The giant deep-sea oyster Neopycnodonte zibrowii Gofas, C. Salas & Taviani, 2009 is a keystone deep-sea habitat builder species. Discovered about fifteen years ago in the Azores, it has been described and assigned to the genus Neopycnodonte Fischer von Waldheim, 1835 based on morphological features. In this study, we generated DNA sequence data for both mitochondrial (COI and 16S) and nuclear (ITS2 and 28S) markers based on the holotype specimen of N. zibrowii to establish a molecular phylogenetic framework for the systematic assessment of this species and to provide a reliable (i.e., holotype-based) reference sequence set for multilocus DNA barcoding approaches. Molecular data provide compelling evidence that the giant deep-sea oyster is a distinct species, rather than a deep-water ecophenotype of Neopycnodonte cochlear (Poli, 1795), with extremely high genetic divergence from any other gryphaeid. Multilocus phylogenetic analyses place the giant deep-sea oyster within the clade “Neopycnodonte/Pycnodonte” with closer affinity to N. cochlear rather than to P. taniguchii Hayami & Kase, 1992, thus supporting its assignment to the genus Neopycnodonte. Relationships within this clade are not well supported because mitochondrial variation is inflated by saturation that eroded phylogenetic signal, implying an old split between taxa within this clade. Finally, the set of reference barcode sequences of N. zibrowii generated in this study will be useful for a wide plethora of barcoding applications in deep-sea biodiversity surveys. Molecular validation of recent records of deep-sea oysters from the Atlantic Ocean and the Mediterranean Sea will be crucial to clarify the distribution of N. zibrowii and assess the phenotypic variation and ecology of this enigmatic species.

Azores, DNA sequences, Gryphaeidae, holotype, molecular systematics, Mollusca, multilocus phylogeny, Natural History Museum

Deep-sea is the Earth’s largest biome but it is still one of the most underexplored regions (Ramirez-Llodra et al. 2010). Deep-sea biodiversity is mostly unknown due to the extreme environmental conditions that limits sampling capabilities (Rogers et al. 2015; Sinniger et al. 2016; Woodall et al. 2018). Along with advances in exploration technologies (Feng et al. 2022), new molecular technologies such as high-throughput sequencing and the molecular identification of multiple species in environmental DNA (eDNA metabarcoding; Taberlet et al. 2012) have boosted deep-sea biodiversity assessments (Guardiola et al. 2016; Everett and Park 2018). However, the low number of reference sequences taxonomically validated in online repository databases (e.g. GenBank) limits the identifications of MOTUs (Molecular Operational Taxonomic Unit), thus reducing the taxonomic resolution of eDNA studies (Ruppert et al. 2019). Studies using an integrative taxonomic approach − combining molecular, morphological and environmental data − on new deep-sea taxa have been carried out in several groups of organisms, such as Anthozoa (López-González et al. 2022), Mollusca (Xu et al. 2019), and some others (Silva et al. 2016; Błażewicz et al. 2019). However, while molecular data are still not available for a great portion of known deep-sea biodiversity, deep-sea exploration has continued, contributing to the discovery of new benthic ecosystems and associated communities. Therefore, there is a great need to constantly improve with reliable reference sequences the taxonomic coverage of deep-sea taxa in repository databases.

In this study, we focused on a keystone deep-sea habitat builder species discovered about fifteen years ago in the Azores Archipelago: the giant deep-sea oyster Neopycnodonte zibrowii Gofas, C. Salas & Taviani, 2009 (Gryphaeidae Vialov, 1936) (Wisshak et al. 2009b). This reef-forming oyster was first observed during a submersible dive along the Faial Channel (480–500 meter depth) (Wisshak et al. 2009b). Deep-sea reefs of N. zibrowii are built by both stacked living and dead specimens on vertical rocky substrate of seamounts, escarpments and in canyons (Beuck et al. 2016), and host peculiar deep-sea communities. Benthic associations between N. zibrowii and the cyrtocrinid Cyathidium foresti Cherbonnier & Guille, 1972 have been documented in the Atlantic Ocean (Wisshak et al. 2009a), and between N. zibrowii and cold-water corals in both the Atlantic Ocean (Van Rooij et al. 2010) and the Mediterranean Sea (Taviani et al. 2017, 2019). Recently, new records and observations on N. zibrowii in the Atlantic Ocean allowed updating its ecology and distribution (Beuck et al. 2016). The giant deep-sea oyster has been meticulously described in terms of external morphology, microstructures of shell and anatomy (Wisshak et al. 2009b). The systematic placement of this species in the genus Neopycnodonte was based on morphological characteristics such as the circular muscle scar, the enlarged vermiculate chomata (see ‘neopycnodontine chomata’ in Harry 1985) and the vesicular structures in the inner shell layer. Neopycnodonte zibrowii is morphologically different from the only extant congeneric species Neopycnodonte cochlear (Poli, 1795) in several characters such as the shell architecture and outline, the absence of the resilifer bulge in the latter species and the shape and thickness of the vesicular microstructures. On the other hand, 15 years on from its discovery, molecular data are still not available for this species, thus limiting the assessment of its phylogenetic position and systematic placement.

The taxonomic assessment of oysters based on morphology can be challenging due to a high shell variability and a low number of diagnostic characters (Lam and Morton 2006; Raith et al. 2015; Salvi et al. 2021). Molecular data have a key role in species delimitation and taxonomic identification of oyster species (Lam and Morton 2003; Bieler et al. 2004; Kirkendale et al. 2004; Al-Kandari et al. 2021; Salvi et al. 2022) and would provide compelling evidence that the giant deep-sea oyster N. zibrowii is a distinct species rather than a deep-water ecophenotype of N. cochlear (Wisshak et al. 2009b).

In this study, we generated DNA sequence data of the giant deep-sea oyster N. zibrowii for both mitochondrial and nuclear markers based on the holotype and performed a multilocus phylogenetic analyses to establish its relationships with other gryphaeids. The main aims of this study are to provide: (i) a molecular phylogenetic framework for the systematic assessment of the giant deep-sea oyster, and (ii) a reliable (i.e., holotype-based) reference sequence set for multilocus DNA barcoding approaches.

BLAST searches using mitochondrial sequences (COI and 16S) of the newly sequenced specimens of H. numisma, N. cochlear and P. taniguchii confirmed the taxonomic identifications of these species (sequence identity of 99–100%). BLAST searches using the mitochondrial sequences generated from the holotype of N. zibrowii recovered as best hits sequences belonging to Gryphaeidae species (COI: sequence identity of 73.2%/72.5%/73.1% with GenBank sequences of Hyotissa sp./Neopycnodonte sp./Pycnodonte sp. respectively; 16S: sequence identity of 87.3%/87.5% with GenBank sequences of Hyotissa sp./Neopycnodonte sp. respectively). This confirms the lack of contamination during the amplification and the affiliation of this species to Gryphaeidae.

The concatenated dataset included 2409 positions (COI: 455, 16S: 449, 28S: 1078, ITS2: 427 positions) and among the 828 variable positions 436 were phylogenetically informative (i.e., parsimony informative). Maximum likelihood and Bayesian trees show two main clades: one including Hyotissa species (uBS = 95; BPP = 1), and the other one including Pycnodonte and Neopycnodonte species (uBS = 83; BPP = 0.94) (Fig. 1). Neopycnodonte zibrowii is nested within the second clade with a sister relationship with N. cochlear (uBS = 56; BPP = 0.72), whereas P. taniguchii is sister to Neopycnodonte species.

Figure 1. 

Bayesian phylogenetic tree of six Gryphaeidae species based on COI, 16S, 28S and ITS2 markers. Nodal supports indicate the values of uBS (upper) and the BPP (lower). The tree is rooted with Magallana gigas which belongs to the sister family Ostreidae Rafinesque, 1815. Specimens sequenced in this study are highlighted in bold.

The COI genetic distances (K2P/p-distance) between N. zibrowii and N. cochlear and between N. zibrowii and P. taniguchii are respectively 35.8%/28.2% and 35%/27.6% (Table 3). The 16S genetic distances (K2P/p-distance) between N. zibrowii and either N. cochlear or P. taniguchii are 13.5%/12.1% (Table 3). The mean interspecific genetic distances (K2P/p-distance) among the six gryphaeid species are 33.7% ± 4.6% / 26.8% ± 3% at the COI and 15.5% ± 4.6%/13.7% ± 3.7% at the 16S. The 28S genetic distances (K2P/p-distance) between N. zibrowii and N. cochlear and between N. zibrowii and P. taniguchii are respectively 2.5%/2.4% and 9%/8.4% (Table 4). The ITS2 genetic distances (K2P/p-distance) between N. zibrowii and N. cochlear and between N. zibrowii and P. taniguchii are respectively 15.8%/14.9% and 38.2%/29.6% (Table 4). The mean interspecific genetic distances (K2P/p-distance) among the six gryphaeid species are 5.5% ± 2.5% / 5.1% ± 2.2% at the 28S and 27.4% ± 16.8% / 22.0% ± 11.9% at the ITS2.

Benthic organisms such as oysters, with extensive phenotypic variation and few diagnostic characters, are prone to misidentification in morphological assessments. The utility of molecular characters for taxonomic identification and systematic assessment of these organisms cannot be overstated and has been proven over and over by studies on true oysters (Lam and Morton 2006; Raith et al. 2015; Salvi et al. 2021), pearl oysters (Cunha et al. 2011), tree oysters (Garzia et al. 2022) as well as gryphaeid oysters (Li et al. 2023). Conchological convergence, phenotypic plasticity, and the occurrence of cryptic species make molecular taxonomic validation of new oyster species necessary to accurately estimate the diversity of these taxa.

Our molecular phylogenetic results clearly demonstrate that N. zibrowii is a distinct species with extremely high genetic divergence from any other gryphaeid at all the markers analysed (Tables 3, 4). Neopycnodonte zibrowii is nested within the clade “Neopycnodonte/Pycnodonte” with closer affinity to N. cochlear rather than P. taniguchii (Fig. 1), and thus supporting its assignment to the genus Neopycnodonte Fischer von Waldheim, 1835 based on morphological features (Wisshak et al. 2009b). Phylogenetic relationships within this clade are not well-supported, like in a previous phylogenetic study including N. cochlear and P. taniguchii and based on COI and 28S markers (Li et al. 2021). However, the extended dataset of our study improved nodal support and allowed us to clarify the source of phylogenetic uncertainty. Indeed, at both mitochondrial markers, values of pairwise genetic distance between N. cochlear / N. zibrowii / P. taniguchii are similar and remarkably high (COI: 35.0–35.8%; 16S: 11.2–13.5%); whereas at nuclear markers the genetic distance between P. taniguchii and N. zibrowii is two-three times higher than between the latter and N. cochlear (Table 4). Such a pattern suggests that mitochondrial variation is inflated by saturation that eroded phylogenetic signal, implying an old split between taxa within this clade. Wisshak et al. (2009b) highlighted a low number of morphological and ecological differences between the genus Neopycnodonte and Pycnodonte Fischer von Waldheim, 1835 and pointed out the need for a systematic revision of the genera. Our results highlight that nuclear data will have a key role in further systematic assessment of these genera.

The availability of taxonomically validated reference sequence is a premise for DNA barcoding and metabarcoding approaches for large-scale, fast, and cost-effective molecular taxonomic identification (Hebert et al. 2003; Moritz and Cicero 2004; Schindel and Miller 2005; Salvi et al. 2020). The mitochondrial COI and 16S are the most common markers in DNA barcoding studies on Ostreidae and Gryphaeidae (Lam and Morton 2003, 2004, 2006; Kirkendale et al. 2004; Liu et al. 2011; Hsiao et al. 2016; Salvi et al. 2021). However, also nuclear rRNA markers such as 28S (Mazón-Suástegui et al. 2016) and ITS2 (Salvi et al. 2014; Salvi and Mariottini 2017) have proven useful for molecular taxonomic identification of oysters. Moreover, rRNA markers are frequently selected as target genes in eDNA metabarcoding projects (Ruppert et al. 2019). In this respect, the set of four reference (holotype-based) barcode sequences of N. zibrowii provided in this study will be useful for a wide plethora of barcoding applications in deep-sea biodiversity surveys. During the last decade, deep-sea oysters from a mounting number of regions across the Atlantic Ocean and the Mediterranean Sea have been morphologically identified as N. zibrowii: from Bay of Biscay (Van Rooij et al. 2010), Gulf of Cadiz (Gofas et al. 2010), Celtic Sea (Johnson et al. 2013), Angola and Mauritania (Beuck et al. 2016), southern Sardinia (Taviani et al. 2017), Sicilian Channel (Rueda et al. 2019) and Gulf of Naples (Taviani et al. 2019). Molecular validation of these records will be crucial to clarify the distribution of N. zibrowii and assess the phenotypic variation and ecology of this enigmatic species. Finally, given the high prevalence of cryptic species in oysters, it is not unlikely that future molecular assessments of deep-sea oysters will disclose new species.

The authors wish to thank Philippe Bouchet (Muséum National d’Histoire Naturelle, Paris) and Serge Gofas (Universidad de Málaga) for the access to holotype material, Paolo Mariottini for the specimen of Neopycnodonte cochlear and to Gustav Paulay (FLMNH) for the specimens of Pycnodonte taniguchii. MG is supported by the SYNTHESYS Project “Synthesys FR-TAF Call3_040” at The Muséum National d’Histoire Naturelle (MNHN, Paris) and the research grant of the Symposium in honour of Philippe Bouchet (World Congress of Malacology 2022 - LMU Munich). This work is part of the PhD thesis of MG under the supervision of DS (“Health and Environmental Sciences” PhD Program, University of L’Aquila).