The general physiognomy, meristic values, squamation and fin shapes of Anthias xanthomaculatus are in general agreement with the genus Odontanthias. In particular, the presence of filamentous segmented rays in the anterior part of the soft dorsal fin is unusual among anthiadine genera, but typical of Odontanthias. Filamentous extension of anterior segmented dorsal-fin rays also occurs in Sacura Jordan & Richardson (1910), Meganthias Randall & Heemstra (2006) and certain species of Anthias Bloch (1792) and Plectranthias Bleeker (1873). However, A. xanthomaculatus differs from Meganthias species in having a large serration on the preopercular angle and in lacking accessory scales on the body. Classification in Anthias is more difficult to reject on the basis of character evidence, as relevant characters are either based on larvae or adult osteology (Baldwin 1990), and are unknown for A. xanthomaculatus. However, given that Anthias is currently restricted to Atlantic and eastern Pacific species (Anderson & Heemstra 2012, Anderson et al. 2017), it seems unlikely that A. xanthomaculatus is correctly classified in that genus.
Among Plectranthias species, filamentous extension of anterior segmented dorsal-fin rays is restricted to just a handful of species, including P. azumanus (Jordan & Richardson 1910), P. exsul Heemstra & Anderson (1983), P. foresti Fourmanoir (1977), P. kelloggi (Jordan & Evermann 1903), P. maculicauda (Regan 1914), P. maugei Randall (1980), P. parini Anderson & Randall (1991) and P. sagamiensis (Katayama 1964). Anthias xanthomaculatus differs from these species and all other members of the genus in various characters, such as having higher numbers of gill rakers (13 + 28 = 41 versus 4–9 + 9–20 = 13–29) and a deeply emarginate caudal fin (versus rounded or truncate to emarginate).
......continued on the next page
Justification for distinction of Sacura from Odontanthias is unclear. In their revision of Sacura, Heemstra & Randall (1979) did not provide characters to distinguish the two genera, and simply referred to Katayama (1959, 1960) for comparisons between Sacura and other genera represented in Japan. Motomura et al. (2017) suggested that the two genera may be distinguished by the presence or absence of teeth “below” (i.e., posterior to) the main vomerine tooth patch. This character was also used as justification for assignment of the central Atlantic Odontanthias cauoh Carvalho-Filho, Macena & Nunes (2016) to the genus (see also Anderson et al. 2017), though that species lacks the typical filamentous extension of anterior dorsal-fin rays. Median posterior extension of the vomerine tooth patch also occurs in species of the New World genus Pronotogrammus Gill (1863). However, median posterior extension of the vomerine tooth patch does not characterise all species in Odontanthias, as it does not occur in O. grahami Randall & Heemstra (see Randall & Heemstra 2006: fig. 1H). We therefore consider the generic distinction of Sacura and Odontanthias unresolved. Regardless, Anthias xanthomaculatus differs from all known species of Sacura in having higher numbers of tubed lateral-line scales (37, discounting the damaged left side count of 33, versus 26–34). Sacura boulengeri Heemstra (1973) is the only species in the genus with X,14 dorsal-fin rays (versus 15–18 in the remaining species), but it differs from A. xanthomaculatus in having fewer tubed lateral-line scales (28–31), and 2 (versus 3) rows of scales between middle dorsal-fin spines and lateral line.
One additional character supports placement of A. xanthomaculatus in either Odontanthias or Sacura: presence of two closely spaced supraneurals. Depending on species, anthiadines have one to three supraneurals, with three being relatively plesiomorphic. In most species with two supraneurals, the two bones are well separated, such that they clearly align with interneural spaces: the first is in the preneural space and the second is in the first interneural space (e.g., Katayama & Amaoka 1986: figs 5C, E and G; Anderson & Heemstra 2012: fig. 2B). In A. xanthomaculatus and species of Odontanthias (including O. cauoh) and Sacura, the two supraneurals are more closely spaced and it is difficult to allocate the second supraneural to either the preneural or first interneural spaces (Figures 2–3; see also Katayama 1959: fig. 27I; Heemstra & Randall 1979: fig. 1).
Notwithstanding the current lack of a clear diagnosis for the genus, we here assign A. xanthomaculatus to Odontanthias, ending its previous classification in Pseudanthias. Because it does not agree with any nominal Odonthanthias described prior to 1979, we further recognise it as a valid species. Of species assigned to Odontanthias, O. xanthomaculatus most closely approaches O. grahami in meristic and morphometric details (see Table 1). The latter species is known only on the basis of the 93.0 mm SL holotype (AMS I.32142-001) from off northern New South Wales, Australia, and three non-type specimens (QM I.21166, QM I.38666, QM I.38967) collected off south-east Queensland. Given the proximity of the Australian localities to New Caledonia, we srongly suspect that the two nominal species are conspecific. The specimens share X,14 dorsal fin rays, similar numbers of lateral-line scales (36–40 in O. grahami versus 37 in O. xanthomaculatus), similar vomerine dentition (at least in lacking a posterior median extension), and remarkably similar live coloration (Figure 4). Differences in the degree of development of head spines, number of spines on the opercle (3 in O. grahami versus 4 or 5), relative length of the third dorsal-fin spine (third spine elongate in O. grahami versus not elongate), and certain live coloration features may be attributable to differences in the size of the specimens (71.8–93.0 versus 42.0 mm SL). In particular, the relatively large spines and serrations on the opercular bones and extra opercle spines are typical of juveniles of the genus (e.g., see Randall & Heemstra 2006: fig. 8). However, pending the discovery and study of additional specimens of appropriate sizes, we refrain from placing O. grahami in the synonymy of O. xanthomaculatus.