A new species of Brachiella (Copepoda, Siphonostomatoida, Lernaeopodidae) from Peninsular Malaysia, with relegation of two genera Charopinopsis and Eobrachiella to junior synonyms of Brachiella

Both sexes of Brachiella malayensis n. sp. are described on the basis of specimens found in the nostrils of narrow-barred Spanish mackerel Scomberomorus commerson (Lacepède) collected off Besut, Malaysia. The female of this species closely resembles those of B. magna Kabata, 1968 and B. cybii Pillai, Prabha et Balaraman, 1982 but is distinguishable mainly by the body size and the proportions of the cephalosome, posterior processes and caudal rami. While examining the male, we noticed a systematic inconsistency in some lernaeopodid genera. The genus Brachiella Cuvier, 1830, represented by its type-species Brachiella thynni Cuvier, 1830, and two monotypic genera Charopinopsis Yamaguti, 1963 and Eobrachiella Ho et Do, 1984, represented by Charopinopsis quaternia (Wilson, 1935) and Eobrachiella elegans (Richiardi, 1880), respectively, share distinct synapomorphies in the embracing (vs. pinching) elongate male maxilliped and the female trunk with a pair of long, cylindrical ventroposterior processes (in addition to a pair of modified caudal rami), both of which are involved in their unique reproductive strategy. The latter two genera are herewith relegated to junior synonyms of Brachiella.


Introduction
A host fish head was purchased at a market in Besut, Terengganu, Malaysia (5.8312°N, 102.5619°E) on 15 October 2019 by the third and fourth authors. Since only the head was available for parasitological examination, the identity of the fish host species was determined using DNA barcoding based on a fragment of the cytochrome c oxidase subunit I (COI) gene in the mitochondrial genome. Approximately 1 cm 2 of tissue of the fish species was removed with scissors and preserved in a sterile 1.5 mL tube containing 95% ethanol. The total genomic DNA of the fish was isolated using a Favorgen DNA extraction Kit (Favorgen Biotech Corp., Ping-Tung 908, Taiwan). The partial COI gene of mitochondrial DNA was amplified using PCR with the universal primers of COI-Fish2 F (5 0 -TCGAC-TAATCATAAAGATATCGGCAC-3 0 ) and COI-Fish2 R (5 0 -ACTTCAGGGTGACCGAAGAATCAGAA-3 0 ) [62]. The PCR was carried out in a 25 lL reaction volume containing 18.2 lL sterile distilled water, 2.5 lL Taq buffer, 2.0 lL dNTP Mix (2.5 mM), 0.5 lL of each primer (10 lM), 0.3 lL of 5 unit lL À1 Taq polymerase (TaKaRa) and 1 lL template DNA (50 ng lL À1 ) on a thermal cycler PCR machine Veriti 96 Well Thermal Cycler (Applied Biosystems, California, USA). The thermal cycling conditions started with initial denaturation at 95°C for 5 min; 35 cycles including denaturation at 95°C for 30 s, annealing at 50°C for 30 s and elongation at 72°C for 10 min; followed by a final extension for 10 min at 72°C and then the PCR product was maintained at 4°C. Sequencing was carried out using a BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems), following the manufacturer's instructions, performed on an ABI Prism 3730XL Genetic Analyzer (Applied Biosystems). Two COI were aligned and edited using the ClustalW multiple sequence alignment program in MEGA 7 [30]. DnaSP software was used to determine the variable sites among the sequence [34]. To discover the origin of the fish species, the sequenced haplotype was queried using the Basic Local Alignment Search Tool (BLAST) against the National Center for Biotechnology (NCBI) nucleotide database and the BOLD identification engine [50]. A top species match was identified with sequence similarity of at least > 90% to avoid false positives.
The copepods were recovered from the nostrils of the fish host and were preserved in 70% ethanol for further examination. The copepods were observed in lactophenol on Humes and Gooding's [17] slides, and illustrated with the aid of a drawing tube attached to a microscope (Olympus BX53). The type specimens are deposited at the South China Sea Repository and Reference Center, University Malaysia Terengganu, Malaysia (UMTCrus 1099 and UMTCrus 1100). The terminology used follows Huys and Boxshall [18] and Kabata [25].

Host identification
After removal of low quality sequences at the 5 0 and 3 0 ends, a 629 bp DNA barcode was obtained in FASTA format. The sequence was deposited in GenBank with the accession number MT423724. The host sequence showed 100% similarity with sequences of the same species respectively in GenBank (MG220579) and in BOLD identification engine [50], including sequences of the same species collected off the coast of China [5]. This confirms the identification of the host species. Diagnosis (amended). Female. Body comprising cylindrical cephalothorax and trunk arranged in line; head slightly enlarged with distinct dorsal shield. Maxillae not medially fused and with bulla attached to host. Maxillipeds substantially displaced in front of maxillae. Trunk cylindrical or more or less dorsoventrally flattened, with pair of dorsoposterior processes (= modified caudal rami) and pair of ventroposterior processes. Gonopores and copulatory pores located ventrolaterally, anterior to caudal rami. Modified caudal rami almost equal to or longer than posterior processes. Antennule 3-or 4-segmented; terminal segment with typical lernaeopodid setation. Antenna heavily sclerotized, with 1-or 2-segmented endopod at angle or parallel to exopod with typical lernaeopodid setation at tip. Mandible with 2 or 3 secondary teeth. Maxillule bilobed, with 2 setae on palp and 3 setae on endite. Maxilliped subchelate; subchela with terminal claw with or without auxiliary spine. Egg sac multiseriate, long.
Male. Body consisting of large cephalothorax with distinct dorsal shield and variably developed trunk. Posterior processes of trunk present or not. Paired gonopores located anterior to caudal rami. Caudal rami unarmed or with rudimentary elements, short or long. Antennule 4-segmented, with typical lernaeopodid setation. Antenna biramous; exopod unilobed, with 2 subterminal elements; endopod 2-segmented, with typical lernaeopodid armament. Mandible and maxillule similar to those of female. Both right and left maxillae linked basally by cuticular tympanum; subchelate, 2-segmented, proximal segment stout, without inner projection at distal corner; distal segment strongly curved inward. Maxilliped embracing (not pinching), elongate, and slender, 2-segmented; corpus with 3 denticulate processes, middle of which bearing element; subchela smoothly curved inward, with denticulate process subterminally and 1 spine or process and 1 small element terminally. Male attached to caudal ramus of female.
Type Remarks. Historically, the taxonomy of Lernaeopodidae was based on female morphology. Also, Wilson [64] depended heavily on female morphology without noticing differences between males of Brachiella and other related genera. Kabata [25] pointed out that the genus Brachiella had accommodated miscellaneous species, and that only the type species B. thynni and B. magna could be adequately assigned to it. This statement was later repeated by Pillai [48]. Boxshall and Halsey [6] listed lernaeopodid genera with respective number of species. However, they mentioned "2" species of Brachiella, without listing the actual species names. In addition, Pillai et al. [49] suggested that B. elegans, B. seriolae and B. gracilis Wilson, 1908 [63] should be re-examined to clearly distinguish Brachiella from other lernaeopodid genera. The study is intended to follow suggestions of some of our predecessors.
Within the last 25 years, many species of the Brachiellagroup have been re-examined and their generic identity confirmed or changed. One of the major obstacles in the classification is the lack of male descriptions in many species.
We decided to include Brachiella cybii in our list of valid species of the newly defined Brachiella, even though the male of B. cybii has not been described. This is because the latter species, recovered from the nasal cavity of Acanthocybium solandri (Cuvier), is very similar to B. magna, collected from the gills of S. commerson, but it differs in the body proportions (see Table 1).
The original description of B. lageniformis does not provide enough detail about the male [58]. A re-description of both sexes of B. lageniformis provided by Kabata [24] showed that the species has no synapomorphy of Brachiella redefined in our paper. It should be newly transferred to Parabrachiella (see below in Section "Discussion"), because it was not listed in Piasecki et al. [46] nor Lebepe and Dippenaar [31].
Brachiella parva has three pairs of short and thick posterior processes in the trunk [40]. Although the male was also described and illustrated by Nuñes-Ruivo [40], the maxillipeds were not described in detail. Judging from the figure of the whole body (Fig. 2i, p. 98), the male maxilliped seems to be short, unlike that of B. thynni.
Another species, that may be erroneously synonymised with B. magna is Epibrachiella magna Song et Chen, 1976 [55]. Its female, however, bears single maxillary processes and three pairs of posterior processes (one pair corresponding to caudal rami) (see Piasecki et al. [46]), and the male does not have long maxillipeds such as seen in B. thynni (see Fig. 8L in [53]). This species is distinctly different from B. magna, and should be moved to the genus Thysanote Krøyer, 1863 as Thysanote magna (Song et Chen, 1976) (new combination).
The genus Charopinopsis was established to accommodate a single species Charopinus quaternius (Wilson, 1935) by Yamaguti [66], in which only the generic diagnosis was provided without any comparison between the newly established genus and its related genus Charopinus. The reason behind Wilson's [65] suggestion about the alleged relation of this copepod with the Charopinus-branch was his false belief about the lack of bulla in C. quaternius. This false assumption was sustained by Yamaguti [66]. Later, Kabata [25] recognized the validity of the monotypic genus in consideration of the evolution of the family Lernaeopodidae, and pointed out its phylogenetic position on the Brachiella lineage. Kabata [21] and Ho and Do [16] carefully redescribed Charopinopsis quaternia. Only Ho and Do [16], however, redescribed the male. Kabata [21] proved that the species indeed has a small bulla and listed Brachiella coryphaenae Pearse, 1952 [42] as a junior synonym of C. quaternia.
The genus Eobrachiella was erected by Ho and Do [16] to accommodate B. elegans. The main differences between Eobrachiella and Charopinopsis were the presence or absence of an The measurements for B. thynni were based on Wilson [64]; measurements for B. elegans, B. quaternia, and B. seriolae were based on figures or measurement of Ho and Do [16]; imMXPs = inner margin of maxilliped subchela. anal tubercle in the female and the shape of the caudal rami in the male [16]. However, these three genera share unique maxillipeds of the males, which can be regarded as a distinct synapomorphy. According to Ho and Do [16], the elongate maxillipeds of the males are most likely specialized to embrace the cylindrical ventroposterior process of the females, rather than pinching to avoid detachment from the females parasitic on fast-swimming pelagic fish. The male maxilliped is evidently characterized by three denticulated pads along the inner margin of the corpus, the middle of which has a setule, and an inner, terminal, denticulated pad of the subchela. Such an embracing maxilliped is found in the following six species, including the new species described herein: B. thynni; B. elegans; B. quaternia; B. seriolae; B. magna; and B. malayensis n. sp. described below. Brachiella seriolae was described by Yamaguti and Yamasu [67], but later, was relegated to a subspecies of Eobrachiella elegans by Ho and Do [16]. However, we think in consideration of the morphology of both sexes that these are two separate species. In females, the proportion of the cephalosome to the trunk and the fine structure of the antenna are different between these species (Table 1). In males, the general body shape and the morphology of the maxillipeds differ remarkably between these species (Table 2). In addition, these parasitized two different host species of the genus Seriola inhabiting the distant localities: B. elegans in the Atlantic and B. seriolae in the Pacific ( Table 1).
Females of these six species share the following features: (1) the female body plan belongs to "Type A" sensu Kabata [25], meaning that the cephalosome and trunk are positioned in line (presumably in a relatively primitive condition); (2) the cephalosome is cylindrical (advanced); (3) pairs of ventro-and dorso-posterior processes of the trunk are well developed, elongate (advanced); (4) the antennae are not prehensile (advanced); (5) the mandible bears primary and secondary teeth (advanced); (6) the maxillae are not medially fused (primitive). Since the six above-mentioned species are well defined by the synapomorphic states of the male maxillipeds and female bodies, we propose herein that Charopinopsis and Eobrachiella are relegated to junior synonyms of Brachiella (new synonymy).
A closely related genus Parabrachiella is now a very specious taxon but unfortunately, in many species only females are known. Since a revision of the genus is beyond the scope of this paper, it should be briefly commented on the basis of the male maxillipeds of the type species as follows: male maxilliped subchelate of pinching type, similar in structure and function to maxilla; corpus unarmed; subchela short, with claw positioned at right angle to shaft; tip of claw fitting depression of medial triangular process of corpus during its closure (cf. Fig. 8, Plate LXIII in [53]). Piasecki et al. [46] transferred B. elegans and B. seriolae, both of which were assigned to Eobrachiella by Ho and Do [16], to the genus Parabrachiella without any comment on Eobrachiella.
The homology of ventro-and dorso-posterior processes of the females of Brachiella can be traced in consideration of the counterparts of the males of B. thynii and B. malayensis n. sp. The dorso-posterior processes represent caudal rami (see Fig. 2C). Antennule (Fig. 1D) 4-segmented with segments distinctly tapering distally; Second segment with reduced minute seta (whip), third segment unarmed, terminal segment equipped with at least six elements (of unequal size) at tip. Antenna (Fig. 1E) biramous, heavily sclerotized; endopod rudimentary, 1-segmented, positioned at right angle to exopod, having three short elements at tip; exopod lobate, apparently unarmed. Mandible (Fig. 1F) with three primary, three secondary and five basal teeth. Maxillule (Fig. 1G) bilobed, with two short setae on palp and three short setae on endite. Maxilla partly broken during removal from host, lacking terminal parts. Maxilliped (Figs. 1H and 1I) with stout corpus with short inner seta midway; subchela with short element at 1/3 length of inner margin, 1 large terminal claw and 1 minute auxiliary spine.
Remarks. Brachiella malayensis n. sp. closely resembles B. magna from Australia and B. cybii from India. These two known species were described only based on a single female. Brachiella magna was found on the gills of S. commerson in Australian waters [22]. Later, Pillai et al. [49] described B. cybii, found in the nasal cavity of A. solandri collected off Trivandrum, India.
Brachiella malayensis n. sp. can be distinguished from B. magna in: (1) body length (6.4 mm vs. 15 mm); (2) different proportions of the cephalosome, trunk, posterior processes and caudal rami (see Table 1); (3) the shape of dorsalmost tooth of the mandibular cutting edge (small vs. large); (4) the number of elements of the maxillulary endite (2 vs. 1); (5) lengths of setae on the maxillary palp (short vs. long); (6) group of denticles near the maxillipedal axilla (absent vs. present). Brachiella cybii was very briefly described on the basis of the habitus only by Pillai et al. [49]. However, B. malayensis n. sp. can be differentiated from it by the body size (6.4 mm vs. 20 mm) and different body proportions ( Table 1).
The discovery of the male of the new species reminded us of similarities of males among the type species of Brachiella, B. thynii, and two other lernaeopodid genera Charopinopsis and Eobrachiella. The latter genera are synonymized with Brachiella in the present study. In fact, the male of B. malayensis n. sp. is very similar to those of B. thynni and B. elegans in having a long trunk, but is differentiated by the trunk longer than the cephalothorax ( Table 2). The male is also distinguished in the shape of the maxillule, maxilla, and maxilliped from these two congeners (Table 2).

Discussion
The family Lernaeopodidae has accommodated 46 genera and over 265 species to date, while 20 genera are monotypic [6]. According to Boxshall and Halsey [6], the taxonomy of this family "remains in urgent need of revision". Although Wilson [64], Kabata [21-23, 25, 26], Kabata and Bowman [27], Castro Romero and Baeza Kuroki [9], and Piasecki et al. [46] intensively reviewed the taxonomy of the family, diagnoses of some genera are still ambiguous. This confusion is mainly due to profound modification in the body plan and superficial similarity of appendages, which sometimes makes the phylogenetic relationships obscure [25]. Kabata's [25] vision of the evolutionary trends and phylogenetic relationships among lernaeopodids is suggestive as a landmark for re-definitions of these problematic taxa.
Brachiella can be well defined by the general body plans and long caudal rami of the females and the cephalic appendages of both sexes, especially the distinct synapomorphic state of the male maxillipeds, whereas it exhibits variability in morphology of other body parts of both sexes (Tables 1 and 2). In females, length ratios of cephalothorax to trunk, degrees of trunk expansion, and relative lengths of posterior processes to caudal rami are variable (Table 1). In males, the development of the trunk is highly variable ( Table 2). Kabata [25] mentioned that the cephalosome, bulla, and trunk of lernaeopodid females are involved in feeding, attachment, and reproduction, respectively. The cephalosome and trunk of the males are also related to feeding and attachment to the mates and to reproduction, respectively. Therefore, it is likely that such interspecific variation depends on their own reproductive and adaptive strategies on different hosts.
The majority of the species related to the Brachiella-branch have been included in the genus Parabrachiella. Those with maxillary processes were relegated to Thysanote [45]. Only some of them were described based on both sexes. Therefore, the genus Parabrachiella constitutes a potential source of Brachiella species (as presently defined), after their males are discovered and properly described. Piasecki et al. [46] provided a checklist of valid species of Parabrachiella covering a total of 67 species. After 10 years, the list should be updated taking into consideration additional publications [14,16,31,37,48].
Recently, intraspecific variation of Parabrachiella platensis Montes, Castro-Romero et Martorelli, 2017 was found between specimens collected from nostrils and fins of Mugil liza Valenciennes. The species identification was confirmed by DNA barcoding using mt-DNA COI (similarity 99.8%) [37]. Minor morphological differences between these specimens were found especially in females: trunk length, shapes of cephalosome, anal slit, posterior truncal processes, labium and maxilla, and armature and elements of antennule, maxillule, and maxilliped [37]. On the other hand, B. magna and B. malayensis n. sp. found from the gills and nostrils of the same host S. commerson, respectively, also had a few differences as in the abovementioned remarks. Thus, in the future, a comparison of the DNA sequences of B. malayensis n. sp. and B. magna might be useful to elucidate the validity of species if such variation also occurs between these two species. Although the distal tips (around the bulla) of the maxillae of B. malayensis n. sp. were accidentally lost during the handling procedure, if the bulla are located at the anteriormost part of the broken right maxilla (left shorter), such short maxillae seem to be adaptive in the narrower habitat, the nostril. The discovery of males of B. magna and B. cybii may confirm the present conclusion.
As Ho and Do [16] have precisely pointed out, males of Brachiella (as Eobrachiella) have an elongate maxilliped which suits embracing the cylindrical caudal ramus rather than pinching, implying avoidance of detachment from the mates on fastswimming pelagic fish. In fact, hosts of Brachiella belong to Carangidae, Coryphaenidae, Peristediidae, Pomatomidae, Sciaenidae, and Scombridae, and Sphyraenidae which are mostly epipelagic [13] (Table 3). Ho and Do's [16] hypothesis is supported by the present study. In situ sites of the males on the females were illustrated in some references. In B. seriolae and B. quaternia, a male of each species was illustrated to attach himself to the base of the ventroposterior process in Figures 10F, 13B, C and 14A, B of Ho and Do [16], respectively. On the other hand, a male adhered to the terminal portion of the ventroposterior process of a female of B. thynni (see [54]). Since paired copulatory pores are located between the ventroposterior processes, the males seem to prefer attachment to the ventroposterior processes rather than the dorsoposterior ones. Moreover, these different positions of the males on the female ventroposterior processes imply that they are capable of moving freely along the processes. The present study has focused more attention on the reproductive organs and behavior of lernaeopodids to improve understanding of the taxonomy and phylogeny of the family. In fact, free-living calanoid copepods exhibit highly variable configurations in the female reproductive systems (e.g., Barthélémy et al. [1]; Ohtsuka and Huys [41]), which was a key innovation to reconsider the phylogeny of the order. Fish name validity verified after Fricke et al. [12]; -: no data provided.