested that Xanthonychidae can be divided into four distinct groups. This finding contrasts with the results of Frank Koehler’s phylogenomic research as reported by Bouchet et al. (2017). The biogeographical distribution also appears to align more with the findings of Calcutt et al. (2020). According to Miller and Naranjo-García (1991), the Monadeniinae are primarily distributed in the North Rocky Mountains, Helminthoglyptinae in the Middle Rocky Mountains, Humboldtianinae in the South Rocky Mountains, Xanthonychidae in Central America, and Epiphragmophorinae in the Andes Mountains of South America. It is essential to note that our taxon sampling of Xanthonychidae focuses primarily on Monadeniinae and might not be representative of the entire diversity within Xanthonychidae.

Pleurodontidae

In the study by Wade et al. (2007), there are three lineages of American Camaenidae. Two of these lineages are now classified as Sagdoidea, with the third lineage now classified as Pleurodontidae. In our study, the Pleurodontidae are sister to the Xanthonychidae. This is consistent with the phylogenetic tree of Sei et al (2017), which shows a sister-group relationship between Pleurodontidae and Xanthonychidae.

Helicodontidae

The Helicodontidae are the sister group of Hygromiidae s.l. + Xanthonychidae + Pleurodontidae in our phylogeny. The phylogenetic position of the Helicodontidae in our tree is consistent with the phylogeny of Gómez-Moliner et al. (2012), in which the Helicodontidae form a paraphyletic group that is slightly outside the Hygromiidae. In the study by Calcutt et al. (2020), the Helicodontidae are the sister group of one of the Xanthonychidae branches that includes Helminthoglypta umbilicata.

Helicidae

The Helicidae are the sister group of the Hygromiidae s.l. + Xanthonychidae + Pleurodontidae + Helicodontidae in our phylogeny. Our Helicidae results are consistent with the results of Wade et al. (2007) but conflict with Razkin et al. (2015). The Murellinae are the sister group of the Ariantinae in our analysis instead of being an early cladogenesis event within the Helicidae, as in the study by Razkin et al. (2015). Every subfamily is robustly supported in our results and in those of Wade et al. (2007) and Razkin et al. (2015).

Sagdoidea

In our phylogeny, Solaropsidae fall to the base of Sagdoidea, with Zachrysiidae falling as the sister to the Sagdidae. Our findings are therefore inconsistent with the phylogenetic topology of Sei et al. (2017) that was based on COII, 16S, and 28S, in which Solaropsidae are sister to the Sagdidae and Zachrysiidae separate as an early event. However, our present study is consistent with Calcutt et al. (2020) in the position of the Solaropsidae as sister to other Sagdoidea indicated by the phylogenetic tree based on CO1, 16S, and ITS2 + 28S.

Author Contribution

G.Z., C.M.W and F.N. designed the study. G.Z. collected, dissected and photographed the Chinese specimens. G.Z. and F.N. worked on morphological description and taxonomic actions. G.Z., P.S.A. and C.M.W. worked on the DNA sequences obtained. G.Z. and C.M.W. worked on bioinformatics analysis and generated the sequence alignment. G.Z. and C.M.W. prepared the primary manuscript, and G.Z., C.M.W., F.N. and P.S.A. revised the manuscript.

Data Availability

DNA sequences are available on Genbank (see Table 1 for accession numbers). Morphological descriptions of new taxa and supplementary files are available at Zoological Journal of the Linnean Society online.

[Version of Record, first published online 23 March 2024, with fixed content and layout in compliance with Art. 8.1.3.2 ICZN. http://zoobank.org/urn:lsid:zoobank.org:pub:C1953944-42BC-4858-81DA-3D800E2E0F54]

Acknowledgements

We are grateful to Yi Wang, Xuyun Qiu, and Junjie Qi for their help in the field. Thanks also go to the Willi Hennig Society for providing TNT. We also thank Dr Ben Rowson and an anonymous reviewer for their helpful comments. This research was supported by the University of Nottingham (G.Z. MRes project). Thanks also to the Tony Whitten Conservation Award 2020 for supporting fieldwork.

Conflict of interest

None declared.

References

© 2024 The Linnean Society of London.

Issue Section:

Original Article

Email alerts

Citing articles via

More from Oxford Academic