Molecular analyses revealed three morphologically similar species of non‐native apple snails and their patterns of distribution in freshwater wetlands of Hong Kong

Effective control of invasive species and conservation of native biodiversity requires accurate species identification. Several species of apple snails (Ampullariidae: Pomacea) from South America have become widespread pests in Asia since their introduction in the early 1980s, but their taxonomic uncertainty has hindered our understanding of the invasive processes. We aim to determine the identity and distribution of Pomacea species in Hong Kong, which has been known as a stepping stone of species invasion.


| INTRODUC TI ON
Biological invasion represents one of the key drivers of biodiversity change by impacting native species and altering the structure of species interaction (Díaz et al., 2019;Sentis et al., 2021). Effective control of invasive species and conservation of native biodiversity demands reliable species taxonomy (Ely et al., 2017;Shaik et al., 2016). Accurate species identification is essential for pest quarantine and early detection of invasive species. Poor taxonomy may lead to confusion of the invasive species and their native relatives, and thus delay the control against invasive species and even endanger the conservation of natives (Hayes, 2021). Moreover, correct species identification serves as a promise to the knowledge of the distribution patterns and population genetic patterns, which contribute to the management of biological invasion. Understanding the distribution patterns is crucial for range prediction and risk prevention of invasion (Zhang et al., 2020). Furthermore, knowledge of the patterns of genetic diversity and genetic structure can help elucidating evolutionary processes and invasive histories (Jackson et al., 2015).
Several apple snail species of the genus Pomacea (Caenogastropoda: Ampullariidae) have successfully invaded tropic and subtropic habitats globally and become the widespread agricultural and environmental pests (Cowie et al., 2017). High morphological symmetry and plasticity make Pomacea species difficult to identify (Hayes, 2021). Taxonomic studies integrating morphological characterization and molecular phylogenetics using mitochondrial genes over the last decade have clarified some of the taxonomic confusion in Pomacea spp. (Hayes et al., 2012;Rawlings et al., 2007;. The accurate taxonomy of apple snails has improved our knowledge on their distributions in their introduced regions around the world, such as in North America (Pierre et al., 2017;Rawlings et al., 2007), Asia (Hayes et al., 2012) and Pacific islands (Tran et al., 2008). Nevertheless, chance of introduction as well as differences in physiological tolerance to environmental factors such as temperature and hypoxia may determine their distribution and dominance patterns in different regions (Matsukura et al., 2016;Mu et al., 2015Mu et al., , 2018Seuffert & Martín, 2021). In many of their invaded regions, the identities of apple snails remain unknown which hinders our understanding of their ecological impacts, and the development of effective management measures (Kwong et al., 2009(Kwong et al., , 2010. At least five non-native Pomacea species (P. canaliculata, P. maculata, P. occulta, P. diffusa and P. scalaris) occur across South-East Asia, among which the first three have caused great economic and ecological impacts . In mainland China, P. canaliculata occurs in 12 southern provinces, but P. maculata only distributes in several locations in the southwestern regions, such as Sichuan Province and Chongqing Municipality, and one population in the eastern Zhejiang Province (Lv et al., 2013;. Pomacea occulta, mis-identified as P. maculata in a previous study (Song et al., 2010), previously considered as a cryptic clade (Lv et al., 2013;, is now recognized as a distinct species . Pomacea occulta exhibits a wide distribution in 11 provinces, often co-existing with P. canalicualta in China . Both P. diffusa and P. scalaris have been considered to possibly have a smaller ecological impact, but their non-native populations have occurred in some regions (Li, 2008;Rawlings et al., 2007). Hayes et al. (2008) reported nonnative populations of P. diffusa in Sri Lanka as a result of the aquarium trade. Although P. diffusa serves as a popular pet for aquarium trade, this species has not yet invaded the wetlands of China (Yang, Liu, Song et al., 2018). The fifth species, P. scalaris, however, have been reported only from southern Taiwan outside its native range Wu et al., 2011).
The Pomacea spp. invaded Hong Kong in the early 1980s (Kwong et al., 2008). Results from two territory-wide surveys showed that non-native apple snails, initially found in the paddy fields and drainage channels of the northern New Territories, had spread to southern New Territories and Tsing Yi Island (Kwong et al., 2008;Yipp et al., 1991). Yipp et al. (1991) reported the presence of Ampullaria levior and Ampullaria gigas in the northern New Territories. Both Ampullaria species are junior synonyms of P. canaliculata (Hayes et al., 2012). However, no specimens collected by Yipp et al. (1991) remain for confirmation of their species identities. In a study mainly aiming to determine the distribution of apple snails in Hong Kong, Kwong et al. (2008) selected 16 most morphologically distinct individuals based on shell size and head-foot colours from 34 freshwater habitats for determination of mitochondrial cytochrome c oxidase subunit I (COI) sequences, which led to the identification of these specimens as P. canaliculata. Nevertheless, due to the small sample size used and considerable intra-specific plasticity in shell morphology in Pomacea, it remains undetermined how many species of apple snails have invaded the freshwater wetlands of Hong Kong.
In the present study, we collected apple snails from five populations in Hong Kong to clarify their species identities and composition through sequencing the COI gene of at least 30 specimens from each population. Because several studies have extensively investigated the species identities of invasive Pomacea spp. in mainland China and from this study will contribute to better understanding their invasive mechanisms and prevention of their continued spread in Asia.

K E Y W O R D S
biological invasion, genetic introgression, mitochondrial COI, nuclear EF1α, phylogenetic analysis, Pomacea, population genetics compared with the global introduced and native populations Lv et al., 2013;Yang, Liu, He, et al., 2018), we compared the genetic diversity of apple snails in Hong Kong with those in mainland China. We included sequences of apple snails from Malaysia because it provides the only country with recent COI gene data on apple snails (Kannan et al., 2020;Rama Rao et al., 2018). Moreover, because introgressive hybridization between Pomacea species has been detected in both Asia and their native ranges (Glasheen et al., 2020;Matsukura et al., 2013;Yang et al., 2020;Yoshida et al., 2014), we also analysed the hybridization patterns of apple snails collected from Hong Kong through genotyping the nuclear elongation factor 1-alpha (EF1α) gene. These sites covered the diversity of habitats inhabited by apple snails in Hong Kong's freshwater environments: two vegetable gardens (HSH and LT), a reconstructed wetland as a compensation for the loss of wetland due to construction of a rail road (KT), a wetland park created as a compensation for the wetland loss due to new town development (WP), and a lotus pond (LMC). We observed a strong sex bias (female/ male = 8/1) in adults and only a few egg clutches in HSH. Sex ratios in other populations were close to 1:1. We deposited the tissue samples used in this study in Zhejiang Museum of Natural History (catalog numbers AS0000001M1-AS000000162M1) as voucher specimens.

| DNA extraction, PCR and sequencing of mitochondrial COI
We extracted total DNA from ~10 mg of foot muscle tissue of each snail using the TIANamp Genomic DNA Kit (TIANGEN) following the manufacturer's protocol, and finally eluted them in 200 μl ddH 2 O.
We amplified the 5′-end portions of the mitochondrial COI from each individual with the universal primer pair LCO1490/HCO2198 (Folmer et al., 1994; Appendix S1: Table S1) in 25 μL PCR reaction, which contained 0.625 U TaKaRa Ex Taq, 1 × Ex Taq Buffer, 5 mM dNTP mixture, 10 μM each primer and 1 μL total DNA template. We performed PCR by initial denaturation at 95°C for 3 min, followed by 35 cycles of 30 s at 95°C, 30 s at 50°C and 60 s at 72°C, then final extension at 72°C for 8 min and termination at 4°C. We visualized and checked for specificity of the PCR products via gel electrophoresis.
We sent single product amplicons to Sangon Biotech for sequencing using the forward primer by using an ABI3730 DNA Analyzer (Applied Biosystems), and checked all the fluorescent peaks of the sequences for errors using the program Geneious v11.1.5 (Kearse et al., 2012). If the sequence had low quality with multiple peaks, we sequenced the amplicon again with the reverse primer. We obtained the target region after removing the primers and error bases at both ends using Geneious.

| COI datasets and haplotypes
A preliminary BLAST search (Johnson et al., 2008) of the GenBank nucleotide database suggested that the sequences of apple snails from Hong Kong generated in the present study matched those of P. canaliculata, P. maculata and P. occulta. Therefore, we included the published COI sequences of the three species that covered the known geographical ranges of populations from mainland China and Malaysia, and also the representative sequences from Argentina and Brazil to form a combined dataset (Appendix S1: We performed global alignment of the dataset using Geneious v.11.1.5 (Kearse et al., 2012) and trimmed the sequences to a consensus sequence dataset to generate haplotypes after alignment.
Because the 521 sequences from Lv et al. (2013) were 503 bp in length, which was the shortest among the sequences from various studies, we trimmed the dataset with a total of 1378 sequences to 503 bp. To evaluate any trade-off of shorter COI sequence on the downstream haplotype analysis, we also developed a consensus dataset of 578 bp with 857 COI sequences excluding the sequence data from Lv et al. (2013). We compared the population genetic diversity results with the datasets of shorter and more sequences versus longer and less sequences. We used DnaSP v5.10 (Librado & Rozas, 2009) to identify haplotypes of Pomacea spp. from the datasets.

| Phylogenetic relationships and genetic divergence
We identified the apple snails from Hong Kong based on the phylogenetic relations and genetic divergence with the known species from mainland China, Malaysia, Argentina, and Brazil. We reconstructed phylogenetic trees using neighbour-joining (NJ), maximum likelihood (ML) and Bayesian inference (BI) approaches using MEGAX (Kumar et al., 2018), IQ-Tree v1.6.12 (Nguyen et al., 2015) and MrBayes v3.2.5 (Ronquist & Huelsenbeck, 2003), respectively. Pomacea diffusa (MF373586) was used as the outgroup. We employed the Kimura-2-parameter (K2P) model with uniform rates among sites for reconstructing the NJ trees. We used jModeltest2 (Darriba et al., 2012) to select the best substitution model for ML and BI analysis. We tested the phylogeny with 1000 bootstrap replicates for both the NJ and ML analyses. For the BI analysis, we conducted two independent runs with four (one cold and three heated) Markov chains for 5 million generations, estimated majority-rule consensus trees by combining results from the duplicate analyses, and sampled Markov chains every 1000 generations with discarding the first 25% as burn-in. When the average standard deviation of split frequencies was below 0.01, the stationarity was considered to have been reached, and the analysis was stopped (Huelsenbeck et al., 2001).

| Population genetic diversity and structure
To reveal the evolutionary relationships among haplotypes, we analysed the network of haplotypes generated from the dataset using TCS v1.21 (Clement et al., 2000). The statistical parsimony connection limit for the network reconstruction was set to 95% and 90% for  (Wright, 1943). A F ST value >0.15 indicates significant differentiation (Frankham et al., 2002), while an Nm value >1 indicates sufficient gene flow to maintain continuity among populations (Excoffier & Lischer, 2010).

| Genotyping the nuclear EF1α gene
For the apple snails from Hong Kong, we further determined the nuclear EF1α genotypes as types of P. canaliculata, P. maculata, or their mix. Individuals with identical genotype determined by the COI and EF1α genes were 'pure' P. canaliculata and P. maculata. Individuals with inconsistent identification based on the two genes were hybrids. We genotyped the nuclear EF1α gene for each snail collected from Hong Kong and diagnosed as P. canaliculata or P. maculata using the primer-specific-multiplex PCR described by Yang et al. (2020).
Although a previous study based on the EF1α gene indicated incomplete lineage sorting between P. occulta and P. canaliculata and P. maculata , we also genotyped the EF1α gene of the snails with a COI type of P. occulta to check the hybridization patterns among these species. The multiplex PCR assay included two pairs of specific primers (Appendix S1: Table S1), including EF3Fc/ EF3Rc targeting on a 125 bp of canaliculata-EF1α, and EF1Fm/ EF3Rm targeting on a 151 bp of maculata-EF1α (Yang et al., 2020).

| Sequences, haplotypes and similarity
Analysis of 503 bp COI sequence in 1378 apple snails from Asia (China, Hong Kong and Malaysia) and South America (Argentina and Brazil) revealed a total of 88 haplotypes (Table 1; Appendix S1: There were 27 haplotypes of P. canaliculata (Hap37-Hap54) and P. maculata (Hap55-Hap63) exclusively from the Argentinean populations (Appendix S1: Table S3). Hap64-Hap88 of P. maculata were generated only from the Brazilian populations (Appendix S1: Table   S3).
The 162 sequences from Hong Kong belonged to 10 haplotypes (Hap1-Hap10, Table 2). The two most common haplotypes were Hap1 (61.1% of total sequences) and Hap2 (19.8%), which were identified as P. canaliculata and presented in all the five populations.
Hap8, the third top abundance haplotype (6.8%), was identified as P. occulta and detected from four populations except the northern vegetable garden in Ho Sheung Heung ( Table 2). The other haplotypes were rare (<5%), distributed in only one or two populations and represented by one to eight individuals. In addition, BLAST analysis against the GenBank nucleotide database identified four unique haplotypes in Hong Kong (Table 1), including Hap5 (P. canaliculata, 1 bp different from the closest haplotype Hap1), Hap7 (P. maculata, 7 bp different from Hap6) and Hap9-Hap10 (P. occulta, 3 bp and 1 bp different from Hap8, respectively). Both Hap5 and Hap 9 were found in Lok Ma Chau; Hap7 and Hap10 were found in Ho Sheung Heung and Wetland Park, respectively (Table 2).
Apple snails from mainland China were more genetically diverse, with 21 haplotypes of P. canaliculata, 3 haplotypes of P. maculata and 6 haplotypes of P. occulta (Table 3). In contrast, the Malaysian Pomacea had only four haplotypes of P. canaliculata and one haplotype of P. maculata (Table 3). Hap1 was the most common haplotype of P. canaliculata from Hong Kong, but it was the second largest haplotype of P. canaliculata from mainland China. Hap2 that had been identified in 507 specimens was the most dominant haplotype in China (Table 3), but it was absent from the Malaysian population.

| Phylogenetic relationships and genetic divergence
The NJ, ML and BI trees of COI sequence all supported the monophyletic clades of P. canaliculata, P. maculata and P. occulta (Figure 2, Appendix S1: Figure S1). Although the posterior probability value for distinguishing P. canaliculata sequences into its own monophyletic clade was lower than 0.7 on the BI tree, both the ML and NJ trees showed high bootstrap values (100) at the major clades (Figure 2, Appendix S1: Figure S1). Both the P. canaliculata and P. maculata Hap8-Hap10 were nested in the P. occulta clade.
The phylogenetics of haplotypes further confirmed the presence of three species of Pomacea in Hong Kong: P. canaliculata (Hap1-Hap5), P. maculata (Hap6-Hap7) and P. occulta (Hap8-Hap10). All five locations contained P. canaliculata, accounting for 85.8% of the studied specimens. We detected P. occulta (8.0%) in all the populations except Ho Sheung Heung, while P. maculata (6.1%) was limited to Ho Sheung Heung and Kam Tin (Table 1; Figure 1). The three species co-existed in Kam Tin. In other locations P. canaliculata either co-existed with P. maculata (Ho Sheung Heung) or with P. occulta (Lok Ma Chau, Wetland Park and Wetland Park) (Table 1; Figure 1).
The intra-specific K2P distances of P. canaliculata, P. macualta and P. occulta each exhibited a dumbbell shape with high probabilities at both ends ( Figure 3). The genetic variations of P. canaliculata and P. macualta were high with a maximum intra-specific K2P distance of 0.069 and 0.060, respectively (Figure 3). The inter-specific K2P distances of P. maculata and P. occulta were of broad ranges from 0.070 to 0.133 and 0.062 to 0.125, respectively (Figure 3).

| Population genetic diversity and structure
The three descriptive indices of population genetic diversity (i.e. Hd, π and k) revealed polymorphisms in Pomacea spp. from Hong Kong, mainland China and Malaysia, except for P. maculata from Malaysia (Table 3). In general, P. canaliculata showed higher genetic diversity with two times higher mean Hd and more than ten times higher mean π and k than P. maculata and P. occulta. Among the three regions, the genetic diversity of P. canaliculata was the highest in mainland

| Genetic introgression
Overall, 49.6% of snails were 'pure' P. canaliculata and there was only one 'pure' P. maculata individual out of the ten maculata-COI specimens. 'Pure' P. canaliculata dominated the Kam Tin, Ho Sheung Heung and Wetland Park populations, while the hybrids of P. canaliculata-COI and P. maculata-EF1α dominated the Lok Ma Chau and Wetland Park populations (Figure 1). Among the P. maculata-COI specimens, the P. canaliculata-EF1α was the most common nuclear genotype, followed by the genotype of mixed-EF1α from Ho Sheung Heung; only one individual of pure P. maculata was detected from Kam Tin ( Figure 1).
Due to incomplete lineage sorting or recent hybridization, we were not able to distinguish P. occulta from P. canaliculata and P. maculata based on the nuclear EF1α sequences . The P. canaliculata-EF1α, P. maculata-EF1α and mixed-EF1α snails of EF1α genotyped were all detected from P. occulta in Hong Kong (Figure 1).

| DISCUSS ION
Based on analysis of the mitochondrial COI barcoding region, we identified three non-native apple snail species, P. canaliculata, P. maculata and P. occulta, from various freshwater wetlands in the New Territories of Hong Kong, but they could not be readily distinguished by shell morphology (see Appendix S1: Figure S2). To our knowledge, our study represents the first report of Pomacea spp.
other than P. canaliculata in natural habitats of Hong Kong, which has enhanced our knowledge on the distribution ranges of P. maculata and P. occulta in Asia (e.g. Hayes et al., 2008;Yang, Liu, He, et al., 2018). Pomacea maculata was restricted to several populations in mainland China (Yang, Liu, He, et al., 2018) and has not been reported from Shenzhen, Guangdong Province of China, which is physically connected with Hong Kong through a network of drainage channels in the same catchment (Yang, Liu, He, et al., 2018).
Moreover, there has been no report of P. occulta as an invasive species other than in the mainland and Hong Kong. Whether P. maculata and P. occulta are truly absent on the wetlands require confirmation based on more apple snail samples. Nevertheless, our current data indicate a risk for P. maculata and P. occulta to spread to the abovementioned areas in future. As Lu et al. (2018) suggested Hong Kong as one of the major stepping-stone for invasive species to the mainland of China, the delineation and identification of invasive species in Hong Kong are essential for early warning and control of species introduction to China and other Asian regions.
Pomacea canaliculata was the most widespread species, being recorded from all of the five sampled natural populations in Hong Kong. Pomacea maculata, however, was restricted to two locations (a vegetable garden in Ho Sheung Heung and an artificial wetland in Kam Tin) in our study, and in each population, this species accounted for the smallest proportion of Pomacea individuals at the sites. These results are in agreement with Hayes et al. (2008) which showed that P. canaliculata is often the dominant species in Asia, and P. maculata is distributed in fewer locations and often co-occurring with P. canaliculata. In contrast, the native range of P. canaliculata is mainly Northern Argentina and Southern Uruguay, while P. maculata occurs in a much larger range than P. canaliculata in South America, extending from most western Brazil to Paraná and Uruguay river basins of Argentina and Uruguay (Glasheen et al., 2020;Hayes et al., 2008Hayes et al., , 2012Martín et al., 2001). It is difficult to determine how the native habitat breadth may have affected the ecological impacts and thus driven the relatively success between the two closely related Pomacea species, except perhaps in more high latitude areas (Matsukura et al., 2013). In Hong Kong, the differences in the apple snail species composition among the sampling sites may have largely relied on the numbers of introductions that have been made, although their differences in tolerance to environmental stressors may have also contributed.
Our phylogenetic analysis showed that the invasive populations of P. canaliculata in Hong Kong were likely the results of repeated introductions from various locations in Argentina (Hayes et al., 2012).
The same conclusion can be applied to P. canaliculata populations in mainland China and Malaysia. We found a novel haplotype of P. canaliculata in Hong Kong. There were also unique P. canaliculata haplotypes to the Malaysian populations. The other haplotypes of P. canaliculata from Hong Kong and Malaysia were likely a subset of those that have been reported from mainland China (Lv et al., 2013;Yang, Liu, He, et al., 2018). However, the dominant haplotypes of P. canaliculata were different among Hong Kong, mainland China and Malaysia populations, which indicated that P. canaliculata was with a complex invasion history in the three regions.
The P. maculata populations of mainland China, Malaysia, Vietnam, Singapore and Cambodia likely came from Brazil Yang, Liu, He, et al., 2018); the Korean populations likely came from Argentina , while the populations in Japan and Thailand likely came from either Argentina or Brazil F I G U R E 5 Haplotype networks revealing the frequencies and relations of three Pomacea species. The 90% parsimony limit has split the network into different subnetworks in Pomacea canaliculata (A-C), P. occulta (D) and P. maculata (E-G). The area of the circles is proportional to the observed frequency of the haplotypes. White circles indicate missing haplotypes [Colour figure can be viewed at wileyonlinelibrary. com] Matsukura et al., 2008 (Byers et al., 2013;Rawlings et al., 2007;Underwood et al., 2019).
In this study, we discovered P. occulta for the first time in Hong Kong and revealed its invasion history similar to that in mainland China. Pomacea occulta co-exists with P. canaliculata in four of the five populations with a lower abundance. The widely overlapped distribution of P. occulta with P. canaliculata indicated their cointroduction from Argentina, as was indicated by  in an analysis of apple snail distribution in mainland China. Pomacea occulta was absent from the vegetable garden in the most northern sampling locality (Ho Sheung Heung), where the sex ratio of the apple snail population was strongly bias. Apple snail populations have shown extremely variable brood sex ratios that were determined genetically but equal sex ratios in populations (Yusa, 2006;Yusa & Suzuki, 2003). Moreover, we did not observe obvious correlation between the species distribution pattern and the habitat type. The bias sex ratio indicates that the apple snail population in Ho Sheung Heung was disrupted, which could be caused by various factors, such as environmental contaminants and conditions (Ciparis et al., 2012). Alternatively, it could be due to sampling bias.
Compared to the native apple snail populations in Brazil and Argentina (total of 53 haplotypes), those in the introduced ranges in Hong Kong, mainland China and Malaysia have lower haplotype diversity. This result can be explained by the founder effect, which is the loss of genetic variation due to the small invasive population.
Nucleotide diversity (π), which is relatively independent of sample size and sequence length, is a more stable index of genetic divergence (Li, 1997). For P. occulta, the genetic diversity of the Hong Kong populations (π = 0.00122) was close to that of mainland China Hybridization can enhance the chances of local adaptation and range expansion into new habitats (Mesgaran et al., 2016;Pfenning et al., 2016). Glasheen et al. (2020) first reported the pattern of hybridization of apple snails in their native ranges. Our results of multiple EF1α genotypes in P. occulta supports the idea that some of the founder individuals might have been hybrids. Hybrids of P. canaliculata and P. maculata in their native areas in Uruguay and Brazil were high, reaching up to 30% (Glasheen et al., 2020). The degree of hybridization of the two species in Hong Kong was even higher, reaching 53%. Similarly, Yang et al. (2020) found more hybrids than pure apple snails in southern coastal areas of mainland China. The higher percentage of hybrids indicated the hybridization has intensified in the introduced ranges through their co-occurrence in habitats, which could have probably promoted the invasion through benefiting from the more intensive genetic exchanges.
The apple snail populations in Hong Kong contained hybrids with different genetic introgression patterns. To our knowledge, the 'pure' P. maculata individual in our study is the first report of 'pure' P. maculata in Asia. No 'pure' P. maculata has been reported from other Asian regions including mainland China, Japan and Korea (Matsukura et al., 2016;Yang et al., 2020), but it is not known whether this conclusion will change when more sequences of the EF1α gene from these regions become available. The genetic introgression among populations may be affected by natural selection imposed through environmental stressors. For example, Matsukura et al. (2013), Matsukura et al. (2016 reported differential cold tolerance abilities among the 'pure' species and hybrids, and suggested that such physiological differences might have determined the distribution of these apple snails in Japan. Hybridization between species can contribute to population fitness by increasing genetic variation to compensate the reduced genetic variation resulted from the founder effect (Pfenning et al., 2016). The enhancement of local adaptation might result the range expansion in new habitats, and thus, the high hybridization rate in Hong Kong will pose invasion risk to China and nearby regions. However, a comprehensive understanding of genetic benefits through hybridization for range expansion in apple snails requires further population genomic research.
In summary, we found P. maculata and P. occulta in Hong Kong for the first time. Our results showed that there were likely multiple introductions of apple snails in Hong Kong, with P. canaliculata and P. occulta having been introduced from Argentina, and P. maculata having been introduced separately from Argentina and Brazil. The low genetic diversity indicated the apple snail populations in Hong Kong indicated the presence of founder effect. Both the repeated introduction and high hybridization might compensate and can enhance genetic variation and thus pose further invasion risk to nearby regions. In other invaded regions where only P. canaliculata has been reported to cause great economic impact (i.e. Taiwan of China, Philippines, Myanmar, Laos; Cowie et al., 2017), molecular studies of Pomacea should be conducted to determine their true species identities, species composition and patterns of hybridization. Our results emphasize the importance of strengthening quarantine measures to prevent the further spread of these apple snails, not only to prevent their further spread, such as P. maculata to the southern provinces of China (i.e. Guangdong, Guangxi and Fujian) and P. occulta to other Asian countries, but also to prevent the gene flow among the different populations of the same species.

ACK N OWLED G EM ENTS
This work was supported by grants from National Natural Science

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.

PEER R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/ddi.13443.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data obtained in this study have been deposited to GenBank under the accession numbers MT806196-MT806357.