Molecular investigation on diversity of the land snail genus Aegista (Gastropoda, Camaenidae) in South Korea

Abstract Aegista Albers, 1850 is a large genus of the land snail family Camaenidae Pilsbry, 1895 and distributed in south, southeast and east Asian countries (from India and Nepal to Korea and Japan). Fourteen species and subspecies of Aegista are known from South Korea. They were described, based only on shell morphology during 1887–1943 and our knowledge on diversity of Korean Aegista has seldom been updated since then. In this study, we provide the report on the first molecular investigation of diversity of Aegista in South Korea, which unmasked some of overlooked diversity of this group.


Introduction
The Korean Peninsula is located at the edge of the eastern part of the Asian continent (Fig. 1).It has had intermittent land-bridge connections with mainland China and its nearby large archipelago (i.e. the Japanese archipelago) through multiple glacial cycles (Kim andKennett 1998, Kong et al. 2006).Moreover, the Peninsula has never been covered by ice sheets during the Quaternary, which makes it possible for terrestrial organisms to inhabit the Peninsula for a long period (Chung et al. 2017).In East Asia, for example, the Korean Peninsula is the only region where an ancient divided member of the salamander family Plethodontidae Gray, 1850 (Amphibia), whose centre of diversity is North and Middle America, is found (Min et al. 2005).The ancestral group of the Asian plethodontid salamander is estimated to have migrated from the North American continent around 45 Mya and Asian plethodontid populations have probably been extirpated, except in the climatologically stable region, the Korean Peninsula (Shen et al. 2015).These complex geographical characteristics make this region attractive for biodiversity and phylogeographic studies.Map showing the sampling localities.The number corresponds to the locality No. in Table 1.Our analyses also included the sequences of the specimens from Japan, Taiwan and Laos.
Aegista Albers, 1850 is a large genus of the land snail family Camaenidae Pilsbry, 1895 and distributed in south, southeast and east Asian countries (from India and Nepal to Korea and Japan) (Hirano et al. 2015, Nurinsiyah et al. 2019).Although some researchers treat Landouria Godwin-Austen, 1918 as a distinct genus (Köhler et al. 2018, Nahok et al. 2021), as suggested by Hirano et al. (2014), we treated members of the genus as Aegista in this study because the monophyly of Landouria is still unclear.Fourteen species and subspecies of Aegista are known from South Korea (Kuroda 1958).They were described during 1887-1943 and, unfortunately, our knowledge on diversity of Korean Aegista has seldom been updated since then.The classification of these Korean Aegista is based only on shell morphology.However, recent molecular phylogenetic studies of land snails have revealed cryptic species within a single morphospecies and demonstrated the unreliability of shell morphology for species identification (e.g.Chiba and Davison (2008), Nantarat et al. (2014)).Hirano et al. (2014) examined a lot of species of Aegista (n >70) from almost all regions of Japan in addition to its surrounding regions such as Taiwan and China and found that multiple phylogenetic clades inhabit the Japanese archipelago, which has a profound biogeographic connection with the Korean Peninsula.Whether the Korean Aegista species include cryptic species and distribution pattern of the phylogenetic lineages in the Korean Peninsula remain to be studied.In the present study, we aim to address this question.

Samples
Individuals of 12 Aegista species and subspecies (Fig. 2) were collected in South Korea from 2018 to 2021 (Table 1).In total, 42 individuals were used in this study and six of them were not identified to the species-level because they were still small juveniles.A small portion of the foot muscle from each snail individual was stored in 99.5% ethanol for DNA extraction.For some species, DNA sampling with body swabbing (Morinha et al. 2014) was used in order to avoid unwanted effects of sample collecting on snail populations.The examined Aegista specimens are preserved in Tohoku University Museum (TUMC) or the authors' collection (Voucher No. in

Phylogenetic analyses
Total DNA was isolated from each foot piece and swabbed subsance using the DNeasy Blood & Tissue Kit or QIAamp DNA Micro Kit (Qiagen), according to the manufacturer's instructions.Fragments of mitochondrial (cytochrome oxidase subunit I [COI] gene) and nuclear (internal transcribed spacer [ITS] 1-2 regions and 5.8S rRNA gene) DNA markers were amplified and sequenced.The PCR conditions and primer sets were decided following Hirano et al. (2014).The PCR products were purified using Exo-SAP-IT (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK).Sequencing was performed using a BigDye™ Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) and electrophoresed using an ABI 3130xl sequencer (Applied Biosystems, Carlsbad, CA, USA).The resulting COI and ITS sequences have been deposited in the DDBJ/EMBL/GenBank database (Table 1).In addition to these new sequence data for Korean Aegista, already existing datasets of the genus and its closely related genus Euhadra Pilsbry, 1890 were obtained from GenBank in order to conduct the phylogenetic analyses (Table 1).
The mitochondrial and nuclear sequences were aligned with MUSCLE v.3.8 (Edgar 2004).
In order to eliminate any uncertainty in the ITS alignments, trimAl (v.1.2) with automated option (Capella-Gutiérrez et al. 2009) was used to exclude ambiguous alignment regions.The phylogenetic trees were obtained for a concatenated dataset (533 sites for COI and 1,147 sites for ITS) using Maximum Likelihood (ML) and Bayesian Inference (BI) methods.
Prior to the ML and BI analyses, ModelFinder (Kalyaanamoorthy et al. 2017) was used to select the optimal partition scheme for codon position and the appropriate models for sequence evolution.As a result, the following models were selected: TIM2+F+G4 model for codon 1 of COI, TIM+F+R4 for codon 2 of COI, GTR+F+ASC+G4 for codon 3 of COI, TVM+F+R3 for ITS in the ML analysis; GTR+G for codons 1, 2 and 3 of COI and ITS in the BI analysis.ML analysis was performed with IQ-TREE (v.1.6.8)using the options of edgeunlinked branch lengths between partitions and a perturbation strength of 1.0 (Nguyen et al. 2014, Chernomor et al. 2016).For the ML analyses, we assessed nodal support by performing ultrafast bootstrap analyses with 10,000 replications in IQ-TREE (Hoang et al. 2018).BI analysis was performed using MrBayes (v.3.2.7)with two simultaneous runs (Ronquist and Huelsenbeck 2003).Each run consisted of four simultaneous chains for two million generations and sampling of trees every 100 generations.We discarded the first 2,001 trees as burn-in after examining the convergence of runs and their effective sample sizes (ESSs) using Tracer (v.1.6) and used the remaining samples to estimate tree topology, branch length and substitution parameters.

Species delimitation analyses
To validate the Korean species of Aegista, three species delimitation analyses were conducted.For topology-based approach, ML Poisson Tree Process model (mlPTP; Zhang et al. (2013)) and the Bayesian Poisson Tree Process model (bPTP; Zhang et al. (2013)) were used and the topology of the obtained ML tree was applied for both models.For the bPTP model, we set default parameters (number of MCMC generation: 100,000; burn-in: 10%).Both analyses were run using web servers (http://species.h-its.org/ptp/).For distance-based approach, Assemble Species by Automatic Partitioning (ASAP) analysis (Puillandre et al. 2020) was performed using the COI dataset.This analysis was run using Kimura (K80) distances with the default setting at the ASAP website (https:// bioinfo.mnhn.fr/abi/public/asap/asapweb.html).The partitions with the first and second-best asap-scores were considered according to Puillandre et al. (2020).

Results
The results of the phylogenetic analyses for the concatenated sequences using ML and BI methods were mostly consistent.Only well-supported clades (the posterior probabilities ≥ 0.95 or ultrafast bootstrap support values ≥ 90%) are considered hereafter.
The 12 Aegista species in South Korea were separated into two major clades (Fig. 3).Subclade A1 was composed of at least nine species of Korean Aegista and subclade A2 was composed of a Korean and two Japanese species.Subclade B2 included the remaining Korean species and showed a sister relationship with subclade B1 composed of two Japanese species.The monophyly of A. chosenica was supported only in the BI tree.
For the Korean species in subclade A1, A2, and B2 the mlPTP, bPTP and ASAP analyses provided the same pattern of species delimitation (Fig. 3) and delineated 17 species (asapscore for the ASAP: 5.00), while the ASAP also partitioned 26 species with the second-best asap-score (6.50).For all analyses, the number of the species delineated was higher than the number of the Korean Aegista species recognised.None of the six small individuals that we could not identify was treated as a species with the identified snails in the species delimitation analyses.

Discussion
Our phylogenetic analyses, using the COI and ITS markers, revealed that multiple phylogenetic lineages of Aegista inhabit the Korean Peninsula and they are included in clade A or B. Hirano et al. (2014) showed that the majority of the Aegista species in the mainland of Japan are included in the two clades.Our result provides an example showing a close relationship between terrestrial biotas of Japan and South Korea.In both clades A and B, the lineages of Korean Aegista are not ancestral, which may suggest dispersal events of their progenitors from the Japanese archipelago to the Korean Peninsula.To test this hypothesis, further studies are needed to examine when Aegista has diversified and compare it with the geohistorical background of the Asian continent and the Japanese archipelago.Discovery of fossils of Aegista or big data arising from NGS technologies will help to address the history of diversification of this genus.Of the 14 Aegista species and subspecies known from South Korea, phylogenetic positions of A. pumilio (Pilsbry & Hirase, 1909) and A. gottschei fusanica (Pilsbry, 1926) are unclear.Although the unidentified small individuals may include these two taxa, further examination using specimens of them is needed to clarify their phylogenetic relationship with the other Korean Aegista.The species delimitation analyses suggested the presence of cryptic species in A. chosenica (Fig. 3).In addition, each of A. gottschei kyobuntonis and A. pyramidata hebes would need to be treated as a distinct species.Kuroda (1958) proposed the scientific name A. pyramidata hebesioides Kuroda, 1958 for the latter because A. hebes (Pilsbry & Hirase, 1905) is known from Taiwan.Although Kuroda's proposal based on the comparison between species and subspecies and thus was regarded as unnecessary, it would need to be reconsidered.The analyses also suggested that several unrecognised species of Aegista inhabit South Korea.It seems that at least three species have been overlooked even if the unidentified individuals include A. pumilio and A. gottschei fusanica.
Previous studies exhibited that Aegista shows a striking divergence in shell morphology and provides an example of parallel evolution of morphological traits (Hirano et al. 2014, Hirano et al. 2015).Our results also found examples of the independent evolution of shell flatness and hair-like ornamentation of the shell.For example, A. tenuissima omorii has a flat shell, while A. chejuensis has a globular shell.

Conclusions
In this study, we provide the report on first molecular investigation of diversity of Aegista in South Korea, which unmasked some of overlooked diversity of this group.To fully understand diversity of Korean Aegista, further studies using additional specimens including A. pumilio and A. gottschei fusanica, and an additional sampling effort, especially in the central region of the Korean Peninsula, are required.

Figure 3 .
Figure 3. ML tree of the Aegista species, based on the concatenated data (1680 bp) of the COI and ITS DNA markers.Each OTU label represents a species name.Numbers at nodes represent Bayesian posterior probabilities (right) and Maximum Likelihood ultrafast bootstrap support values (left).Only Bayesian posterior probabilities ≥ 0.95 and bootstrap support values ≥ 90% are shown.Scale bar indicates 0.07 substitutions per site.Arrow heads indicate clades and subclades.Blue shadows mark the Korean Aegista species.Black bars represent putative species suggested by the species delimitation analyses.

Table 1 .
List of the species included in the molecular phylogenetic analyses.Asterisks indicate the sequences newly obtained in this study.