A new species of the genus Pinctada is described from samples collected from the east coast of Phuket Island, Thailand in the Andaman Sea. Pinctada phuketensis sp. nov. is distinguished from other species on both molecular and morphological data. Morphologically, the valves of P. phuketensis are characterized by a slightly developed to undeveloped posterior auricle, a small, narrow slit-like byssal notch, the absence of hinge teeth, and a pale to transparent non-nacreous border, with a few dark brown or red blotches. This new species resembles P. fucata but differs by its smaller size and the absence of hinge teeth. Phylogenetic analyses based on both mitochondrial (COI) and nuclear (18S rDNA, ITS1 and ITS2) genes and species delimitation using COI data strongly support that P. phuketensis is a distinct species, which is closely related to Pinctada albina and Pinctada nigra.

Indian Ocean, mollusk, new species, phylogeny, taxonomy

Pearl oysters in the genus Pinctada (Röding, 1798) (family Pteriidae) are widely distributed from shallow to deep waters of the tropical and subtropical regions between the Indo-Pacific and western Atlantic (Wada and Tëmkin 2008; Cunha et al. 2011). Pinctada contains approximately 20 species according to the latest taxonomic records of MolluscaBase (2022). Several Pinctada species are used widely in pearl aquaculture and in industry including the Akoya pearl oyster Pinctada fucata (Gould, 1850) in Japan (Matsuyama et al. 2021); the black-lip pearl oyster Pinctada margaritifera (Linnaeus, 1758) in the South Pacific and Indo-Pacific Islands (Aideed et al. 2014; Ky et al. 2019); and the silver-lipped pearl oyster, Pinctada maxima (Jameson, 1901) in western Australia (Steve et al. 2021).

While the pearl farming industry has expanded rapidly during recent decades, our understanding of biodiversity, evolution, and conservation of Pinctada species is still limited. Traditionally, systematics and taxonomy of Pinctada species have primarily focused on morphological parameters (Hynd 1955, 1960; Wada and Tëmkin 2008). The identification of Pinctada species is largely based on the soft tissues and shell characteristics; however, such morphological features vary greatly and are sometimes indistinguishable between species, particularly if the specimens are young (Ranson 1961; Wada and Tëmkin 2008). Accordingly, these studies are relatively complicated and challenging due either to their non-discrete differentiation or high levels of morphological variation (Cunha et al. 2011; Scuderi et al. 2019). In order to address these problems related to morphology-based taxonomy, molecular approaches, together with detailed comparative morphology have been increasingly applied to elucidate the classification, distribution pattern and evolutionary history of Pinctada species (Yu and Chu 2006; Tëmkin 2010; Cunha et al. 2011; Lal et al. 2018; Reisser et al. 2019). Additionally, a recent proposal to raise the infraspecific taxon P. margaritifera persica to specific rank as P. persica (Jameson) has been supported primarily by partial mitochondrial cytochrome oxidase subunit 1 (COI) sequences and two different species delimitation methods (general mixed Yule-coalescent: GMYC and Automatic Barcode Gap Discovery: ABGD) (Sharif Ranjbar et al. 2016). The aforementioned study clearly confirms the potential of DNA sequences to unveil hidden diversity, geographic origin, and phenotypic plasticity of pearl oyster Pinctada species.

In the Southeast Asian region, nine species of Pinctada are currently recognized: P. albina (Lamarck, 1819), P. chemnitzii (Phillipi, 1849), P. fucata, P. imbricata Röding, 1798, P. maculata (Gould, 1850), P. margaritifera, P. maxima, P. nigra (Gould, 1850) and P. radiata (Leach, 1814) (Cheah 2007; Sanpanich and Duangdee 2018; MolluscaBase 2022). All nine Pinctada species have been recorded in Thailand (Wells et al. 2021). However, diversity and taxonomic studies of Thai Pinctada species have relied heavily on morphological features, and the research is outdated when compared with studies from other areas, such as the Central Pacific Ocean (Yu et al. 2006; Miyake et al. 2016; Saruwatari et al. 2018) and Indo-West Pacific (Colgan and Ponder 2002; Cunha et al. 2011; Reisser et al. 2019). Considering this fact, we postulate that the diversity of Pinctada species has yet to be fully revealed in Thai waters and adjacent areas.

Among Thai species, P. fucata and P. maxima are the main species used for pearl culture in Phuket, the island province off the western coast of Thailand (Bussarawit 1995; Kanjanachatree et al. 2004). As a consequence of great abundance and high demand for these two Pinctada species, most previous studies focused on their life cycle, physiology and cultivation techniques (Kanjanachatree et al. 2004, 2006; Piyathamrongrut et al. 2009), whereas little is known about their biodiversity and genetic resources. We recently collected several Pinctada specimens from Phuket, and some of them were quite different in external appearance from other reported Pinctada species in this area. Accordingly, the present study aims to clarify the taxonomic status of these recently collected Pinctada specimens based on morphological and molecular analyses.

A total of 15 pearl oyster specimens were collected around Dok Mai Island (7°47.84'N, 98°31.84'E), Phuket Province, western coast of Thailand by SCUBA diving. All specimens were allocated a registration code (NMR) to facilitate sample tracking. A small piece of adductor muscle from each oyster was preserved in 90% ethanol for DNA analyses. For morphological observation, we carefully examined both shell and soft body features (Wada and Tëmkin 2008), especially shell shape, hinge teeth pattern, posterior auricle and byssal notch. All characteristics were observed under the stereomicroscope. Voucher specimens were deposited at Kasetsart University Museum of Fisheries (Natural History Museum) mollusk collection (KUMF.MOLL.), Faculty of Fisheries, Kasetsart University, Thailand.

Genomic DNA extraction from mantle tissue was performed using NucleoSpin Tissue Kit (Macherey-Nagel, Germany). Mitochondrial cytochrome oxidase subunit 1 (COI) gene, nuclear 18S rDNA gene and nuclear ribosomal DNA internal transcribed spacer 1 and 2 (ITS1 and ITS2) regions were selected for molecular phylogenetic analysis according to previous studies (e.g., Yu and Chu 2006; Tëmkin 2010; Cunha et al. 2011; Sharif Ranjbar et al. 2016). Primer details, PCR amplification profile and procedure followed Folmer et al. (1994) for COI, Tëmkin (2010) for 18S rDNA and Yu and Chu (2006) for ITS1 and ITS2. PCR was carried out using PCR Master Mix solution (i-Taq^TM) (iNtRON Biotechnology DR, South Korea) in a total volume of 20 μl, consisting of 10 μl of i-Taq, 10 pmol of each primer and 2 μl of DNA (~ 10–20 ng). PCR products were purified with ExoSAP-IT (USB, Cleveland, Ohio USA) and then sequenced commercially (U2Bio Inc., Seoul, South Korea).

Newly generated sequences, including seven COI sequences, six 18S rDNA sequences and five ITS1 and ITS2 sequences, were deposited in NCBI. All sequences were edited, assembled, and aligned for individual and concatenated data sets using the Geneious Prime software package (Biomatters, available from http://www.geneious.com/) with the MAFFT sequence alignment algorithm, and were further manually refined. Additional sequences of Pinctada species were retrieved from NCBI and included in the alignment (Suppl. material 1). Pteria (Scopoli, 1777) species were selected as outgroups.

Phylogenetic trees were reconstructed for both individual (COI and 18S rDNA) and concatenated data sets (ITS1 + ITS2) using maximum likelihood (ML) implemented in IQ-TREE (Minh et al. 2020) and Bayesian inference (BI) implemented in MrBayes v.3.2 (Ronquist et al. 2012). ML analyses were carried out with the best-fit model selection implemented in ModelFinder (Kalyaanamoorthy et al. 2017). The nodal support values were estimated using the nonparametric bootstraps with 1000 replicates. For BI analyses, the nucleotide models of substitution were selected using Kakusan 4 (Tanabe 2011). BI analyses were conducted by two parallel runs of four Markov chains for a million generations with sampling every 1000 generations. The first 2500 trees (burn-in) were removed before determining a consensus tree and posterior probabilities. The best partition schemes (partitioned by codon position for COI dataset and by gene for ITS datasets) and substitution models of ML and BI methods for all datasets are listed in Suppl. material 2. Both ML and BI trees were edited and visualized with the program FigTree v.1.4.4 (Rambaut 2019).

Additionally, due to low variation of nuclear DNA sequences among species, we utilized only COI sequences for the three different species delimitation methods: the general mixed Yule-coalescent (GMYC) model (Pons et al. 2006), the Bayesian Poisson tree processes method (bPTP) (Zhang et al. 2013) and Assemble Species by Automatic Partitioning (ASAP) (Puillandre et al. 2021). The single-threshold GMYC analyses were performed via GMYC web server (http://species.h-its.org/gmyc/) using an ultrametric input tree. Ultrametric tree was constructed using BEAST v.2.5 (Bouckaert et al. 2019) with the uncorrelated lognormal relaxed clock, the GTR + I + R model and a coalescent tree prior. For the bPTP analyses, BI tree was used as input and implemented by web server (http://species.h-its.org/ptp/) with the setting of 100,000 MCMC generations and thinning value of 100. Additionally, the ASAP approach was applied using a web server (https://bioinfo.mnhn.fr/abi/public/asap/asapweb.html). The nucleotide substitution model K2P was selected, and other parameters were set to their default values.

Our analyses using morphological and genetic data clearly distinguished the Pinctada samples recently collected from Dok Mai Island, Phuket Province, western coast of Thailand, from other Pinctada species in the region. Accordingly, these pearl oysters should be recognized as a new species, and we have named it as Pinctada phuketensis sp. nov. This new species can be distinguished from other members of the genus by its smaller size, a subquadrate shell with moderately long ligament, slightly developed to undeveloped posterior auricle, the absence of hinge teeth, a pale to transparent non-nacreous margin with dark brown or black blotches, and brownish stripes on the external surface. A morphological comparison of P. phuketensis and some closely related species is presented in Table 1.

Among the Pinctada species distributed in Southeast Asian waters, the new species of P. phuketensis morphologically resembles P. fucata, P. nigra and P. albina, but can be distinguished from these three species based on shell shape, hinge teeth and anterior/posterior auricles. Both P. fucata and P. nigra can be easily distinguished from P. phuketensis by having conspicuous hinge teeth. In addition, P. fucata can be separated from P. phuketensis by being larger in size and having a large and developed anterior auricle (Takemura and Okutani 1958; Colgan and Ponder 2002; Wada and Tëmkin 2008), while P. nigra clearly differs from our new species by exhibiting a large and developed posterior auricle and deep posterior sinus (Deng et al. 2019) (Table 1). On the other hand, P albina and P. phuketensis are similar in having no hinge teeth, but they can be differentiated by the characteristics of byssal notch and anterior border. Pinctada albina is distinguished from our new species by having a broad byssal notch and anterior border that projects well beyond the reference line (a line drawn at right angles to the hinge line through the byssal notch) (Hynd 1955) (Table 1).

Among the Indo-Pacific Ocean species, our new species, P. phuketensis closely resembles P. sugillata (Reeve, 1857) from Australia in having a weakly developed to undeveloped posterior ear and a nearly 1:1 ratio of the hinge line to the antero-posterior axis of the shell (Hynd 1955, 1960). However, these two species differ in shell shape (slightly oblique in P. phuketensis and very oblique in P. sugillata), hinge teeth (absent in P. phuketensis and present in P. sugillata) and byssal notch (narrow and slit-like in P. phuketensis and moderately wide and slit-like in P. sugillata) (Table 1; Hynd 1955, 1960). Unfortunately, genetic data for P. sugillata are not available. Further studies on P. sugillata will be useful for confirming that these two species are distinct.

Our phylogenetic analyses and species delimitation approach showed that P. phuketensis is genetically distinct from other described Pinctada species. While our observations indicated that our new species is morphologically similar to P. fucata, genetic analyses revealed the distant phylogenetic relationship between these two species, implying that morphological traits probably do not reflect their real evolutionary history. Additionally, our phylogenetic analyses showed that P. phuketensis is more closely related to P. albina and P. nigra than to P. fucata. We also found that phylogenetic relationships of some Pinctada species in this study had weak nodal support and were incompletely resolved. It is apparent that further work on Pinctada species based on combined data of different genetic markers and more expansive sampling from different geographic regions will uncover their diversity, phylogenetic relationships and evolutionary patterns.

This work is submitted as part of the Master’s thesis by S. Somrup to the Dual-degree Program of Kasetsart University and Ocean University of China, sponsored by the China Scholarship Council through Ocean University of China. This work is also partially supported financially by Faculty of Fisheries, Kasetsart University (N. Muangmai), Phuket Pearl Co., Ltd. and National Research Council of Thailand (NRCT) and Biodiversity-Based Economy Development Office (Public Organization) (S. Winitchai). Our special thanks also go to Graham Oliver and two anonymous reviewers, who made a number of valuable suggestions for the improvement of our manuscript, and to Dave Anderson for critically reading and correcting the manuscript.