Comatula Lamarck 1816

Comatula Lamarck, 1816

Table 1, Figs. 3–4

Type species. Comatula solaris Lamarck, 1816.

Other included taxa (5). Asterias pectinata Linnaeus, 1758; Comatula cratera HL Clark, 1916a; Comatula micraster AH Clark, 1909b; Comatula rotalaria Lamarck, 1816; Comatula tenuicirra AH Clark, 1912a.

Diagnosis. Mouth central in fully developed individuals; arms usually ten (rarely to 17 in C. pectinata, and to 20 and rarely 27 in C. rotalaria); centrodorsal pentagonal to circular, thin, sometimes reduced and lying within radial circlet (Fig. 3 C, E, K, P–R); cirri absent or present (few to complete ring; sometimes only in interradial angles of centrodorsal); all brachitaxes of two ossicles united by syzygy; syzygies at 1+2, 3+4 on undivided arms (Fig. 3 N, P); distal intersyzygial interval irregular, 2–11 (Fig. 3 F); combs of triangular teeth confluent with edge of pinnule away from arm, tapering distally, either to a point or with smaller distal teeth (Fig. 3 I, L, O) (modified from Hoggett & Rowe (1986)).

Distribution. Indo-western Pacific from tropical and warm temperate Australia (Abrolhos Islands, WA, to Sandon Bluffs, NSW), west to the Andaman Islands and Gulf of Martaban, Burma, east to New Caledonia, and north to Taiwan and Okinawa (AH Clark 1931; Chen et al. 1988; Kogo 1998; Meyer 1986; Rowe & Gates 1995). A single specimen attributed to C. solaris and supposedly collected in the Society Islands (French Polynesia) has been lost (AH Clark 1931). No other specimen of Comatulidae has ever been collected this far east. Depth range: 0– 110 m.

Molecular results. Specimens of Comatula were recovered as a pair of sister clades, one comprised of 20- armed Comatula rotalaria and the other of 10-armed C. solaris, C. pectinata, and a Comatula sp. from Singapore (Fig. 4) (see Remarks below).

Specimens identified in the field as C. pectinata and C. solaris and, separately, C. pectinata and C. purpurea (from Australia) differ by less than 1% difference in COI (Fig. 4). The type specimens of C. solaris (MNHN EcCs1009, 1010) and C. purpurea (ZMB) were collected along the western or northwestern coast of Australia (Freycinet 1815, AH Clark 1931), whereas Linnaeus (1758) gave the type locality for his Asterias pectinata as “Indian seas” (“in mare Indico”) (ZMLU), which could be anywhere in the Indian Ocean or East Indies. Rowe & Gates (1995) previously synonymized C. purpurea under C. pectinata, which we consider valid here. The one specimen identified as C. pectinata recovered within a clade with three attributed to C. solaris is here considered C. solaris.

Remarks. Comatula differs from all other comatulids in having a syzygy in the IBr series and at 1+2 and 3+4 on all arms (Fig. 3 H, N, P). Although some of the six current nominal species may be synonyms, as diagnostic characters are often overlapping meristic, size-related, and qualitative, at least a few are likely valid. Both C. rotalaria and C. cratera have ten or fewer cirri when small but lose them with growth. In larger specimens with similar arm lengths, C. cratera has no more than ten arms, whereas C. rotalaria typically has 20 (up to 27) and exhibits an apparently unique method of increasing arm number: the second brachial in a 10-armed specimen develops an arm instead of exterior P1 (AH Clark 1915:131; 1931). Even in large specimens, these exterior arms may remain shorter than the interior arms (Messing et al. 2006). Small, poorly known C. micraster also has 10 arms and no cirri and might be attributable to either of the preceding two. These three taxa have non-overlapping, but adjacent, ranges from the Gulf of Martaban to Sandon Bluffs, NSW, Australia, warranting further molecular investigation to delineate their status. Likewise, descriptions of C. pectinata and C. solaris widely overlap, and specimens identified as each at Lizard Island had almost identical COI sequences. In this case, those attributed to C. pectinata were collected from reef habitat, whereas C. solaris was found under rubble in a non-reef sediment environment — suggesting that they cannot be distinguished by habitat as advocated by Messing et al. (2006). Comatula purpurea and C. tenuicirra are distinguished from the two preceding by cirri restricted to the interradial angles of the centrodorsal, with C. tenuicirra set apart only by its proportionally more elongated cirrals.

Comatula sp. from Singapore has longest cirrals with L/W ratio ~2.0, as in C. tenuicirra, and cirri restricted to one side of the centrodorsal, as in one of the C. tenuicirra type specimens (NBC V.ECH.C 2102.2); however, it has long oral combs composed of numerous tall teeth, whereas the type specimens of C. tenuicirra have short combs of up to ~10 separated teeth and taper to a distal point. Comatula rotalaria, also from Singapore, keys out as C. pectinata; it has a complete ring of XV cirri and combs of 7–12 separated teeth that taper distally.