Introduction
The subfamily Monotaxinae of the actinopterygian fish family Lethrinidae comprises medium- to large-sized carnivorous species occurring in inshore waters over sandy or rubble substrates of coral reefs (Carpenter and Allen, Reference Carpenter and Allen1989). Commonly known as large-eye seabreams, the monotaxines are important food fishes, which are sold fresh, frozen, or dried throughout the tropical Indo-West Pacific. With 11 currently recognized species, Gymnocranius Klunzinger, 1870 is the largest genus among the four valid genera within the Monotaxinae. The three other genera are Gnathodentex Bleeker 1873 (one species), Monotaxis Anonymous [Bennett], 1830 (two species), and Wattsia Chan & Chilvers, 1974 (one species) (Carpenter and Allen, Reference Carpenter and Allen1989; Chen et al., Reference Chen, Ho and Borsa2016, Reference Chen, Miki and Borsa2017; Chen and Borsa, Reference Chen and Borsa2020; Fricke et al., Reference Fricke, Eschmeyer and van der Laan2022).
Fishes in the genus Gymnocranius are characterized by having an ovate, laterally compressed body, generally overall silvery in colour with a pattern of five to eight narrow, transverse dark bars along the body, particularly noticeable in juveniles and pre-adults; the eye is relatively large; the profile of the head in front of the eye is slightly to markedly convex, and the snout slope relatively steep (Carpenter and Allen, Reference Carpenter and Allen1989). Largely due to the great similarity in shape and colour among the species, morphology-based species recognition has long been challenging (Sato, Reference Sato, Uyeno, Arai, Taniuchi and Matsuura1986). Gymnocranius has thus been considered one of the taxonomically most problematic tropical marine fish groups, and useful information such as life history and catch statistics have remained scant (Sato, Reference Sato, Uyeno, Arai, Taniuchi and Matsuura1986; Carpenter and Allen, Reference Carpenter and Allen1989). An updated morphological key for 12 species including one undescribed species in this genus has been proposed by Chen et al. (Reference Chen, Miki and Borsa2017). Several species including G. frenatus Bleeker, 1873, G. grandoculis (Valenciennes, 1830), G. microdon (Bleeker, 1851), G. oblongus Borsa et al., 2010, G. satoi Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013, and G. superciliosus Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013 usually possess blue ornamentation on the snout and cheeks.
Large-eye seabream specimens morphologically similar to G. superciliosus, a species so far only reported from the tropical western Pacific (Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013; Miki et al., Reference Miki, Sakurai and Iwatsuki2014; Chen and Borsa, Reference Chen and Borsa2020), were noticed by the first author (WJC) at the Victoria fish market on Mahé Island in the Seychelles during a visit in April 2016. Other Gymnocranius species for sale at the Victoria fish market included G. elongatus Senta, 1973 and G. microdon. Like G. superciliosus, this large-eye seabream had a relatively elongate body with reddish pectoral, dorsal, and caudal fins and featured a distinctive eyebrow-like pattern above the eye, characters that together excluded all other large-eye seabreams then known (Carpenter and Allen, Reference Carpenter and Allen1989; Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013; Chen et al., Reference Chen, Miki and Borsa2017). Upon closer examination, however, the general body shape and proportions of these specimens seemed marginally different from G. superciliosus. Three specimens were purchased, measured, and photographed and one of them was preserved as voucher. A series of underwater pictures of large-eye seabreams from Reunion Island (Western Indian Ocean) taken by C. Schilling (GSM diving club, Saint-Gilles-les-Bains) and posted online in June 2018 were seen by the last author (PB) and found to include the same fish. An additional specimen was collected in April 2019 by PB and WJC at fish stalls in Le Port, Reunion Island and kept as voucher. In June 2019, S. Paijo (BarCore project, Indonesia) looked for large-eye seabreams in the Peunayong and Lampulo fish markets in Banda Aceh (Andaman Sea) and sampled two G. superciliosus-like individuals (including one voucher specimen and a fin clip of another individual). Meanwhile, one of the co-authors, based in Mahé (JN) surveyed the Victoria fish market between 2013 and 2018, where he regularly sighted specimens of a species similar to G. superciliosus, of which photographs were retained and later examined in this study.
A recent multiple-gene study exploring the phylogeny and species diversity of the Monotaxinae (Chen and Borsa, Reference Chen and Borsa2020) has confirmed the validity of the 11 Gymnocranius species described so far and that of one undescribed species (Gymnocranius sp. D; Chen et al., Reference Chen, Miki and Borsa2017); two additional species from the western Indian Ocean provisionally named Gymnocranius sp. F and Gymnocranius sp. G were reported, which formed a monophyletic group with G. obesus Chen et al., Reference Chen, Miki and Borsa2017. The molecular analyses of the three large-eye seabream specimens sampled by WJC from the Seychelles in April 2016 designated them as of yet another distinct species, provisionally referred to as ‘Gymnocranius sp. H’ (Chen and Borsa, Reference Chen and Borsa2020). Gymnocranius sp. H was placed within a separate clade mostly comprising Gymnocranius spp. with blue ornamentation on snout and cheek, and was resolved as the sister group of G. superciliosus (Chen and Borsa, Reference Chen and Borsa2020).
The objectives of the present paper are (1) to morphologically and genetically compare the Gymnocranius sp. H specimens from the Indian Ocean with the G. superciliosus material available to us; (2) to formally describe Gymnocranius sp. H as a new species.
Materials and methods
Material examined
Gymnocranius sp. H specimens to be chosen as type material of the new species included two specimens from the Western Indian Ocean preserved at the ichthyological collections of the National Taiwan University Museums in Taipei under collection nos. NTUM 12894 and NTUM 16756 and two specimens from Western Indonesia deposited in fish collections in Australia (NTM S.10771-004 and CSIRO H7306-13). Additional specimens from the Seychelles and Aceh (Table 1) were photographed and measured at their sampling site; fin clips were retained. Comparative material of G. superciliosus used for morphological examination included six specimens deposited at Muséum National d'Histoire Naturelle, Paris [MNHN ICOS-00715 (New Caledonia, voucher), MNHN 2009–0010 (New Caledonia, holotype), MNHN 2009-0011 to −0013 (New Caledonia and Fiji, paratypes)], six specimens deposited in the Miyazaki University Fisheries Science collections [MUFS 42021, MUFS 42771, MUFS 42776, MUFS 42785, MUFS 43073, and MUFS 43076 (Okinawa, vouchers)], as well as NTUM10729 (Marshall Islands, voucher), and United States National Museum, Washington DC USNM 443295 (South China Sea off Luzon, voucher). Another specimen from Aceh (Andaman Sea) was deposited in the Museum Zoologicum Bogoriense in Cibinong, Indonesia under collection no. MZB.26789. Preserved tissue samples of other individuals of the two species were also used for molecular examination. Three new sequences of the mitochondrial cytochrome c oxidase I (COI) gene and two new sequences of the nuclear early growth response protein 2B (EGR2B) gene were generated in this study according to the laboratory protocols and procedures of Ward et al. (Reference Ward, Zemlak, Innes, Last and Hebert2005) and Chen and Borsa (Reference Chen and Borsa2020). Homologous COI gene sequences of Gymnocranius sp. H and G. superciliosus were searched in the GenBank (https://www.ncbi.nlm.nih.gov/genbank/; Benson et al., Reference Benson, Cavanaugh, Clark, Karsch-Mizrachi, Lipman, Ostell and Sayers2017) and BOLD (www.boldsystems.org; Ratnasingham and Hebert, Reference Ratnasingham and Hebert2007) repositories. When such sequences were identified from the databases, we recorded the distribution information registered with the samples to further inform the distribution map of the two species (Figure 1). The list of all specimens used for the present work, together with sampling details and GenBank and BOLD accession numbers are provided in Table 1.
Table 1. List of specimens and sequences of G. indicus sp. nov. and G. superciliosus examined for the present work, with sampling location, sampling date when available (NA not acknowledged), and GenBank or BOLD accession nos. for COI gene sequences, and GenBank accession nos. for EGR2B gene sequences
Asterisk indicates holotype. CSIRO Australian National Fish Collection, Hobart; MNHN, Muséum national d'histoire naturelle, Paris; MUFS, University of Miyazaki, Fisheries Science, Japan; MZB Museum Zoologicum Bogoriense, Cibinong; NTM Museum and Art Gallery of the Northern Territory, Darwin; NTUM, National Taiwan University Museums, Taipei; USNM National Museum of Natural History, Washington DC.
a Same specimen as the one represented as ‘Gymnocranius microdon’ in O'Neall et al. (Reference O'Neill, Rahmat, Proctor and White2023) (J.J. Pogonoski, pers. comm.).
b Same specimen as the one represented as ‘Gymnocranius sp.’ in Gloerfelt-Tarp and Kailola (Reference Gloerfelt-Tarp and Kailola2022) (B.C. Russell, pers. comm.).
c Individual morphologically identified as G. superciliosus but carrying G. indicus sp. nov. mitochondria and EGR2B alleles.
Figure 1. Map of the Indo-West Pacific showing the occurrence points of the material examined for the present study including material reported previously (Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013; Miki et al., Reference Miki, Sakurai and Iwatsuki2014; Chen and Borsa, Reference Chen and Borsa2020), including: Gymnocranius indicus sp. nov. (solid circles) and G. superciliosus (open circles). Asterisks (*) indicate type localities; numbers indicate sample sizes per locality; light-blue spots represent coral reefs. See Table 1 for additional details.
Measurements
The methods of measuring and counting and the terminology followed Carpenter and Allen (Reference Carpenter and Allen1989). Standard length (SL), largest body depth (BD), body depth at origin of first dorsal fin (BDd), body depth at origin of first anal fin (BDa), and pre-dorsal (PD), pre-pelvic (PP), and pre-anal (PA) lengths were measured to the nearest millimetre. Head length (HL), snout length [SN; measured without lips as in Carpenter and Allen (Reference Carpenter and Allen1989)], eye diameter (ED), inter-orbital width (IOW), and median ray of caudal fin length (MRC) were measured to the nearest half-millimetre using a vernier caliper. Box-plot distributions were calculated under R v. 4.2.2 (R Core Team, Reference R Core Team2020). Colour patterns were recorded from the photographs of two live and 12 freshly caught specimens.
Data analysis
Nucleotide sequences were aligned manually with Se–Al v. 2.0 (Rambaut, Reference Rambaut1996). The software PAUP* (Swofford, Reference Swofford2002) was used to compute p-distance, and to visualize the diagnostic characters at the COI gene and EGR2B gene loci within and between Gymnocranius indicus sp. nov. and G. superciliosus. We chose the uncorrected p-distance metric over other nucleotide distances because of its universality, its smaller variance, and its adequacy when nucleotide divergence is less than 10% (Kumar et al., Reference Kumar, Tamura and Nei1993).
Principal component analysis (PCA; Pearson, Reference Pearson1901) was run on a matrix of 29 Gymnocranius sp. H / G. superciliosus individuals (15 from the Indian Ocean, 14 from the Pacific Ocean), defined by seven morphological variables (BDd, ED, HL, PA, PD, SN, and depth of caudal peduncle) using FactoMineR v. 2.7 (Lê et al., Reference Lê, Josse and Husson2008) under R v. 4.2.2 (R Core Team, Reference R Core Team2020). All measurements to be included in the PCA dataset were made on photographs (Borsa et al., Reference Borsa, Béarez, Paijo and Chen2013; Miki et al., Reference Miki, Sakurai and Iwatsuki2014; Figure 2; Supplementary Plates S2–S4) so as to have all specimens measured in a standardized way, to enable meaningful comparisons. Morphological variables were expressed as percentages of standard length and these ratios were subjected to arcsine-square root transformation (Sokal and Rohlf, Reference Sokal and Rohlf1995). Clusters of individuals were defined by the agglomerative hierarchical clustering (Sneath and Sokal, Reference Sneath and Sokal1973) algorithm implemented in FactoMineR (Husson et al., Reference Husson, Josse and Pagès2010).
Figure 2. Gymnocranius indicus sp. nov. (A) Photograph of holotype, no. NTUM 16756 from Reunion Island (NTUM /W.-J. Chen). (B) Photograph of paratype, no. NTUM 12894 from Victoria, Seychelles (NTUM /W.-J. Chen). (C) Photograph of individual sighted underwater at ca. 18 m depth at Pierre-du-Préfet off Saint-Gilles, Reunion Island (C. Schilling, 31 March 2022).
Results
Morphological measurements and meristic counts on vouchered Gymnocranius sp. H specimens are provided in Table 2, together with homologous data for the type material of G. superciliosus. Gymnocranius sp. H essentially differed from G. superciliosus by its larger eye. Eye diameter (ED) was about equal to the IOW (ratio of ED to IOW = 0.97–1.21) and always larger than MRC, with ratio of ED to MRC = 1.09–1.29, while eye diameter of G. superciliosus was smaller than both IOW and MRC. Other body proportions did not clearly differ between Gymnocranius sp. H (Table 2) and G. superciliosus: the ratio of SL to BD was 2.61–2.69 in Gymnocranius sp. H, vs 2.51–2.75 in G. superciliosus; the ratio of SL to HL was 2.79–3.10 in Gymnocranius sp. H, vs 3.02–3.40 in G. superciliosus. Also, the profile of the head in front of the eye was generally more prominent and the slope of the snout, steeper in Gymnocranius sp. H (Figure 2) than in G. superciliosus. No blue ornamentation on the snout and cheeks was visible in the Gymnocranius sp. H material examined (Figure 2; supplementary Plates S2, S3).
Table 2. Measurements on the type material of Gymnocranius indicus sp. nov. and comparison with type material of G. superciliosus
CSIRO Australian National Fish Collection, Hobart; MNHN, Muséum national d'histoire naturelle, Paris; NTM Museum and Art Gallery of the Northern Territory, Darwin; NTUM, National Taiwan University Museums, Taipei.
a From Borsa et al. (Reference Borsa, Béarez, Paijo and Chen2013), except BD and SL.
b Measured as in Carpenter and Allen (Reference Carpenter and Allen1989).
Two disjunct clusters of individuals were observed with PCA, and confirmed by hierarchical clustering (Figure 3). Principal component 1 represented over 45% of the total variance. One cluster included 12 of the 13 individuals sampled from the Indian Ocean; the other cluster consisted of all 14 individuals sampled from the tropical western Pacific, including the type material of G. superciliosus. The exception was individual MZB.26789 from Aceh, which clustered with G. superciliosus (thus morphologically identified as this species according to PCA) while possessing Gymnocranius sp. H mitochondria (Table 3) and nuclear EGR2B alleles (Table 4). Comparisons based on the arcsine-square root transformed foregoing measurements in Gymnocranius sp. H (N = 12) and G. superciliosus (N = 15) are presented as box-plot pairs in supplementary Plate S1. Among the most influential variables in the PCA was the arcsine-square root transformed ratio of ED to standard length, which was highly significantly larger in Gymnocranius sp. H than in G. superciliosus (supplementary Plate S1), confirming the quasi-diagnosticity of eye diameter to distinguish the two species. Other highly discriminant variables were the arcsine-square root transformed ratios of HL, PA, PD, and SN to standard length (Figure 3; supplementary Plate S1). Multivariate morphological analysis thus formally confirmed the morphological distinctness of the two species.
Figure 3. Principal component analysis (PCA) based on seven morphological measurements as defined in text. (A) Principal plane of PCA; envelopes delimit the two main clusters identified by hierarchical clustering; specimen numbers as in Table 1, first column. Asterisks designate holotypes. (B) Correlation circle; BDd body depth at origin of dorsal fin; CP depth of caudal peduncle; ED, eye diameter; HL, head length; PA, pre-anal length; PD, pre-dorsal length; SN, snout length. (C) Boxplots representing the distribution of standard length in each cluster.
Table 3. Matrix of pairwise genetic distances (p-distance) at the COI locus, deduced from the nucleotide sequences sampled in 11 individuals morphologically identified as Gymnocranius indicus sp. nov. (GspH) or G. superciliosus (Gsup)
Sample numbers as in Table 1, first column.
a Individual morphologically identified as G. superciliosus but carrying G. indicus-like mitochondria.
Table 4. Matrix of pairwise genetic distances (p-distance) at the EGR2B locus, deduced from exon 1 to intron region nucleotide sequences (450 bp) sampled in ten individuals morphologically identified as Gymnocranius indicus sp. nov. (GspH) or G. superciliosus (Gsup)
Sample numbers as in Table 1, first column.
a Individual morphologically identified as G. superciliosus but genetically characterized as G. indicus sp. nov.
Genetic comparison at intra- and inter-specific levels of Gymnocranius sp. H and G. superciliosus was made at the COI and EGR2B loci (Tables 3 & 4). Gymnocranius sp. H was differentiated from its sibling G. superciliosus by 3.6–4.6% (mean = 3.9%) and 0.67–0.74% (mean = 0.67%) nucleotide sequence divergence at the COI and EGR2B loci, respectively. Genetic diversity at the COI locus appeared to be slightly higher within G. superciliosus (mean pairwise p-distance = 0.003) than in Gymnocranius sp. H (mean pairwise p-distance = 0.001; Table 3). Nucleotide sequences at locus COI nos. MT888958 and MT888959 in GenBank and FOAM404-10 in BOLD, all three labelled G. microdon, were characteristic of Gymnocranius sp. H.
Discussion
Morphological characters that might be useful for distinguishing species in other families of ray-finned fishes are remarkably uniform in Gymnnocranius spp. (Carpenter and Allen, Reference Carpenter and Allen1989). As a matter of fact, the diagnosis of most species in the genus Gymnocranius involves body proportions, body and fin colours, and caudal fin shape (Sato, Reference Sato, Uyeno, Arai, Taniuchi and Matsuura1986; Carpenter and Allen, Reference Carpenter and Allen1989; Chen et al., Reference Chen, Miki and Borsa2017). Carpenter and Allen (Reference Carpenter and Allen1989) emphasized the importance of colour patterns as diagnostic characters for identification of species in Monotaxinae.
In a phylogenetic analysis based on four gene markers, we have previously referred to the new species described here as Gymnocranius sp. H, the Indian-Ocean sibling of G. superciliosus from the tropical western Pacific (Chen and Borsa, Reference Chen and Borsa2020). Gymnocranius sp. H fell under G. superciliosus according to the identification key of Chen et al. (Reference Chen, Miki and Borsa2017: their table 3), except the blue dots on the cheek which were apparently absent. Gymnocranius sp. H has remained largely overlooked in the ichthyological literature. There is no mention of any Gymnocranius that match this species by external morphology and colour patterns in fish guides from the Western Indian Ocean or from the East Indian Ocean (Smith and Smith, Reference Smith and Smith1963; Kyushin, Reference Kyushin1977; Randall, Reference Randall1992, Reference Randall1995; Allen and Erdmann, Reference Allen and Erdmann2012). An exception we found is the recently updated edition of Gloerfelt-Tarp and Kailola's compendium of trawled fishes of the northwestern Australia – southern Indonesia region. This guide includes a picture of a 195-mm long individual of this species (labelled ‘Gymnocranius sp.’) captured off Tanah Bala Island off West Sumatra (Gloerfelt-Tarp and Kailola, Reference Gloerfelt-Tarp and Kailola2022). A photograph of this species was published (as ‘G. griseus’) in the IFREMer online identification sheets of marine species (Evano, Reference Evano2016). Another photograph was published (as ‘G. microdon’) in the fishIDER identification sheets for the commercial fishes of Indonesia (Proctor et al., Reference Proctor, White and O'Neill2018; O'Neill et al., Reference O'Neill, Rahmat, Proctor and White2023). This species was again reported (as G. cf. superciliosus) in the ‘Sous les mers’ online identification sheets for scuba divers at Reunion Island (Schilling, Reference Schilling2018). Another underwater picture of an individual of this species, erroneously labelled ‘Gymnocranius microdon’, from off Utende, Tanzania in July 2021 has recently been posted in the iNaturalist community forum (https://forum.inaturalist.org/; page consulted 15 October 2023).
That an Andaman-Sea G. superciliosus individual (as identified by morphometrics) possesses Gymnocranius sp. H mitochondria and alleles at the nuclear marker scored in the present study suggests introgression of the genome of one of the two species by genetic material of the other species in this region of the Indo-West Pacific. Further investigation using a larger sample size of individuals and genome-wide locus sampling is warranted, as it will help address the phylogeographic structure and history of the two species in the Andaman Sea and adjacent regions.
Large-eye sea breams of the genus Gymnocranius are fishes of high commercial interest throughout the tropical Indo-West Pacific (Coleman, Reference Coleman1981; Carpenter and Allen, Reference Carpenter and Allen1989). The discovery of a new Gymnocranius species is therefore significant. Gymnocranius sp. H was just one of four potentially undescribed Gymnocranius species included in Chen and Borsa's (Reference Chen and Borsa2020) phylogeny of the Monotaxinae. The three other species have been provisionally referred to as Gymnocranius sp. D, Gymnocranius sp. F and Gymnocranius sp. G. The description of the Indo-West Pacific distributed Gymnocranius sp. D is currently in preparation. The two other species, both possibly endemic to the Western Indian Ocean, remain to be investigated in more depth.
Taxonomy
Gymnocranius indicus sp. nov
ZooBank no. urn:lsid:zoobank.org:act:3CD85B8F-19F8-4EBA-B8DE-CB4AE4C818FC. Proposed vernacular names: Indian-Ocean eyebrowed large-eye seabream (English); empereur à sourcils de l'océan Indien (French); ikan kakap putih alis Samudera Hindia (Indonesian); capitaine sourcil (Mascarene Islands); kapten blan grolizye (Seychelles). See Tables 2–4; Figures 2 & 3; supplementary Plates S2, S3. Previously reported as Gymnocranius sp. (Gloerfelt-Tarp and Kailola, Reference Gloerfelt-Tarp and Kailola2022), Gymnocranius sp. H (Chen and Borsa, Reference Chen and Borsa2020) and Gymnocranius cf. superciliosus (Schilling, Reference Schilling2018). Misidentifications: Gymnocranius microdon (non Bleeker, 1851) (O'Neill et al., Reference O'Neill, Rahmat, Proctor and White2023); Gymnocranius griseus (non Temminck and Schlegel, 1843) (Evano, Reference Evano2016; upper photograph).
Type material
Holotype: NTUM 16756 (Table 2; Figure 2A), 281 mm SL, Reunion Island, Western Indian Ocean, 3 April 2019. The specimen was captured using baited bottom handline by ca. 20 m depth on Reunion Island's western shore. Paratypes: NTUM 12894 (Table 2; Figure 2B), 291 mm SL, Victoria, Seychelles, 16 April 2016; NTM S.10771-004 (Gloerfelt-Tarp and Kailola, Reference Gloerfelt-Tarp and Kailola2022: 210; Table 2), off West Sumatra, Eastern Indian Ocean; CSIRO H7306-13 (Table 2), Pelabuhan Ratu, West Java, Eastern Indian Ocean.
Nucleotide sequences of the type material of Gymnocranius indicus sp. nov. deposited in GenBank have the following registration numbers. Holotype: OQ517983 (COI gene); paratype from Seychelles: MT607095 (COI gene), MT606910 (cytochrome b gene), MT607230 (early growth response protein 2B gene), and MT607051 (rhodopsin gene). The partial nucleotide sequence of the COI gene of the paratype from West Java has been deposited in BOLD (as ‘G. microdon’) under accession no. FOAM404-10.
Other specimens examined
Gymnocranius indicus sp. nov.: specimen nos. 1–3, 5–13, 15 and 16 listed in Table 1. Gymnocranius superciliosus: specimen nos. 17–32 listed in Table 1, including type material of this species.
Description
The description of the new species is based on 16 specimens, using discriminant morphometric and genetic approaches (Tables 2–4; Figure 3). Gymnocranius indicus sp. nov., as all species in the genus Gymnocranius, is characterized by a continuous dorsal fin with ten spines and ten to 11 soft rays (ten spines and ten soft rays in Gymnocranius indicus sp. nov.); anal fin with three spines and nine to ten soft rays (three spines and ten soft rays in Gymnocranius indicus sp. nov.); pectoral fin rays 14; rear part of cheek with three to five transverse scale rows (four in Gymnocranius indicus sp. nov.); remainder of cheek, preorbital, snout, and interorbital region scaleless; inner surface of pectoral fin base scaleless; body laterally compressed and ovate; profile of the head in front of the eye convex, slope of the snout relatively steep; adult specimens often develop a bony ridge on the nape and a bony shelf above the anterior part of the eye; mouth small, posterior part of the jaw usually anterior to the level of the anterior edge of the eye; each jaw with two or three slender canines at the front, conical teeth on the sides, and a range of numerous villiform teeth behind the front teeth; eye relatively large, a pair of close-set, round nasal openings on each side of the snout in front of the eyes, usually a thin flap of skin on the rear edge of the anterior opening.
The specimens of the new species possess the following combination of characters: body elongated, 2.7–2.9 times in standard length (2.7 in holotype); pored scales on lateral line 47–49; forehead prominent; lower edge of eye slightly above a line from tip of snout to middle of caudal fin; presence of distinctive and conspicuous dark patch above eye; eyes protruding and large, the eye diameter (ED) reaching 37–40% of head length (39% in holotype), usually close to or slightly smaller than inter-orbital width (IOW) with an ED/IOW ratio of 0.91–1.21 (1.03 in holotype); caudal fin moderately forked, its lobes slightly convex inside, the median rays nearly shorter than ED with an ED/MRC ratio of 1.09–1.38 (1.29 in holotype); body colour silvery, sides without visible transversal dark bars in adults, blackish spot at basis of scales forming longitudinal rows on back; pectoral, dorsal, and caudal fins reddish to red; front row of six pre-dorsal scales, arranged in a ‘V’ pointed backwards (Figure 2; supplementary Plates S2, S3).
Nucleotide sequences at the COI locus that can be used as DNA barcodes to identify Gymnocranius indicus sp. nov. are available from the GenBank sequence repository under accession nos. MT607095, MT888958, MT888959, OQ517983, and OQ517984; and from the BOLD repository under record no. FOAM404-10. Similarly, nucleotide sequences at the EGR2B locus are available from GenBank under accession nos. MT607230, MT607231, OR785721, and OR785722.
Diagnosis
The specimens of Gymnocranius indicus sp. nov. possess the same combination of morphological traits that distinguishes G. superciliosus from all other known Gymnocranius spp. in Chen et al.'s (Reference Chen, Miki and Borsa2017) key to species of this genus: general body shape oblong to elongate, SL/BDd ratio 2.7–3.1, forehead prominent; caudal fin large, moderately forked with a subtle middle notch, its lobes slightly convex inside; flanks silvery; anal fin, caudal fin, dorsal fin, and pectoral fin reddish to red; distinctive dark patch above eye. Protruding, large eyes (eye diameter reaching 37–40% of head length) is the main distinctive trait to diagnose Gymnocranius indicus sp. nov. from G. superciliosus. The PCA graph (Figure 3) demonstrated complete separation of two clusters along PC1, one including the type material of G. superciliosus, the other one including type material of the new species. The size of the specimens examined in this analysis ranged from 189 to 341 mm. Therefore, the combination of quasi-diagnostic characters ED, HL, PA, PD, and SN measured in the present study was diagnostic all over this range of size.
Along the COI gene, the following apomorphic sites have unique nucleotides shared by all seven Gymnocranius indicus sp. nov. sequences examined so far, that distinguish it from all other species in the genus Gymnocranius: nos. 360 (T vs C), 405 (T vs C), 453 (T vs C) and 477 (C vs A or G), where the numbering of nucleotide sites starts from the first nucleotide of the gene. These genetic characters remain diagnostic for individuals of any size, from egg to adult, except for possible hybrids or introgressed individuals which would, for example, possess the mitochondrial type of the female ancestor. They are also diagnostic for body parts, including fillets, offal, scales, blood, mucus, etc. DNA sequences are indeed powerful characters for identifying individuals to species, valid for any age and a wide array of preservation conditions, potentially including stuffed and alcohol-preserved specimens.
Habitat and distribution
The type locality is Reunion Island in the Western Indian Ocean. Gymnocranius indicus sp. nov. was recorded (as either ‘G. superciliosus' or ‘Gymnocranius sp. H') from Reunion Island and the Seychelles in the western Indian Ocean (Schilling, Reference Schilling2018; Chen and Borsa, Reference Chen and Borsa2020). Based on these and additional records from the present study, its distribution range extends from the western Indian Ocean (Tanzania, Seychelles, Reunion, and Mauritius) to the Andaman Sea (Aceh) and to the eastern Indian Ocean off West Sumatra and West Java (Figure 1; Table 1). Underwater sightings of solitary individuals have been made by C. Schilling at ca. 18 m to ca. 25 m depth on the outer slope of the reef off Saint-Gilles on the western coast of Reunion Island, above a bottom constituted by sand, lava boulders, and sparse coral colonies (supplementary Plate S3A). An additional underwater sighting has been reported from off Tanzania.
Etymology
Epithet indicus is the latin translation of ‘Indian’, a reference to the geographic distribution of this species, as inferred from the available records.
Remarks
As reported above, Gymnocranius indicus sp. nov. has previously been confused with G. microdon and G. griseus. The inferred multigene phylogeny revealed that Gymnocranius indicus sp. nov. is closely related to neither G. microdon or G. griseus (Chen and Borsa, Reference Chen and Borsa2020). Morphologically, Gymnocranius indicus sp. nov. differs from G. microdon in general body shape (oblong in Gymnocranius indicus sp. nov. vs oblong to elongate in G. microdon), in snout shape (more protruding and steeper in G. microdon than in Gymnocranius indicus sp. nov.) and in the shape of caudal fin (moderately forked with parenthesis-shaped edge in Gymnocranius indicus sp. nov. vs forked and concave in G. microdon). Gymnocranius indicus sp. nov. differs from G. griseus by the following suite of characters: (i) body shape oblong to elongate (high-bodied to oblong in G. griseus; Chen et al., Reference Chen, Miki and Borsa2017); (ii) silvery, mostly immaculate flanks (vs several transversal dark bars in G. griseus; Chen et al., Reference Chen, Miki and Borsa2017); (iii) number of pre-dorsal scales in front row (six in Gymnocranius indicus sp. nov. vs nine to 11 in G. griseus; Chen et al., Reference Chen, Miki and Borsa2017). Moreover, based on the genetically validated records available, the geographic distributions of Gymnocranius indicus sp. nov. and G. griseus are disjunct, as G. griseus is confined to the southern Japanese archipelago and Taiwan, where G. indicus sp. nov. is absent (Chen et al., Reference Chen, Miki and Borsa2017; Chen and Borsa, Reference Chen and Borsa2020).
Notice
The present article in portable document (.pdf) format is a published work in the sense of the International Code of Zoological Nomenclature [International Commission on Zoological Nomenclature (ICZN), Reference International Commission on Zoological Nomenclature2012] or ICZN Code. Hence, the new name contained in the present article is effectively published under the ICZN Code. The present article and the new nomenclatural act it contains have been registered in the online registration system for the ICZN, ZooBank (http://zoobank.org/). The online version of this article is available from the JMBA journal website (https://www.cambridge.org/core/journals/journal-of-the-marine-biological-association-of-the-united-kingdom/), and from the HAL (https://cnrs.hal.science/; Daphy and Ha–Duong, Reference Daphy and Ha–Duong2010) repository.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315424000274.
Acknowledgements
We are especially grateful to C. Schilling for providing us with excellent underwater photographs from Reunion Island and to S. Paijo (Proyek BarCore, Lampuuk) for help in sampling specimens from Aceh. We also thank three anonymous reviewers for helpful suggestions and B.C. Russell for pointing out the latest edition of the Trawled fishes of southern Indonesia and northwestern Australia book. An MZB collection number for the voucher from Aceh was kindly allocated by H. Haryono (BRIN, Cibinong). D. Pitassy (Smithsonian Institution, Washington DC) granted us access to the high-resolution photograph of G. superciliosus specimen no. USNM 443295 from the South China Sea. J.J. Pogonoski and B.C. Russell assisted in obtaining measurements on the two paratypes from Western Indonesia located in Australian museums. P. Béarez (CNRS, Paris), A. Collet (IRD, Nouméa), J.-L. Justine (IRD, Nouméa) and D. Ponton (IRD, Nouméa) helped us collect and analyse important biological material in the early stages of our project on the systematics of the Monotaxinae, to which the present article is our eighth contribution.
Data availability
All nucleotide sequences were deposited in the GenBank-Nucleotide open-access repository (https://www.ncbi.nlm.nih.gov/nucleotide/). Pictures of all specimens examined are provided in this article or in its supplementary material. Other data not included in this article, or in its supplementary material, or in open-access repositories will be made available upon reasonable request.
Author contributions
Designed the study: PB, WJC; contributed reagents or materials or analysis tools: PB, WJC, RM, JN; analysed and interpreted the data: PB, WJC; wrote the paper: PB, WJC; read, edited, and approved the final draft: PB, WJC, RM, JN.
Financial support
This work was supported in part by the Ministry of Science and Technology, Taiwan (MOST 104-2611-M-002-002-MY3, -111-2611-M-002-025 -, and -112-2611-M-002-025- to WJC) and in part by the French government through Fonds Pacifique, INEE-TAAF Iles Eparses research consortium, and IRD grants to PB.
Competing interest
The authors declare they have no competing financial interests or personal relationships that potentially could have influenced the work reported in this paper.
Ethical standards
All specimens examined were obtained from fish markets. No live specimen was manipulated.
