A new genus and species of the family Pseudocyclopidae, Sipadantonius roihani gen. et sp. nov., was described based on specimens collected using a light trap in the marine cave of Sipadan Island, Sabah, Malaysia. The new genus is most related to Pinkertonius, primarily based on the similarity observed in the armament of ancestral segment IV of the male antennules, the armament of the female P5 Exp-3, the segmentation of the male P5, the armament of the maxillular basal exite, and the relative length of the ancestral segment XXVII of the antennules. Nevertheless, it distinguishes itself from Pinkertonius and all other genera of the family by the absence of the lateral seta of the basis of all swimming legs, the presence of an inner seta on the coxa of the female P5, the reduction of furcal setae I and III, as well as the specific armament of the ancestral segment XX of the antennules and the maxillular coxal endite. The female of Sipadantonius roihani gen. et sp. nov. has aesthetascs on the ancestral segments IV and XX of the antennules, as well as six setae on the maxillular coxal endite, exhibiting the most plesiomorphic characteristics of the family Pseudocyclopidae. The latter characteristic has not been recorded in the order Calanoida. It was hypothesised that the new species was a particle feeder living in the pelagic zone of the marine cave. The existence of the new species supported the assumption that the regional distribution of the family Pseudocyclopidae exhibited the Tethyan track, which might have been the subsequent result of the colonisation of the habitats prior to the closure of the Tethys Sea.

Crustacea, Southeast Asia, systematics, taxonomy, Zooplankton

Based on the phylogenetic research conducted by Bradford-Grieve et al. (2014), 14 calanoid genera were recognised as members of the family Pseudocyclopidae Giesbrecht, 1893, which included Badijella Kršinic, 2005, Boholina Fosshagen, in Fosshagen & Iliffe, 1989, Brattstromia Fosshagen, in Fosshagen & Iliffe, 1991, Exumella Fosshagen, 1970, Exumellina Fosshagen, 1998, Hondurella Suárez-Morales & Iliffe, 2007, Normancavia Fosshagen & Iliffe, 2003, Pinkertonius Bradford-Grieve, Boxshall & Blanco-Bercial, 2014, Placocalanus Fosshagen, 1970, Pseudocyclops Brady, 1872, Ridgewayia Thompson I.C. & Scott A., 1903, Robpalmeria Fosshagen & Iliffe, 2003, Stargatia Fosshagen & Iliffe, 2003 and Stygoridgewayia Tang, Barron & Goater, 2008. The family was united with the family Epacteriscidae Fosshagen, 1973, thereby forming the superfamily Pseudocyclopoidea Giesbrecht, 1893, which was the basal clade of the order Calanoida Sars, 1903, based on molecular phylogenetic study (Bradford-Grieve et al., 2014). Among the listed genera, Pinkertonius was regarded as the basal taxon of the family (Bradford-Grieve et al. 2014).

Pseudocyclopid copepods are prevalent in shallow benthopelagic or anchialine cave habitats within tropical and subtropical marine waters globally, as documented by various researchers (e.g., Fosshagen and Iliffe 1989, 1991, 1998, 2003; Boxshall and Jaume 2012; Bradford-Grieve et al. 2014; Ohtsuka et al. 1996, 2000; Suárez-Morales and Iliffe 2007; Figueroa 2011; Moon and Soh 2014). Additionally, some genera and species are recognised from freshwater subterranean habitats, including Stygoridgewayia trispinisa Tang, Barron & Goater, 2008, Boholina laorsriae Boonyanusith, Wongkamhaeng & Athibai, 2020 and Boholina reducta Tran & Chang, 2020 (Tang et al. 2008; Boonyanusith et al. 2020; Tran and Chang 2020).

In the past decade, a great deal of research was conducted regarding the zooplankton ecosystem of Malaysian waters (Rezai et al. 2009; Yoshida et al. 2012; Nakajima et al. 2013; Balqiah and Rahim 2021), resulting in the documentation of more than 230 species. Nevertheless, the documentation regarding the prevalence of endemic zooplankton species remained insufficient. To date, only four species were recognised as endemic or host-specific, namely Brachiella malayensis Ohtsuka, Piasecki, Ismail & Kamarudin, 2020, Kensakia parva Harris V.A. & Iwasaki, 1997, Labidocera jaafari Othman, 1986 and Sipadania celerinae Humes & Lane, 1993. A recent biological survey conducted at the Sipadan Turtle Tomb yielded a significant number of specimens of calanoid copepods, which could not be assigned to any described species.

Morphologically, the new species appeared to resemble the genus Pinkertonius in the armament of the ancestral segment IV of the male antennule, the armament of the female P5 Exp-3, and the segmentation of both rami of the male P5. However, it bore a specific combination of the characteristics, particularly regarding the armament of the swimming legs and mouthparts; therefore, the status of a new genus was justified for the Malaysian specimens, and the name of Sipadantonius gen. nov. was proposed. The description and illustrations of the new taxon were provided hereafter.

Specimens were collected in the Turtle Tomb of Sipadan Island, Sabah, Malaysia, on 3 August 2023 (Fig. 1). Sampling was conducted utilising a modified light trap at a depth of 20–24 m, which was deployed at the bottom substrate for ~ 24 hours. The specimens utilised for morphological descriptions were preserved in 10% buffered formalin prior to examination. Copepods were meticulously sorted under a stereomicroscope, followed by fixation and storage in 70% ethanol in the laboratory. Subsequently, a few specimens were soaked in the mixture of glycerol and 70% ethanol (approximate ratio 1/10 v/v) for 30 min. After that, animals of both sexes were individually transferred into a drop of 40% glycerol on a glass slide and dissected under a stereomicroscope. The material was ultimately mounted with a coverslip and subsequently examined under a Nikon ECLIPSE E200 compound light microscope at 1000 × magnification.

Figure 1. 

Geographical location of the type locality of Sipadantonius roihani gen. et sp. nov.

The habitus and appendages were examined and subsequently drawn utilising a drawing tube attached to a compound microscope at 400 × and 1000 × magnifications, respectively. Description was made by adopting the terminology established by Huys and Boxshall (1991) and employing the following abbreviations: ae = aesthetasc, I = spine, Endp = endopod, Exp = exopod, Endp-1 (2, 3) = proximal (middle, distal) segment of the endopod of the swimming legs, Exp-1 (2, 3) = proximal (middle, distal) segment of the exopod of the swimming legs, P1–P6 = first to sixth swimming legs, Seta I–VII = first to seventh furcal seta, Seta I = anterolateral accessory seta, Seta II = anterolateral seta, Seta III = postereolateral seta, Seta IV = outer terminal seta, Seta V = inner terminal seta, Seta VI = terminal accessory seta, and Seta VII = dorsal seta.

The order Calanoida had been divided into nine superfamilies by Andronov (1974). Subsequently, after Fosshagen and Iliffe (1985) raised the family Platycopiidae Sars, 1911 to full order rank, namely Platycopioida Fosshagen, 1985 and following the introduction of two new superfamilies by Park (1986), ten superfamilies were recognised. They included Augaptiloidea Sars, 1905, Bathypontioidea Brodsky, 1950, Centropagoidea Giesbrecht, 1893, Clausocalanoidea Giesbrecht, 1893, Epacteriscoidea Fosshagen, 1973, Eucalanoidea Giesbrecht, 1893, Megacalanoidea Sewell, 1947, Pseudocyclopoidea Giesbrecht, 1893, Ryocalanoidea Andronov, 1974, and Spinocalanoidea Vervoort, 1951, in which Andronov (2007) suggested the merger of Epacteriscoidea within Pseudocyclopoidea. According to a morphology-based phylogenetic study, the monophyly of numerous recognised superfamilies was established (Bradford-Grieve et al. 2010); the finding was subsequently corroborated by the multiple gene-based phylogenetic analysis (Blanco-Bercial et al. 2011). Bradford-Grieve et al. (2014) indicated that the families Boholinidae, Pseudocyclopidae, Ridgewayiidae, and Epacteriscidae were included in the re-definition of the superfamily Pseudocyclopoidea, based on the recovery of the monophyletic relatedness among the superfamilies Epacteriscoidea and Pseudocyclopoidea. Although the clade was formed with low jackknife support due to a considerable number of homoplastic characters, evidence indicated that each superfamily possessed a distinct set of morphological characteristics that reflected the unique genetic basis of the superfamily. This may have been the reason that the previous gene-based phylogenetic analysis of the Calanoida aligned with the earlier morphology-based phylogenetic inference (Park 1986; Bradford-Grieve et al. 2010, 2014). In the case under consideration, it was observed that most of the ancestral characteristics of Pseudocyclopoidea were retained in the new species. They included a triangular rostral plate without a rostral filament, a complete separation of ancestral segment IV, which resulted in antennules with 27 segments, a maxilliped with six-segmented Endp, P5 with both rami three-segmented, separated genital apertures each hidden under its operculum, and presence of inner seta on Exp-1 of the female P5 (Andronov 2014). Furthermore, the weak expression of the sexual dimorphism of P5 and the configuration of swimming legs, which were regarded as indicative of the origin from a single species of the Pseudocyclopoidea (Andronov 2014), is relevant for the new species. The characteristic of P2 Exp-3 with two outer spines was observed in other families, including Calanidae Dana, 1846, Paracalanidae Giesbrecht, 1893, Pseudodiaptomidae Sars, 1902, Sulcanidae Nicholls, 1945, and Tortanidae Sars, 1902. The representatives of these families have P5 bearing Exp-3 with two outer spines at most. Furthermore, the considerable reduction of Enp of the female P5 and the strong expression of sexual dimorphism of P5 were evident in most of the aforementioned families (Boxshall and Halsey 2004; Andronov 2014).

Bradford-Grieve et al. (2014) noted certain modifications of the mandible, maxilla, and maxilliped that were associated with the families Epacteriscidae and Pseudocyclopidae, thereby indicating their feeding habits (e.g., Fosshagen 1968; Fosshagen et al. 2001; Jaume and Humphreys 2001; Fosshagen and Iliffe 2003, 2004a, 2004b, 2007; Suárez-Morales and Iliffe 2007; Boonyanusith et al. 2020). Recently, the family Epacteriscidae is comprised of Balinella Fosshagen, Boxshall & Iliffe, 2001, Bofuriella Fosshagen, Boxshall & Iliffe, 2001, Bomburiella Fosshagen, Boxshall & Iliffe, 2001, Cryptonectes Fosshagen & Iliffe, 2004, Iboyella Boxshall & Jaume, 2003, Oinella Fosshagen, Boxshall & Iliffe, 2001, Bunderia Jaume & Humphreys, 2001, Edaxiella Fosshagen, Boxshall & Iliffe, 2001, Epacteriscus Fosshagen, 1973, Enantronoides Fosshagen, Boxshall & Iliffe, 2001, Enantiosis Barr, 1984, Enantronia Fosshagen, Boxshall & Iliffe, 2001, Gloinella Fosshagen, Boxshall & Iliffe, 2001, Minnonectes Fosshagen & Iliffe, 2004, Erebonectes Fosshagen, in Fosshagen & Iliffe, 1985, and Miheptneria Fosshagen & Iliffe, 2004) had been considered as raptorial feeders, based on the characteristics such as the gnathobase of the mandible with enlarged ventral teeth, the mandibular palp with reduced Endp and the well-developed Exp forming the main axis of the palp, the transformation of setae on the distal part of the maxilla and maxilliped to stout, elongate, spinous setae or the reduction of setae on both appendages, and the maxilliped bearing a shortened Endp. On the other hand, the feeding habits of the family Pseudocyclopidae were observed to be more variable. Genera such as Boholina, Hondurella, Ridgewayia and Stygoridgewayia were presumably particle feeders (Fosshagen and Iliffe 2003; Suárez-Morales and Iliffe 2007; Tang et al. 2008; Boonyanusith et al. 2020), Exumella were considered to be a benthic scavenger, based on the reduction of the seta of the Endp of the mandible (Jaume and Boxshall 1995), in addition to Exumellina and Stargatia are believed raptorial feeders, deduced from the paddle-like adaptive feature of the mandible and the Endp of maxillule (Fosshagen and Iliffe 1998). The new species was presumed to be a particle feeder, demonstrating a close affinity with the family Pseudocyclopidae from a taxonomic perspective based on the four following morphological characteristics, which include: (1) the mandible armed with numerous small teeth on the cutting-edge of the gnathobase; (2) the well-developed Endp of the mandible with four and more than nine setae on the proximal and the distal segments, respectively, and representing the main axis of the mandibular palp; (3) the maxilla with elongate endite on the basis and armed with normal plumose setae; (4) the Endp of the maxilliped elongate. The new species exhibited a modification of a seta on the Endp of maxilliped, also observed in Brattstromia (Fosshagen and Iliffe 1991); however, the shape of the transformed seta differed between the new species and the genus mentioned above. In Brattstromia, the tip of the seta was serrated and sharpened, which related to its cutting function (Fosshagen and Iliffe 1991). In the new species, the tip of the seta was ornamented with long spinules, which bore a resemblance to the feather-like spine that has frequently been observed in the cave-dwelling harpacticoid genus Elaphoidella (Brancelj et al. 2010). It was hypothesised that the transformed seta on the maxilliped of the new taxon could be utilised as a collecting apparatus, in accordance with the assumption posited by Brancelj et al. (2010), who postulated that the modified seta plays a role in collecting food particles, such as fine detritus or bacteria.

The existence (= the presence/absence) of aesthetascs on certain segments of the antennule was employed in morphology-based phylogenetic analysis of the superfamily Pseudocyclopoidea (Bradford-Grieve et al. 2014). The presence of aesthetascs on all other ancestral segments was utilised for the analyses; I, II, III, VII, XVI, XXII, XXIII, XXIV, and XXVI–XXVIII of the ancestral segments were refrained. This was likely attributable to the low phylogenetic implication of the characteristic, which may have indicated the consistency in the existence of aesthetascs on the aforementioned ancestral structures in the superfamily Pseudocyclopoidea. Pinkertonius and Sipadantonius gen. nov., as well as the three basal taxa of the superfamily, namely Caiconectes, Erebonectoides, and Azygonectes, consistently exhibited a presence of aesthetascs on the ancestral segments I, III, VII, XVI, and XXVII–XXVIII and consistently lacking aesthetascs on the ancestral segments II, XXII, XXIII, XXIV, and XXVI (Fosshagen and Iliffe 1994, 2007; Bradford-Grieve et al. 2014). Thus, the presence or absence of aesthetascs in the segments mentioned above was presumably plesiomorphic in the superfamily. In contrast, the retention of aesthetascs on the ancestral segments IV, XIX and XX of the antennule in both sexes indicated that the new species also exhibited an affinity to the Epacteriscidae. Nevertheless, an inconsistency in the presence of an aesthetasc on a particular ancestral segment was frequently observed in the Epacteriscidae, a close relative of the Pseudocyclopidae, thereby suggesting that the characteristic was homoplastic. The initial instance of an inconsistent appearance in the retention of aesthetascs in the Epacteriscidae was observed in the genus Oinella; only the taxon lacked an aesthetasc on the ancestral segment IV of the female antennule (Fosshagen et al. 2001), whereas this characteristic was evident in females of all other genera. The retention of aesthetascs on ancestral segments XIX and XX was sporadically observed within the superfamily Pseudocyclopoidea, predominantly among the Epacteriscidae. In Caiconectes, an aesthetasc was present on the ancestral segments XIX and XX of the female antennules and on the segment XIX of the males (Fosshagen and Iliffe 2007). In the female of Erebonectoides and Oinella, as well as the male of Enantiosis, Bomburiella and Exumella (Fosshagen and Iliffe 1994; Jaume and Boxshall 1995; Fosshagen et al. 2001), an aesthetasc was documented to be present in the ancestral segment XIX of the antennule, and the retention of an aesthetasc on ancestral segment XX has been reported in the females of Gloinella (Fosshagen et al. 2001). The examples supported the assumption that the retention of aesthetascs in these three segments could occur independently within the superfamily Pseudocyclopoidea. Furthermore, the presence of a distolateral process on the P1 Exp-2 and the absence of the rostral filament supported the placement of the new species within the family Pseudocyclopidae. Although the process was identified in several other genera, including Badijella, Boholina, Stygoridgewayia, Pseudocyclops and Ridgewayia (Ohtsuka et al. 1999, 2000; Kršinic 2005; Tang et al. 2008; Boonyanusith et al. 2020), it had not been reported in Epacteriscidae. In contrast, the rostral filament, which was consistently present in the representatives of Epacteriscidae, was absent in the new species. Moreover, the rostral filament has been reported from numerous genera of Pseudocyclopidae, such as Badijella, Robpalmeria, Normancavia, Exumella and Pinkertonius. The characteristics of their buccal appendages exhibited an adaptive feature of particle feeders or scavengers. Furthermore, the maxillular coxal epipodite which bore seta 9, as observed in the new species, differed from most of Epacteriscidae, in which seta 9 was absent (Jaume and Boxshall 1995; Fosshagen and Iliffe 2003; Kršinic 2005; Bradford-Grieve et al. 2014).

Five plesiomorphic characteristics of Pinkertonius were noted: (1) the female antennule segmentation between the ancestral segments II and III; (2) the male antennule with aesthetasc on the ancestral segment IV; (3) Endp-1 of the male right P5 with medial seta; (4) Exp-3 of the female P5 with medial seta 4; (5) the P5 with three-segmented Endp in both sexes. The new species exhibited a close phylogenetic relatedness with Pinkertonius, characterised by the presence of the combination of characteristics (2), (4), and (5), which served to distinguish the new species and Pinkertonius from all other Pseudocyclopidae. The characteristic (2) demonstrated a high phylogenetic signal in morphology-based phylogenetic analyses of the superfamily (Bradford-Grieve et al. 2014), having been shared by nearly all of the Epacteriscidae and three additional genera, namely Azygonectes, Erebonectoides and Caiconectes. The assumption regarding the close relationship between Sipadantonius gen. nov. and Pinkertonius was supported by the observation that the characteristic of the P5 with a three-segmented Endp in both sexes was shared by a few numbers of calanoid genera of Pseudocyclopidae which includes Sipadantonius gen. nov., Pinkertonius and Exumella. Nevertheless, the reduction of the Exp was deemed relevant in Exumella (Jaume and Boxshall 1995; Suárez-Morales and Iliffe 2005). Furthermore, the Endp of the male P5 was derived in various patterns of segmentation reduction in other Pseudocyclopidae. For instance, it was observed that the Endp of the male P5 exhibited a two-segmented ramus on one side and a three-segmented on the other one in Brattstromia, Exumellina, and Stargatia (Fosshagen and Iliffe 1985, 1998, 2003). In contrast, it bore a two-segmented ramus on both sides in Badijella and one-segmented rami on both sides in Boholina, Hondurella, Stygoridgewayia or Robpalmeria (Fosshagen and Iliffe 1989, 2003; Kršinic 2005; Suárez-Morales and Iliffe 2007; Tang et al. 2008). The sole significant distinction observed in the male P5 Endp between the new species and Pinkertonius was the reduction of the inner seta on the Endp-1.

Certain characteristics link Pinkertonius to the family Pseudocyclopidae, including: (1) the existence of an aesthetasc on a certain segment of the female antennule; (2) segmentation and armament of the Endp of the mandible; (3) the armament of epipodite of the maxillule; (4) ornamentation of the basis of the P1, and (5) segmentation and ornamentation of the male P5 (Bradford-Grieve et al. 2014). Based on the aforementioned characters, the new species and Pinkertonius were differentiated from one another by the characters (1), (2), and (5). Nevertheless, this study posited that the new species and Pinkertonius were close relatives due to their shared characteristics of having aesthetascs on the ancestral segment IV of the male antennule. Furthermore, the variation in the number of armature elements of Endp-2 of the mandible and ornamentation of Exp-2 of the male right P5 had commonly been observed in Pseudocyclopidae, even among the representatives of a monophyletic clade. This study, thus, posited that the differences in these two characters between the new species and Pinkertonius were merely the result of the reduction/transformation of the structure within the closely related phylogenetic lineage. The two examples that supported the argument were the differences in the number of the armature element of Endp-2 among the genera Hondurella and Stygoridgewayia, as well as the differences in the ornamentation with the medial process of Exp-2 of the male right P5 among the genera Placocalanus and Pseudocyclops. Hondurella and Stygoridgewayia were identified as the sister taxon, constituting a monophyletic clade in morphology-based phylogenetic analyses (Bradford-Grieve et al. 2014). However, Hondurella bore nine setae on Endp-2 of the mandible instead of 11 setae presented in Stygoridgewayia (Suárez-Morales and Iliffe 2007; Tang et al. 2008). Likewise, Placocalanus and Pseudocyclops constitute a monophyletic clade; however, the ornamentation with a medial process on Exp-2 of the male right P5 was present in Pseudocyclops but not in Placocalanus (Bradford-Grieve et al. 2014). The number of setae on Endp-2 of the mandible varied between six and eleven setae among the representatives of Pseudocyclopidae (Bradford-Grieve et al. 2014). Furthermore, the ornamentation with the medial process of Exp-2 of the male right P5 was absent in numerous genera, including Boholina, Hondurella, Stygoridgewayia, and Placocalanus (Ohtsuka et al. 1996; Suárez-Morales and Iliffe 2007; Tang et al. 2008; Boonyanusith et al. 2020).

It was noted that the new species bore six elements on the maxillular coxal endite and furcal rami, lacking furcal seta I and seta III. The presence of six elements on the maxillular coxal endite constituted the most plesiomorphic characteristic among the representatives of Calanoida, in which the maximum number of setae on the endite was five (Boxshall and Halsey 2004). On the other hand, the reduction of furcal seta I and sata III could be considered an apomorphic characteristic of the new species. These characteristics supported the placement of the new species outside either the family Pseudocyclopidae or the superfamily Pseudocyclopoidea. Nevertheless, the setation of the armament of the maxillular coxal endite and furcal rami exhibited variability, even within the same genus. The occurrence of six elements on the maxillular coxal endite was reported from the order Misophrioida, as well as the families Canuellidae Lang, 1944 and Longipediidae Boeck, 1865 (Boxshall and Halsey 2004). Examples of the difference in the setation of the maxillular coxal endite among representatives of a genus were Misophriella Boxshall, 1983 and Longipedia Claus, 1863. Six elements were present on the endite of both M. schminkei Martínez Arbizu & Jaume, 1999 and L. ulleungensis Bang, Moon & Back, 2021 (Martínez Arbizu and Jaume 1999; Bang et al. 2021). Furthermore, five elements were present in M. tetraspina Boxshall, 1983 and L. koreana Bang, Moon & Back, 2021 (Boxshall 1983; Bang et al. 2021).

A distinctive characteristic of the new species was the reduction of furcal setae I and III. In Pseudocyclopoidea, the complete reduction of furcal setae was reported for the furcal setae I and II, with the latter being specific to Miheptneria (Bradford-Grieve et al. 2014). In other families, the reduction of furcal setae was reported in Temorites Sars, 1900 of the Bathypontiidae Brodsky, 1950, and Fosshagenia Suárez-Morales & Iliffe, 1996, of the Fosshageniidae Suárez-Morales & Iliffe, 1996 (Sars 1900; Suárez-Morales and Iliffe 1996; Fosshagen and Iliffe 2004b). However, the retained setae were interpreted as seta III to seta VI (Bradford-Grieve et al. 2014). Thus, the reduction of furcal seta III was identified as a distinctive characteristic of the new species, thereby facilitating the classification of this new species within a newly established genus. An inconsistency in the setation of the furcal rami was found in certain other copepod taxa, even within the same genus, suggesting an influence of the ecological differentiation or the niche partitioning on the setation of furcal rami. An example was the genus Pseudograeteriella Brancelj, Boonyanusith & Sanoamuang, 2019 in which the variation of the armament of the furcal rami was found, even the two most closely related representatives were encountered within the same cave (Sanoamuang et al. 2019). Based on the evidence mentioned earlier, this study assumed that the presence of six elements on the maxillular coxal endite represented merely the retention of the plesiomorphic characteristic in the new species, while the reduction of the armament of furcal rami was influenced by the ecological differentiation, as had been observed in other Copepoda. Therefore, the new species could be classified in the family Pseudocyclopidae.

The superfamily Pseudocyclopoidea had frequently been reported from cave habitats, and it appeared likely that no relevant adaptive feature was specific to copepods inhabiting this type of environment. It was possible that such a morphological adaptation/modification was related to the zones of the water column in which the copepods lived (Boonyanusith et al. 2020). In general, the relatively long antennules (i.e., surpassing the prosome) were frequently observed in pelagic representatives collected from the water column of the cave, such as Caiconectes, Azygonectes, Exumellina and Stargatia (Fosshagen and Iliffe 1998, 2003, 2007). Conversely, the relatively short antennule, as found in some epacteriscids and pseudocyclopids like Stygoridgewayia, Boholina, and Ridgewayia, was deduced as a hyperbenthic adaptive feature (Fosshagen and Iliffe 1998; Tang et al. 2008; Boonyanusith et al. 2020). Other characteristics observed consistently in benthic copepods (Fosshagen 1973) included the compact body and strong spines on the exopods of the swimming legs (Bradford-Grieve 2002). Following the aforementioned assumption, the new taxon would be the pelagic adaptive Calanoida, characterised by long antennules surpassing the posterior margin of the second urosomite and conspicuous oar-shaped outer spines on the exopods of all swimming legs. In addition, it was posited that the hypothetical calanoid ancestor, as well as most of its relatives, are epi- or hyperbenthic colonisers (Huys and Boxshall 1991; Bradford-Grieve 2002). Thus, the colonisation of the pelagic realm was probably secondary, occurring subsequent to the divergence of the new species from the epibenthic ancestor.

The two assumptions regarding the pelagic adaptation and the secondary colonisation of the pelagic realm of the new species were likely interconnected with the geological history of the sampling area and the ecological characteristics of the cave. From a geological perspective, the Turtle Tomb is located beneath Sipadan Island, which is the top of a volcanic cone located within the complex zone formed by the convergence of three tectonic plates, namely the Philippines Sea, Indian and Pacific Plates (Bellon and Rangin 1991). Historically, the volcanic cone primarily originated from the vertical rising of the magma, followed by the extinction of the volcano and the subsequent growth of the corals on the extinct volcano. This indicated the richness of volcanic activities at the seabed in this area, attributable to the convergence of the Eurasian, Pacific and India-Australian Plates since the Miocene (Yang et al. 2024). The environmental catastrophes and toxic gases likely prompted an adaptation of the epibenthic ancestor of the new species for colonisation of the pelagic environment prior to the colonisation in the cave. Based on the richness of organic particles and the stability of the environmental conditions, the cave functions as a refugee for the new species.

The second possibility was hypothesised based on the ecological characteristics of the cave. The new species was collected in the Turtle Tomb, recognised as one of the largest marine caves of Southeast Asia, located at Sipadan Island, from ~ 20 meters below the sea surface within the cave. The cave has a large entrance measuring ~ 20 meters wide and a horizontal gallery with depths varying from ~ 18–21 meters. Crystal-clear water with settled sediment at the sampling site indicates a slow movement of water currents. The broad space of the cave, coupled with the slow movement of water currents, likely facilitates an adaptation towards the planktonic lifestyle, enabling organisms to navigate through the free water masses within the cave. Exumellina is another example of a copepod exhibiting a planktonic lifestyle, having been collected in the pelagic zone of a marine cave (i.e., Norman’s Pond Cave). Exumellina is characterised by a slender body and relatively small and equal outer spines of Exp of P1–P4 (Fosshagen and Iliffe 1998).The disjunct biogeographic distribution of cave-dwelling calanoids in the two extremities of the Tethys Sea has suggested the colonisation of the habitats where they inhibited prior to the closure of the Tethys Sea (Bradford-Grieve 2002; Belmonte 2022). In the family Pseudocyclopidae, five genera constitute a monophyletic clade, in which the genus Ridgewayia is regarded as the sister taxon of the other five genera, including Boholina, Hondurella, Placocalanus, Pseudocyclops, and Stygoridgewayia (Bradford-Grieve et al. 2014). Although Ridgewayia and Pseudocyclops are cosmopolites, the regional distribution of the other genera is relatively restricted, having been recorded from merely one or two geographical regions. In one extremity of the Tethys Sea, which corresponds to the western coast of the Pacific Ocean, the new species and the two genera of the aforementioned clade have been reported from Southeast Asia, including Boholina and Pseudocyclops (e.g., Fosshagen and Iliffe 1989; Ohtsuka et al. 1999; Boxshall and Jaume 2012; Moon and Soh 2014; Tran and Chang 2020). Additionally, Stygoridgewayia, Placocalanus, and Ridgewayia (two out of six species of the typica species-group, including R. boxshalli Barthélémy, Ohtsuka & Cuoc, 1998 and R. flemingeri Othman & Greenwood, 1988) have been reported along the western coast of the Pacific Ocean, (Ohtsuka et al. 1996, 2000; Tang et al. 2008), forming the Indo-West Pacific community. In the other extremity of the Sea, which corresponds to the coast of the Atlantic Ocean and the Mediterranean, Placocalanus and Hondurella have been found from the Honduran coast of the Caribbean Sea of the North Atlantic Ocean (Fosshagen 1970; Suárez-Morales and Iliffe 2007). Placocalanus is the only genus that has been found in the localities beyond the Caribbean Sea and the Mediterranean; moreover, the six other genera, namely Exumella, Exumellina, Stargatia, Brattstromia, Normancavia, and Robpalmeria, have been found only from these two regions (e.g., Fosshagen 1970; Grahame 1979; Fosshagen and Iliffe 1991, 1998, 2003; Jaume and Boxshall 1995; Suárez-Morales and Iliffe 2005). The disjunct pattern of the regional distribution previously described appeared to resemble that of the genus Ridgewayia, as most species of the typica species-group had been documented from localities beyond the Caribbean Seas and the Mediterranean. In contrast, nearly all the gracilis species and the marki species-groups had been found in the North Atlantic/Mediterranean (Ohtsuka et al. 2000). Ohtsuka et al. (2000) posited that the existence of a certain morphological species-group (= marki species-group) in both the Indo-West Pacific region and the North Atlantic/Mediterranean served to confirm the faunistic link between these two regions. Furthermore, the presence of Hondurella and Placocalanus constituted another example of the phenomenon mentioned above. Thus, the regional distribution of the family Pseudocyclopidae exhibited the Tethyan track.

Gratitude was extended to Roihan Han, Boy Jay Alexius Petrus, Ag Ahmad Zaki Abu Bakar, Rex Andoh, Jackson Tang, Maxime Cheminade, Por Parasu Komaradat and Fitria Mock for their assistance with the Turtle Tomb cave studies. Gratitude was expressed towards Dr. Nasrulhakim Maidin, Mr. Jufri Hj. Nasri, and Mr Johny Buis from the Sipadan Island Park (Sabah Parks) and Sabah Biodiversity Council (SaBC) for granting permission to conduct these investigations. The authors thank the reviewers for their constructive comments and helpful suggestions, which significantly enhanced the manuscript.