Warenophis andersi, Stöhr, 2025

Stöhr, Sabine, 2025, Two new species of brittle stars (Echinodermata: Ophiuroidea), living epizoic on a stalked crinoid, European Journal of Taxonomy 1022, pp. 176-201 : 190-194

Warenophis andersi sp. nov.

urn:lsid:zoobank.org:act:

Figs 3–4

Diagnosis

The disc is round, slightly domed, dorsally covered by large, irregular scales, with triangular, contiguous radial shields, that are ⅓ as long as the disc diameter. Granules and spines are completely absent from the disc. The dorsal arm plates are fan-shaped and contiguous. The ventral disc is covered by smaller round scales. At the apex of the jaw there are two elongated tooth papillae, along each jaw edge sit three small, round lateral oral papillae. The second tentacle pore is superficial outside the mouth slit. The up to four arm spines are flattened, wide, triangular, serrated, the dorsalmost spine is the largest. There is a single small, oval tentacle scale. The spine articulations on the lateral arm plate have a tongue-like dorsal extension.

Etymology

The specific epithet honours Anders Warén, who collected the species and is a mentor and long-time friend of the author.

Type material

Holotype

TONGA – N Ha’apai group • 19°03.7′ S, 174°18.98′ E; depth 523–806 m; 14 Jun. 2000; Anders Warén leg.; Bordau stn DW1595; Warén dredge; on Neogymnocrinus richeri (SMNH-225978); on SEM stub; MNHN, MNHN-IE-2023-164 .

Paratypes

TONGA – N Ha’apai group • 2 specs; same collection data as for holotype; in ethanol, arm ossicles on SEM stub; SMNH, SMNH-Type-10005 .

Description

Holotype

DORSAL DISC. 4.8 mm dd, pentamerous. Round, domed disc, centre with irregular scales of variable size, no granules or spines. Radial shields scalene triangular, about as long as ⅓ of dd, completely contiguous. Primary rosette not distinguishable. Large rectangular interradial scale separates pairs of radial shields.

VENTRAL DISC. Covered by scales smaller than on dorsal disc. Oral shield rhombic, 1.5 times as wide as long, distal edge convex, madreporite larger, no hydropore visible. Adoral shields crescent-shaped, proximal edge slightly concave, restricted to proximal edges of oral shield. Short jaws with two fingerlike tooth papillae, teeth wide, rounded. Three minute, round granule-like lateral oral papillae. Large superficial second tentacle pore outside mouth slit, with oval, scale-like adoral shield spine at edge of adoral shield.

ARMS. Dorsal arm plates fan-shaped, as long as wide, contiguous. Lateral arm plates forming part of dorsal arm. Up to four arm spines, two dorsal spines largest, flat, triangular, proximal edge convex, distal edge straight, perpendicular to arm, imbricated, roof-like horizontal. Spines smaller on distal arm, but same flat shape throughout arm. Two ventral spines shorter, narrower, flattened. All arm spines with serrated edges. Ventral arm plates widely axe-shaped, twice as wide as long, lateral edges deeply excavated, distal edge straight, proximal edge wide flat angle, barely contiguous. Single round tentacle scale, not completely closing pore.

COLOURATION. Dark green, similar to host N. richeri , faded in alcohol.

Ossicles

LATERAL ARM PLATES. Outer side with three to four spine articular structures on elevated part, shaped as two lobes, almost vertical in position, dorsal lobe tongue-shaped widened, ventral lobe short, muscle opening larger than nerve opening. Proximal outer edge of LAP with median process (spur). Stereom finely porous, along proximal edge a finer meshed band. Vertebral articular structures on inner side as angled flat ridge with finer pores than surrounding stereom, and round low knob on ventro-proximal part. Large pores distal to ridge, flat spur median on distal part.

ARM SPINES. Articular structures as oval, flat, smooth knob with median slit and deep groove dorsodistal to knob. On largest spine, articular structures offset to distal part of lateral edge.

VERTEBRAE. Zygospondylous articulation with zygosphene as long as zygocondyles, extending beyond lower ends of zygocondyles. No dorsal keel-like structure.

Paratypes

Specimen A: 4.1 mm dd, disc damaged, but plates and scales similar to holotype. Two arms completely broken off, three at distance from disc. Arms and ventral disc like holotype. Tentacle scale smaller than in holotype. Arm pieces present in sample.

Specimen B: juvenile of 2.2 mm dd, dorsal disc with pentagonal central primary plate, slightly smaller, rounded square radial primary plates, small pentagonal proximal interradial plate, larger rectangular distal interradial plate. Radial shields scalene triangular, contiguous. Oral papillae as in other paratype and holotype. Large tentacle pores with single minute scale. Small arm pieces in sample, all but one arm broken off close to disc.

Remarks

The taxonomic affinities of Warenophis andersi gen. et sp. nov. can be narrowed down to two families, Ophiacanthidae and Ophiotomidae . These are closely related and share numerous characters. All known ophiotomid genera have disc spines or granules, but the ophiacanthid genus Ophiohamus O’Hara & Stöhr, 2006 completely lacks disc armament, thus resembling the new species ( O’Hara & Stöhr 2006). The flat, serrated arm spines of W. andersi are similar to those of Ophiotomidae , particularly Ophiocopa Lyman, 1883 and Ophiopristis Verrill, 1899 . The analysis of the lateral arm plates showed that both families have a tendency to a tongue-like extension of the dorsal lobe of the spine articular structures ( Fig. 4E, G, I, K), particularly obvious in Ophiolimna ( Fig. 4G). The sigmoidal fold that is typical for the order Ophiacanthida is weakly expressed in the new species. The LAP of W. andersi has a proximal spur, which is more similar to Ophiotomidae than to Ophiacanthidae . Ophiotomidae LAPs have less conspicuous striations than those of Ophiacanthidae or they are limited to a small area, the new species lacks striations. The wavy border along the row of spine articulations on the LAPs is missing in W. andersi , but a slight elevation is present. The vertebral articular structure on the inner surface of the LAPs has the shape of a digit one ( Numberger-Thuy & Thuy 2020) or an upside down check mark in Ophiacanthidae ( Fig. 4F, H), but has an elongated dorsal extension in some Ophiotomidae , e.g., in Ophiomitra leucorhabdota ( Fig. 4I), and in the new species it is a diagonal ridge, similar to Ophiocopa spatula ( Fig. 4C, L). This ridge could be homologous to the long part of the digit one or a shortened version of the ophiotomid ridge, with the short downwards pointing proximal part missing. The articular structures on the arm spines are similar between the new species and Ophiacantha and Ophiocopa ( Fig. 4M–P), perhaps slightly more similar to Ophiocopa . The evidence gathered from the new species is inconclusive, but Ophiacanthida is the most likely order. Ophiacanthidae is polyphyletic, and Ophiotomidae may need to be revised due to its type species appearing to be an ophiacanthid (O’Hara, unpubl. data). Family placement is proposed as incertae sedis until more data are available and the taxonomic issues with these families have been resolved.

The Bayesian phylogenetic inference places W. andersi gen. et sp. nov. on the same major branch as Ophiacanthida , but not close to Ophiacanthina O’Hara, Hugall, Thuy, Stöhr & Martynov, 2017 ( Fig. 5). The tree is more comb-shaped within the clade that holds the Ophiacanthina and its allies, compared to the tree inferred from the dataset excluding W. andersi ( Fig. 6). The addition of the new species to the dataset has caused Ophioleucida O’Hara, Hugall, Thuy, Stöhr & Martynov, 2017 to form a sister group to the rest of the clade, albeit still paraphyletic for Ophiernidae O’Hara, Stöhr, Hugall, Thuy & Martynov, 2018 , and the Ophiodermatina Ljungman, 1867 are left paraphyletic for Ophiomyxidae Ljungman, 1867 . Ophiocopa spatula is morphologically similar to Ophiocamax vitrea Lyman, 1878 and thus clusters with Ophiocamacidae O’Hara, Stöhr, Hugall, Thuy & Martynov, 2018 , leaving the Ophiotomidae paraphyletic. The addition of more species ( Ophiactis profundi Lütken & Mortensen, 1899 , two Ophiolimna , and O. spatula ) has caused small changes in the tree structure and node support values, compared to previously published versions of the dataset ( Thuy & Stöhr 2016; Stöhr 2024), but the division into major clades is the same. The dataset for W. andersi is incomplete and lacks all inner disc skeletal characters, because a lack of material did not allow the sacrifice of a specimen for dissociation of the disc ossicles. The homologies of the oral papillae could not be assessed, since the necessary ontogenetic series is not available, but the proximal papillae below the teeth were interpreted as tooth papillae and it is assumed that there should be infradental papillae and an adoral shield spine at the second tentacle scale. Incomplete data can have a negative effect on the analysis, as was also observed in the test run with only LAP characters (Supp. file 2), which resulted in a comb-shaped tree structure with broken up Amphilepidida O’Hara, Hugall, Thuy, Stöhr & Martynov, 2017 . Only the Ophiacanthida still formed a clade similar to the complete analysis, and W. andersi is confirmed in the clade that contains the Ophiacanthida , but not within that order. Molecular data may also be needed, but the type material is suspected to have been preserved in denatured ethanol originally and none of the unique specimens could be sacrificed with such uncertain expected outcome.