Atopocottus Bolin, 1936

Matsunuma, Mizuki, Sato, Mao & Kai, Yoshiaki, 2017, Redescription of Atopocottus tribranchius (Cottidae) from Japan with comments on the generic diagnosis and distribution, Species Diversity 22, pp. 87-97 : 88-93

publication ID

https://doi.org/ 10.12782/sd.22_87

persistent identifier

https://treatment.plazi.org/id/BD2087A7-063D-D877-FE8B-F8B9BA42FA49

treatment provided by

Felipe

scientific name

Atopocottus Bolin, 1936
status

 

Atopocottus Bolin, 1936 View in CoL

[Japanese name: Kawari-anahaze-zoku]

Diagnosis. Dorsal-fin rays VI–VIII-10–12; anal-fin rays 6–9; pectoral-fin rays 17–19; pelvic-fin rays I, 2; dorsal fins broadly separated into spinous and soft-rayed portions ( Fig. 1 View Fig ); short row of 3–5 deeply embedded lateral-line scales with 4–6 pores restricted to anterior portion of body, no other scales on head and body; lateral-line scales with a simple tube, non-serrated posterior margin ( Fig. 2A View Fig ); cephalic sensory pores on head moderately developed ( Fig. 3A, B View Fig ), supraorbital canal with 4 pores (SO1–2 and 4–5), lacking SO3; occipital canal with a single and four paired pores, including posterior medial pore (OCP), lateral anterior central pores (OCLA), lateral posterior central pores (OCLP), lateral anterior pores (OLA), and lateral posterior pores (OLP), lacking anterior medial pore (OCA) and minute supplementary pores; terminal mandibular pores paired but not united as single large pore; postpelvic process somewhat rounded, slightly elongate and well separated from each other ( Fig. 2B View Fig ); head and nasal spines absent, preopercular spines present; uppermost preopercular spine with 1–6 dorsal branches, becoming smaller toward spine tip, some branches with two spinous points in large specimens ( Fig. 4 View Fig ); skin flaps (including supraocular tentacles and orbital cirri) absent on head and body.

Comparisons with Antipodocottus and Stlengis . Although Bolin (1952) and Fricke and Brunken (1983, 1984) considered Antipodocottus to be related to Stlengis, Dewitt (1969) suggested that the former was in fact more similar to Atopocottus . Nelson’s (1985, 1990) comparisons of Antipodocottus with Stlengis and Atopocottus (based on osteological and external characters) indicated that Antipodocottus was more closely related to Atopocottus , with Stlengis appearing to be closer to Icelinus , although he was unable to conduct a detailed osteological examination of Atopocottus . Recently, Knope (2013) provided a molecular phylogenetic hypothesis for Cottidae based on mitochondrial and nuclear genes, and indicated a robust sister relationship between At. tribranchius and Stlengis misakia ( Jordan and Starks, 1904), and Icelinus was recovered in a monophyletic group with other several northern Pacific genera. Antipodocottus was not included in his analysis. Accordingly, this study focused on the examination of external and selected osteological characters of Atopocottus (described for the first time) so as to compare that genus with Antipodocottus and Stlengis .

Comparisons of selected characters among the three genera are shown in Table 1 [revised from Nelson (1985: table 1)]. The generic placement of S. mesembrinus follows Nelson (1990) and Nelson and Carpenter (1999). No differences occurred in general skull components, including the five infraorbitals [ Nelson (1985, 1990) included the lacrimal in the infraorbital series]. Similarly, the caudal skeleton was consistent among the three genera, each having three epurals and a plate-like hypural complex with a central split ( Nelson 1985: figs 1, 2, 5; Figs. 2C View Fig , 3C, D View Fig ). However, generic differences were apparent in the pelvic girdle, the pelvis of At. tribranchius being relatively broad and short, its length subequal to its width ( Fig. 2B View Fig ), compared with the relatively narrow and long (length about 1.5 times its width) pelvis in Antipodocottus and Stlengis (see Nelson 1985: fig. 4). The postpelvic process of At. tribranchius is somewhat round- ed and widely separated each other, but more pointed and closed in Antipodocottus and Stlengis . The nasal bone of At. tribranchius is smooth, lacking a spine as in Antipodocottus (except An. elegans -variably present or absent), whereas it has a distinct spine in S. misakia ( Nelson 1985, 1990; this study). The uppermost preopercular spine shape of At. tribranchius is unique among the three genera, comprising 1–6 dorsal branches, their length and width becoming shorter and narrower, respectively, toward the spine tip (some branches with two spinous points in large specimens) ( Fig. 4 View Fig ). In contrast, Stlengis possesses 5 (rarely 4) mostly simple branches on the uppermost preopercular spine, all of subequal length and width ( Fig. 5A View Fig ), as in Antipodocottus ( Bolin 1952; DeWitt 1969; Fricke and Brunken 1983; Nelson 1985, 1990).

The cephalic sensory systems of At. tribranchius and Antipodocottus are similar, sharing poorly developed canals with fewer pores, compared with Stlengis . Atopocottus tribranchius is characterized by four supraorbital pores and a few supplementary pores on the inner side of the frontal region ( Fig. 3A View Fig ), as opposed to 11 supraorbital pores and numerous minute supplementary pores in that region in Stlengis ( Fig. 6 View Fig ). The number and position of pores on the frontal of At. tribranchius ( Fig. 3C View Fig ) is closely similar to those of Antipodocottus (see Nelson 1985: fig. 1A; Nelson 1990: fig. 2A). Moreover, At. tribranchius has the occipital canal with a posterior medial pore, lateral anterior and posterior central pores, and lateral anterior and posterior pores on the parietal region. The occipital canal of Stlengis , on the other hand, has many anterior and posterior branches and supplementary pores. The terminal mandibular pores of At. tribranchius are paired [not united as a single large pore as in some species of Antipodocottus ( An. elegans and An. galatheae ) and Stlengis ( Table 1)].

The lateral-line scale morphology of At. tribranchius and Antipodocottus clearly differs from Stlengis , as point- ed out by Nelson (1985). Atopocottus tribranchius and Antipodocottus possess non-overlapping lateral-line scales with a simple tube (see Nelson 1985: fig. 3A; Fig. 2A View Fig ). In contrast, Stlengis is characterized by overlapping lateralline scales, with curved dorsal and ventral wings forming dorsal and ventral pores along the lateral-line canal and the strongly serrated posterior margin (see Nelson 1985: fig. 3B). Within Stlengis , the lateral-line scales almost reach the caudal-fin base in Stlengis distoechus Bolin, 1936 and S. misakia or the anterior portion of the caudal peduncle in Stlengis osensis Jordan and Starks, 1904. The lateral line is incomplete in At. tribranchius , with a row of 3–5 scales restricted to the anterodorsal portion of the body. In Antipodocottus , the lateral-line scale condition is variable, incomplete with a row of 12–14 scales restricted to the anterodorsal portion of the body in An. galatheae , but complete and reaching the caudal-fin base in An. megalops ( Table 1). With the exception of the lateral-line scales, the body of At. tribranchius is naked. This condition is similar to Antipodocottus , although An. elegans and An. mesembrinus possesses a lateral row of non-overlapping, plate-like scales below the second dorsalfin base ( Fricke and Brunken 1983, 1984). In contrast, S. distoechus and S. osensis are characterized by dorsal and/or ventral rows of well-developed overlapping ctenoid scales; S. distoechus has a dorsal row of scales along the dorsal fin bases to the dorsal portion of the caudal-fin base ( Fig. 7A View Fig ), and S. osensis , an additional ventral scale row above the anal fin to the ventral portion of the caudal-fin base ( Fig. 7C View Fig ). The remaining congener, S. misakia , possesses only lateralline scales ( Fig. 7B View Fig ).

Atopocottus tribranchius further differs from Stlengis in lacking cirri on the orbit surface, compared with numerous minute cirri in the latter ( Fig. 7B View Fig ). In contrast, Antipodocottus usually has a moderately long supraocular tentacle [see Bolin (1952: fig. 1), DeWitt (1969: fig. 1) and Fricke and Bolin (1984: fig. 1); Table 1].

Atopocottus is further distinguished from Antipodocottus by fewer lateral-line scales [3–5 (vs. 5–33 in the latter)], the postpelvic process being somewhat rounded and widely separated each other (vs. pointed and slightly separated), the uppermost preopercular spine with 1–6 dorsal branches, their length and width becoming shorter and narrower toward the spine tip, some with two spinous points (vs. 2–5 simple, similarly-sized dorsal branches), and absence of a supraocular tentacle [vs. present (unknown in An. mesembrinus )]. Atopocottus may also be diagnosed by the lower numbers of dorsal-fin soft rays (10–12) and analfin rays (6–9), compared with Antipodocottus (12–15 and 10–12, respectively) ( Table 1). However, further detailed investigations of the internal and external characters of Atopocottus and Antipodocottus are necessary to determine their monophyly and phylogenetic positions.

Comparisons among the three genera suggested that Atopocottus is more similar to Antipodocottus rather than Stlengis as in the previous studies [e.g., Nelson (1985, 1990)]. Fricke and Brunken (1984) suggested a difference in depth distribution between Antipodocottus and Stlengis . Among the three genera, At. tribranchius inhabits the shallowest depths (118–247 m), compared with Antipodocottus (ca. 100–735 m depth) and Stlengis (ca. 130–460 m depth) ( Fricke and Brunken 1984; Stewart 2015; Ohashi et al. 2017; this study).

Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF