Tenuicephalus squamilabrus, Schwarzhans & Møller, 2021

Schwarzhans, Werner W. & Møller, Peter R., 2021, Revision of the ‘ dragon-head’ cusk eels of the genus Porogadus (Teleostei: Ophidiidae), with description of eight new species and one new genus, Zootaxa 5029 (1), pp. 1-96 : 81-92

Tenuicephalus squamilabrus n. sp.

Figs. 2 , 42–43 , 48 , 51 , Tab. 1–7

Material examined ( 11 specimens): Holotype: AMS I36456-007, 175 mm SL, 13°21’N 124°12’E, 1037–1100 m; paratypes: CAS 50160, 127– 140 mm SL (2 specimens), R / V Anton Bruun, Cruise 8, 21°18’S 36°18’E, 1510–1600 m, 2 October 1964; MNHN uncat., 138–167 mm SL (3 specimens), Sta. CH 128, off Madagascar, 18°05’S 42°53’E, 1930 m, collected by A. Crosnier, 16 January 1975; BSKU 82353-56 , 106– 173 mm SL (4 specimens), 01°59’S 157°12’E, 1610 m R / V Hakuho-maru, 3 m beam trawl, 04 January 1968; BSKU 98880, 174 mm SL, 01°59’S 157°12’E, 1610 m, R / V Hakuho-maru, 3 m beam trawl, 4 January 1968.

Diagnosis. Precaudal vertebrae 15–16, 2–4 loin vertebrae (last precaudal vertebrae without ribs); long gill rakers on first gill arch 17–21; pectoral-fin rays 16–18; HL:HD 1.37–1.55; maximal HD:HD through center of eye 1.75–1.90; weak head spines on ethmoidal and interorbital; opercular spine sharp, extruding; opercular flap small; scales on occiput, cheeks, opercle and maxilla; vomer narrow, naked or with few teeth or one row of teeth; palatines narrow with 2–3 rows of teeth; basibranchial tooth patch small, narrow with 2 rows of teeth; otolith small, 5.0–6.5 % in HL, with single colliculum; OL:OH = 1.14–1.20; OL:TCL = 1.94–2.30.

Description. Meristics: precaudal vertebrae 15–16, last 2–4 vertebrae without ribs; pectoral-fin rays 16 (16– 18); D/V = 5 (5–6); D/A = 23 (22–23); V/A = 15 (15–17); long gill rakers on lower gill arch 18 (17–21). Gill rakers in holotype on lower first gill arch with six short rakers, followed by a series of 18 long rakers. The lower ten of those intercepted by single plate shaped rakers. Upper gill arch with three short rakers intercepted by five plate shaped rakers.

Morphometrics: in % of SL: HL 14.7–16.2; maximal HD 9.5–11.7; HD through center of eye 5.0–6.4; bony interorbital width 2.3–3.7; snout length 4.9–5.5; upper jaw length 10.0–11.1; predorsal 16.5–17.4; preanal 28.9– 30.8; prepelvic 12.2–13.3; prepectoral 15.8–17.7; pectoral length not measurable. Relations: HL:HD = 1.37–1.55; maximal HD:HD through center of eye 1.75–1.90; HL to snout length 2.64–3.28; preanal to predorsal 1.69–1.84; predorsal to prepectoral 0.95–1.08.

Overall slender fish with long tapering tail, but stout, short head and moderately long snout. Maximal size of fishes investigated 175 mm SL (holotype). Head with distinctly concave dorsal profile, with few weak spines such as ethmoidal and interorbital; opercle with small but sharp, extruding spine. Eye moderately small located in strongly asymmetric orbit. Maxilla extending far beyond eye, strongly widened posteriorly; supramaxilla very narrow. Infra-/postorbital and mandibular-preopercular pores very wide. Head with many scales (preserved only in holotype) on cheeks and opercle, few scales on occiput, and many small scales on rear part of maxilla and very small scales on supramaxilla. Opercular flap moderately small. Lateral line not discernable.

Dentition. All teeth tiny and cone-shaped. Vomer naked or with a very narrow dentition patch with 1–2 rows of few teeth anteriorly, naked posteriorly; palatines with a narrow dentition patch with 2–3 rows of teeth. Premaxilla tooth patches not fused anteriorly; 5 rows of teeth in middle part. Dentary tooth patches fused anteriorly; ca. 5 rows anteriorly and 2 rows posteriorly. Median basibranchial tooth patch small and narrow with 2 rows of teeth.

Otolith morphology (n = 4). Size up to 1.35 mm in length ( CAS 83066) (holotype 1.35 mm); OL in % HL = 5.0–6.5; OL:OH = 1.14–1.20; OH:OT = 2.00–2.23. Otolith nearly round in shape with all rims regularly curved and smooth and without angles. Inner face flat, with short sulcus positioned near center of inner face; OL: TCL = 1.94–2.30. Sulcus with shallow, undivided, uniform, oval, moderately wide colliculum. Dorsal field wide, without distinct depression; ventral field smooth. Outer face smooth, convex, with anteriorly shifted umbo .

Coloration. Live coloration not known. Color of preserved dark brown or uniformly black.

Discussion. Tenuicephalus squamilabrus is characterized by extensive head squamation including the maxilla and supramaxilla, which however is only reasonably preserved in the holotype, and the small otolith (OL in % HL of 5.0–6.5), but otoliths are preserved only in about one-third of the specimens. In the absence of these two prime diagnostic features, the distinctly concave dorsal head profile and the resulting high ratio of the maximal HD to the HD through the center of the eye of 1.75–1.90 (vs 1.35–1.65, rarely 1.70) is the most distinctive character. For further differentiation see above discussion to T. melampeplus and T. multitrabs .

Distribution. The distribution of Tenuicephalus squamilabrus overlaps with that of T. multitrabs in the western Pacific and both species have been caught in close vicinity in the Lagonoy Gulf of the Philippines. Else, T. squamilabrus had been caught in the Madagascar Channel. Tenuicephalus squamilabrus does not seem to venture far away from the continental slopes and for instance has not been caught along the Ninety-East Ridge like T. multitrabs .

Etymology. From squama (Latin) = scale and labrum (Latin) = lip, referring to the many small scales on maxilla and supramaxilla.

Ecology and biogeography

Porogadus and Tenuicephalus are widely distributed beneath the oligotrophic tropical and subtropical seas of the world and in some regions reach into the temperate zone. Their live-style is benthopelagic below the oxygen minimum zone and to the best of our knowledge all catches were made by deep-sea bottom trawls. OTSB hauls (Otter Trawl Semi-Balloon) appear to have yielded the best results for catching larger benthopelagic fish at great depth. For instance many of the Porogadus specimens represented in the BMNH collection from off Cabo Verde Islands and Mauretania below 3000 m depth have been caught with this method (e.g., Merrett in Partridge 2000). While Porogadus specimens are generally few in individual hauls they can locally become quite abundant as exemplified by a Galathea II trawl in the Bay of Bengal that yielded 45 specimens of P. melanocephalus , as yet the highest record of a Porogadus species in any individual trawl.

The majority of Porogadus species and all species of Tenuicephalus occurs bathyal at depths between c. 1000 and 3500 m but there are specimens of primarily bathyal species that have occasionally been caught at depth greater than 3500 m, i.e., P. atripectus , P. dracocephalus , P. miles and P. promelas ( Figs. 49–51 ). In addition there appears to be some depth segregation in the diversity within certain groups indicated by species having been caught regularly or exclusively at great depths. These are Porogadus gracilis and P. melanocephalus below 3000 m; P. caboverdensis , P. mendax and P. turgidus below 4000 m; and P. abyssalis below 5000 m. In addition there is a live photograph of a putative Porogadus specimen taken by a ROV of the NOAA ship Okeanos Explorer at 5856.8 m on the abyssal plain of the central Pacific near Phoenix Islands ( Fig. 5 ) which would represent the deepest record for the genus. These species are amongst the deepest dwelling teleost fishes and they recruit only from the Porogadus miles and Porogadus gracilis groups, the latter being entirely restricted to below 3000 m water depth ( Fig. 49 ). Interestingly, the three abyssal species of the Porogadus miles group so far have only been found in a relatively confined area on the abyssal plains off NW-Africa. This is an unusual diversity center but would be in line with observations made by Merrett & Marshall (1981) that the species diversity off NW-Africa was considerably higher than off NE-America (for bathyal fish communities), but dominated by smaller species, which they related to the higher primary productivity related to the Mauretanian upwelling system even at considerable distance from shore. The reason, however, for the unusual diversity of Porogadus in the abyssal zone off NW-Africa remains elusive. Watling et al. (2013) do not show a specific abyssal bioprovince in the NE-Atlantic. We are uncertain whether this feature represents a genuine diversification event in the region, which we cannot readily explain or if it is a product of sampling bias in which case the same or other abyssal Porogadus species might be expected in other ocean basins. The abyssal regions of the world oceans are still severely undersampled, even though the use of deep-sea remotely operated vehicles (ROV) for scientific purposes have greatly increased our knowledge in recent years ( Chave & Mundy 1994; Linley et al. 2017; Mundy et al. 2018; Bell et al. 2012, 2016, 2107; Raineault et al. 2018, 2019, 2020) and abyssal biogeography of fishes therefore has remained largely elusive (Priede et al. 2020). The abyssal region discussed here off NW-Africa is located south of the warm, high salinity Mediterranean outflow zone that reaches down to about 2000 m ( Potter & Lozier 2004; Filipelli 2014) and far off from the coastal upwelling system in the Mauretanian Sea ( Pelegrí et al. 2017). Seamounts in the vicinity have recently become known for large deep-sea sponge populations ( Ramiro-Sánchez et al. 2019). Two of the three deep-water species of the Porogadus gracilis group occur in the Indian Ocean ( P. gracilis and P. melanocephalus ) and one of them in the southwestern Pacific Oceans ( P. gracilis ). This is a remarkable under-representation compared to the deep Atlantic and it is not entirely clear, if and how sampling bias may play a role in this discrepancy. The third abyssal species of the Porogadus gracilis group, P. abyssalis lives on the central Atlantic abyssal plains and represents the deepest dwelling species of the entire genus.

The Porogadus gracilis group is noticeable for a reduced ossification, reduced otolith size and low level head armature. We hypothesize that these features reflect an adaptation of the fishes for living at great (abyssal) depth generally below the carbonate compensation depth (CCD). In contrast to this group the three abyssal species of the Porogadus miles group, i.e. P. caboverdensis , P. mendax and P. turgidus do not show a reduced ossification or reduced otolith size or even reduced head armature. This in our view would indicate a less advanced adaptation to the abyssal zone compared to the species of the Porogadus gracilis group, which in turn might be a result from them having settled at great depth in more recent times.

Not surprisingly, the deep-water species mentioned above which live on the abyssal plains can occur far out from the continental breaks, but some of the bathyal species can also occur at great distance from the nearest continent ( Figs. 49–51 ). Porogadus miles holds the record with over 2500 km away from the nearest continental mass on the Emperor Seamount Chain ( Fig. 49 ). Tenuicephalus multitrabs also occurs off the continental breaks by 1000 to 2000 km on the Ninety East Ridge ( Fig. 51 ). Other species have been found regionally restricted and on or close to continental slopes, which is a pattern found in all groups of both genera ( Figs. 49–51 ).

A recent study by Agusti et al. (2015) revealed the ubiquitous presence of healthy photosynthetic cells, dominated by diatoms, down to 4,000 m, which together with fecal pellets probably represent the basis for the deep-water food chain. Deep-sea fishes are the top predators in this short trophic sequence. Carter (1984) studied the feeding strategy and functional morphology of several deep-sea ophidiids at great detail including Porogadus miles , P. catena and T. silus . The species of both genera are remarkable for their tiny, uniform, cone-shaped teeth on dentary, premaxilla and vomer, which in the genus Porogadus are usually distributed on many rows with the outermost one(s) on the dentary sometimes bent outwards, whereas the number of teeth rows are reduced in Tenuicephalus . The many developed rakers on the first gill arch support filter function, sometimes additionally enhanced by their bladed structure (e.g. in P. guentheri ). The mouth is terminal or slightly inferior with a wide gape. In Carter’s (1984) assessment of the stomach content of the investigated ophidiids, substantial differences were observed in composition and diversity of the food as well as the content of sediment in the intestines the latter being indicative for infaunal feeding. The two Porogadus and the Tenuicephalus species studied showed no or negligible amounts of sediment inside and a high dependency on a prime food source. T. silus stomach content was mostly filled by calanoid copepods plus mysadaceans while in P. catena the diet contained primarily calanoid copepods plus substantial amounts of cumaceans. In P. miles the prime diet consisted of gammarideans. Carter (1984: 160) studied a total of twelve deep-sea ophidiids for their feeding strategy and grouped P. catena , P. miles , T. silus and Penopus macdonaldi in a group morphologically characterized by, among others: “a slender and highly attenuate body, narrow and short pectoral fins with low aspect ratios, moderate to large gape, slightly reduced upper jaw protrusibility, moderate to large eye, greater number of gill rakers on the anterior arch with high filtering capacity of the branchial sieve and similar configuration and shape of bony elements of the feeding apparatus” (Carter 1984: 82–83). Porogadus catena and Tenuicephalus species are further characterized by a short, terminal snout and a reduction of head ossification (moderate in P. catena , very strong in Tenuicephalus spp. ). Carter (1984) concluded that these functional morphological traits combined with the selective pelagic copepod diet of P. catena and T. silus point to a “hovering, passive (not far) off-bottom foraging and facultative benthopelagic feeding strategy”. The much more robust, anteriorly depressed head and subterminal mouth of P. miles and Penopus spp. would suggest bottom or near bottom feeding (Carter 1984). Other typical features observed in Porogadus and Tenuicephalus species is the often lacking or rejuvenated tip of the whip-like tail, which probably represents an escape mechanism against larger predators. The highly deciduous nature of the body scales may serve a similar purpose. The very strongly reduced head ossification in Tenuicephalus likely supports a live-style predominantly off the sea-bottom, while the heavily armored, spinous heads of the “dragon-head” Porogadus species may represent a defensive mechanism against food competitors or larger predators that may be foraging on or near the bottom.

Porogadus and Tenuicephalus species show varied geographic distribution patterns ( Figs. 44–48 ). Some species are widely distributed beneath the tropical and subtropical oceans but there appears to be none with a circum-global distribution. Other species are rather restricted geographically. The most widespread species is P. miles known from the western and eastern Atlantic, throughout the Indian Ocean and the western Pacific as far east as the Emperor Seamounts but absent from the eastern Pacific ( Fig. 45 ), where the closely related P. longiceps is represented. A number of species are known from the Indo-West Pacific but are missing from the Atlantic or eastern Pacific, e.g., P. dracocephalus , P. gracilis , P. trichiurus , T. multitrabs and T. squamilabrus ( Figs. 44–48 ). Few species are seemingly restricted to the Atlantic, e.g., P. abyssalis and P. catena ( Figs. 46–47 ). Finally, there are several species restricted to rather small geographic areas although in some instances their actual distribution pattern might be vague because of rather few catches. The more common regionally restricted species are P. atripectus , P. longiceps and P. promelas in the eastern Pacific, P. guentheri off southern Japan, P. lacrimatus and P. solomonensis chiefly in the Solomon Sea, P. melanocephalus in the Bay of Bengal, P. caboverdensis , P. mendax and P. turgidus off NW-Africa, and T. silus in the Caribbean. The most widely distributed species ( P. miles and T. multitrabs ) inhabit bathyal terrain and not surprisingly are also the ones capable of venturing furthest away from the continental masses. The abyssal species of the Porogadus gracilis group and the three deep NE-Atlantic species P. caboverdensis , P. mendax and P. turgidus show a rather restricted distribution pattern to few abyssal plain regions but in this case we are not certain about possible sampling biases. The narrow distribution areas of P. guentheri (off southern Japan), P. melanocephalus (Bay of Bengal) and T. silus (Caribbean) are based on many specimens caught in different cruises over some time and therefore appear to be genuine.

Another interesting aspect in the distribution pattern of Porogadus and Tenuicephalus species is the occurrence of morphologically similar and putatively closely related species pairs. One such pair is composed of P. miles (Atlantic to West Pacific) and P. longiceps (East Pacific), another of P. catena (Atlantic) and P. promelas (East Pacific). Both species pairs probably resulted from allopatric speciation that occurred after the rise of the Isthmus of Panama and the separation of the adjacent deep-sea of the tropical Atlantic and tropical East Pacific. Following current consensus, the final separation of the tropical West Atlantic from the East Pacific occurred between 4.7 to 3.0 Ma during the early to middle Pliocene, but shoaling and separation at bathyal depth may have occurred as early as 13 Ma during the Middle Miocene Climate Transition (MMCT) (e.g., Burton et al. 1997; Jackson & O’Dea 2013). However, pelagic larval stages of bathyal fishes like in Porogadus may have facilitated exchange of populations for a considerable time into the gateway shoaling process. In the case of the separation of populations that may have given rise to P. miles and P. longiceps another barrier of exchange must have occurred in the Central Pacific. We speculate that the drivers for separation there may be related to the cool water regime established along the North Pacific rim latest with the initiation of the northern hemisphere glaciation that started at about 3.0–2.5 Ma ( Anderson Dahl 2009), while the wide Central Pacific deep-water gap is a much older permanent feature. Warm pulses that could have facilitated some exchange of warm-water biota along the North Pacific rim could potentially have occurred as late as 5.0–3.5 Ma that culminated in the mid-Pliocene Climate Optimum (e.g., Morley & Dworetzky 1991; Gladenkov 1994; Tsuchi 2002). The closing process of the adjacent East Pacific and Caribbean basins apparently also led to an increase of the upwelling in the East Tropical Pacific that induced an expansion of the oxygen minimum zone in the intermediate ocean layer, i.e. the modern oceanographic conditions from about 4.2 Ma ( Kamikuri et al. 2009). This is supported by the first occurrence of endemic East Pacific mesopelagic fishes in deep-sea sediments on the Pacific coast of Panama from about 2.5 Ma ( Schwarzhans & Aguilera 2013). While low oxygen zones will act as an effective barrier to most deep-water taxa, it has been observed that some ophidiids can tolerate remarkable low concentrations (Gallo et al. 2018). Whether this is the case for species of Porogadus (or Tenuicephalus ) is unknown. In conclusion, a series of plate tectonic events and oceanographic and climatic changes during the time interval from about 4.5 to about 2.5 Ma is thought to be responsible for an effective isolation of the tropical East Pacific deep-sea fish fauna both from the West Pacific as well as the West Atlantic. Consequently, we assume that phylogenetic divergence of the assumed species pairs mentioned above took place during this time interval.

Following our synonymization of P. nudus with P. miles , and P. subarmatus with P. catena we do not recognize an allopatric separation of species across the Atlantic. However, we postulate P. miles and P. mendax to form a species pair in the subtropical NE-Atlantic that has resulted from sympatric depth segregation. The interrelationships of two other abyssal species in the Cabo Verde and Canaries basins, P. caboverdensis and P. turgidus , within the Porogadus miles group are less clear. The isolating mechanism for the NE-Atlantic abyssal species is elusive (see above) and hence we are unable to identify potential causes and timing. Aspects that could potentially be linked to speciation events in the deep-sea of the region are the origination and development of the warm, high salinity, high-density Mediterranean outflow zone in intermediate depths of the adjacent NE-Atlantic ( Potter & Lozier 2004, Filipelli 2014) and the coastal upwelling system in the Mauretanian Sea ( Pelegrí et al. 2017). The Mediterranean outflow zone was established sometime after the reconnection of the Mediterranean with the NE-Atlantic after the Messinian Salinity Crisis. Filipelli (2014) concluded that from 4.5 to 3.5 Ma the flow was weak and assumed its full extend only after 3.2 Ma. The timing of the formation of the coastal upwelling system in the Mauretanian Sea is still incompletely known. Diester-Haass & Schrader (1979) found indications of coastal upwelling as early as early Miocene in the DSDP well 369A off the Canary Islands, which however diminished towards late Miocene. This is consistent with observations made by Schwarzhans (1993) who described early Pliocene sciaenid otoliths from northern Morocco of taxa that today are only found southwards of the Mauretanian upwelling zone in tropical West Africa. These findings indicate a weakening or intermission of the Mauretanian upwelling which would allow marine tropical West African fishes to expand northwards at the time. In any case, given the close morphological similarity of P. miles and P. mendax (see above) we would assume a young dichotomy, possibly younger than 5.0 to 4.5 Ma.

Other potential species pairs are the abyssal P. abyssalis in the Central Atlantic and P. gracilis in the Indo-West Pacific and the bathyal P. trichiurus on the continental rise in the Indo-West Pacific and P. guentheri off southern Japan. Timing and mechanism that led to the separation of the species in these two pairs is unknown to us, however in the case of P. guentheri it coincides with an abundance of endemic fish species offshore southern Japan and Taiwan in shallow and deep water (e.g., Masuda et al. 1984, Schwarzhans & Møller 2008, Schwarzhans & Prokofiev 2017). Porogadus guentheri and P. trichiurus are unique amongst Porogadus species in their specialization of the developed gill rakers on the first gill arch likely enhancing their filter function. In P. guentheri the elongate gill rakers are blade-like extended while in P. trichiurus they are very narrowly placed without many intermittent short plates. For the remainder of the Porogadus species and all of the Tenuicephalus species close interrelationships that could be interpreted as ‘species pairs’ are not obvious.

Figures

FIGURE 2. Panel of comparative head drawings: from top to bottom: Porogadus (P. miles), Tenuicephalus (T. squamilabrus n. sp.) and selected putative related genera; Alcockia (A. rostrata), Bathyonus (B. laticeps) and Penopus (P. microphthalmus).

FIGURE 5. Live figures from ROV dives. 5A–B, putative long-nosed Porogadus specimen; photo taken on 23.03.2017 during Okeanos EX 1703, dive 17 at 4°22’31”S, 172°59’52”W at 5856.8 m. 5C, putative short-nosed Tenuicephalus specimen; photo taken on 20.09.2015 during Okeanos EX 1504, leg 4, dive 7 at 15°12’N, 168°03’W, maximal depth of dive 1950.7 m. Photographs courtesy of the NOAA Office of Ocean Exploration and Research.

FIGURE 42. Tenuicephalus squamilabrus n. sp., holotype, AMS I.36456-007, 175 mm SL, 13°21’N, 124°12’E, 1037–1100 m.

FIGURE 43. Tenuicephalus squamilabrus n. sp.; 43A–B, holotype, AMS I.36456-007, 13°21’N, 175 mm SL, 124°12’E, 1037–1100 m, head drawings; 43C–D, CAS 50160, paratype, 140 mm SL, 21°18’S, 36°18’E, 1510-1600, otolith (C inner face, D ventral view); 43E–F, holotype, AMS I.36456-007, 175 mm SL, 13°21’N, 124°12’E, 1037–1100 m, otolith (E inner face, F ventral view).

FIGURE 44. Geographic distribution of the species of the Porogadus miles group.

FIGURE 45. Geographic distribution of the species of the Porogadus trichiurus group.

FIGURE 46. Geographic distribution of the species of the Porogadus gracilis group.

FIGURE 47. Geographic distribution of the species of the Porogadus catena group.

FIGURE 48. Geographic distribution of the species of the genus Tenuicephalus.

FIGURE 49. Depth versus distance from shelf plots for specimens of the Porogadus miles group (upper panel) and the Porogadus gracilis group (lower panel).

FIGURE 50. Depth versus distance from shelf plots for specimens of the Porogadus trichiurus group (upper panel) and the Porogadus catena group (lower panel).

FIGURE 51. Depth versus distance from shelf plots for specimens of the genus Tenuicephalus.