Trimma hamartium, Winterbottom, 2018

Winterbottom, Richard, 2018, A new species of Trimma (Pisces; Gobiidae) from the South-West Islands of Palau, western Pacific Ocean, Zootaxa 4370 (2), pp. 123-136 : 124-135

publication ID

https://doi.org/ 10.11646/zootaxa.4370.2.2

publication LSID

lsid:zoobank.org:pub:F04C3156-3AB0-4790-A139-7CE5585A97BF

DOI

https://doi.org/10.5281/zenodo.3510400

persistent identifier

https://treatment.plazi.org/id/83652B24-C908-4821-9C3B-B6E3F4BEEAF9

taxon LSID

lsid:zoobank.org:act:83652B24-C908-4821-9C3B-B6E3F4BEEAF9

treatment provided by

Plazi

scientific name

Trimma hamartium
status

sp. nov.

Trimma hamartium View in CoL new species

Mistaken Pygmygoby Figs. 1–7 View FIGURE 1 View FIGURE 2 View FIGURE3 View FIGURE 4 View FIGURE 5 View FIGURE 6 View FIGURE 7 .

Trimma preclarum View in CoL —non Winterbottom, 2006:63 View Cited Treatment ( Fiji specimens only); Winterbottom & Hoese, 2015:73 View Cited Treatment (?—Australian material); Sunobe et al., 2017 (specimens used for sex determination only, see supplement).

Material examined. The description is based primarily on the type specimens, although data from all specimens from the South-West Islands of Palau (n = 488) were recorded for the fifth pelvic-fin ray (unbranched or branched), the number of sensory papillae in row c, and the relative lengths of the second and third dorsal spines of the first dorsal fin (where undamaged).

Holotype. ROM 83314, 19.2 About ROM mm SL male, Hatohobei State, Helen Reef , outer reef on W side, N of channel but S of large wreck, 2.91223° N, 131.7443° E, huge variety of hard corals, some alcyonarians, hydroids, sponges, tunicates; steep reef slope with small caves and overhangs, some with sandy floors, 23–34 m, rotenone, 20080924, Field No. RW 08-37, R. Winterbottom et al. GoogleMaps

Paratypes. ROM 102746, 43 (9.6–18.4), collected with the holotype. ROM 102747, 19 (16.2–19.2), collected with the holotype (data taken for detailed description). ROM CS 1971, 5 (10.3–19.5), collected with the holotype (now cleared and stained).

Non-type material: ROM 83027, (25.3), Sonsoral State, off SE tip of Pulo Anna, about 10 m from surf line, 4.6525° N, 131.95438° E, Halimeda , heavy cover of hard corals (Acropora, Pocillopora , Millepora , plate corals, lettuce corals), U-shaped channel about 10 m wide at seaward side and 3 m wide at shore side, floor 17 m at outside shelving to 9 m at tip, numerous caves at base of tip, floor of channel mostly of live coral, 7–15 m, rotenone, 20080913, Field No. RW08-09, R. Winterbottom et al. ROM 83062, 20 (14.8–23.8), Hatohobei State, off northern tip of Tobi Island on E side, 3.01175° N, 131.13182° E, extensive hard coral cover, hydroids in caves, steep reef slope to 15–20 m, then vertical wall with caves and undercuts with sandy floors, 18–30 m, rotenone, 20080915, Field No. RW08-12, R. Winterbottom et al. ROM 83073, 51 (14.5–20.6), Hatohobei State, Tobi island, off southern-most tip of fringing reef, 2.99735° N, 131.12303° E, good hard corals (Acropora, Pocillopora , plate coral), hydroids in caves, sponges, almost vertical wall with several large caves with sand/rubble floors, undercuts, and a few small platforms, 16–28 m, rotenone, 20080915, Field No. RW08-13, R. Winterbottom et al. ROM 83149, 13 (13.6–18.3), Hatohobei State, Helen Reef, N side of pass into lagoon, 2.87298° N, 131.7423° E, star corals, galaxids, Acropora, cactus corals, many gorgonians, hydroids, brown encrusting algae, slope (45°) on N side of channel about middle of pass, beach rock and coarse sand, 8–20 m, rotenone, 20080917, Field No. RW08- 18, R. Winterbottom et al. ROM 83195, 4 (13.0–17.3), Hatohobei State, Helen Reef, N side of pass near seaward entrance, 2.87448° N, 131.74018° E, numerous hard corals on coral rock base, steep (70°) rocky slope flattening out to sand and rubble bottom near bottom of channel, 14–23 m, rotenone, 20080918, Field No. RW08-20, R. Winterbottom et al. ROM 83220, 20 (13.3–18.6), Hatohobei State, Helen Reef, at marker buoy on W side of 'no take' zone in middle of lagoon, 2.879722° N, 131.775° E, some live coral (lettuce, Acropora, Seriatopora ), a few soft corals, rubble ridge just W of W side of old river channel (which leads to main channel), 15–28 m, rotenone, 20080919, Field No. RW08-23, R. Winterbottom et al. ROM 83232, 3 (15.9–16.1), Hatohobei State, Helen Reef, S side of pass about 100 m E from outer reef entrance, 2.87228° N, 131.7382° E, sea fans, sea whips, some hard corals, hydroids, green algae, steep (about 90°) wall of channel, mostly limestone, with numerous very small caves, some with sand floors, 20–35 m, rotenone, 20080920, Field No. RW08-25, R. Winterbottom et al. ROM 83235, 3 (15.2–17.5), Hatohobei State, Helen Reef, S side of main channel at about half-way point, 2.87145° N, 131.75652° E, variety of hard corals (Acropora, Pocillopora , star corals), Millepora , sponges, sea-whips (antipatherians), hydroids, soft corals, rocky 'bommies' with large patches of sand (patch reef), 8–20 m, rotenone, 20080920, Field No. RW08-26, R. Winterbottom et al. ROM 83248, 22 (10.8–17.7), Hatohobei State, Helen Reef, outer reef near NE tip, 2.96687° N, 131.83512 E, large beds of Halimeda , some hard corals ( Meandrina , star corals) and hydroids, sponges, ascidians, 75° drop-off slope with small caves, overhangs, crevices, some coral rubble, 18–26 m, rotenone, 20080921, Field No. RW08-27, R. Winterbottom et al. ROM 83254, 25 (10.5–19.5), Hatohobei State, Helen Reef, E side of outer reef, about 6 km SSE of island in lagoon, 2.92278° N, 131.83118° E, mixed hard corals (Acropora, Pocillopora , star and cactus corals, lots of Halimeda ; sloping reef top (20°) over lip of drop-off slope (about 90°), some overhangs and sand-floored small caves, 4–25 m, rotenone, 20080921, Field No. RW08-28, R. Winterbottom et al. ROM 83261, 82 (10.6–16.6), Hatohobei State, Helen Reef, outer reef along SE edge, 2.82773° N, 131.78827° E, sea fans, hard corals (variety), soft corals (Nephytes), hydroids in overhangs, steep drop-off slope (ca 85°) with small caves, overhangs, some sand and rubble, 20–35 m, rotenone, 20080922, Field No. RW08-29, R. Winterbottom et al. ROM 83272, 19 (10.0–20.8), Hatohobei State, Helen Reef, south end 'embayment', vertical drop-off from reef top (no reef slope), 2.79918° N, 131.75505° E, sponges, hydroids, sea-fans, ascidians, some small heads of hard corals, vertical wall covered with sponges, small hard corals, ascidians with large sandyfloored cave (floor at 27 m, hydroids in cave), then sand/rock/patch reef, 22–34 m, rotenone, 20080922, Field No. RW08-30, R. Winterbottom et al. ROM 83304, 28 (10.2–18.5), Hatohobei State, Helen Reef, southern margin near E end, 2.79887° N, 131.75553° E, hydroids, some hard coral, sea fans, sponges, ascidians, cave (ca. 15 m wide, 4 m high) and surrounding area with sand/silt floor, floor at 27 m, 24–32 m, rotenone, 20080923, Field No. RW08- 33, R. Winterbottom et al. ROM 83324, 21 (9.7–20.1), Hatohobei State, Helen Reef, S coast near E point, 2.79933° N, 131.75477° E, lots of hard and soft corals, some sea fans, sponges, vertical wall of drop-off, to sill above cave sampled in RW08-33, 12– 30 m, rotenone, 20080924, Field No. RW08-38, R. Winterbottom et al. ROM 83335, 26 (9.8–17.8), Hatohobei State, Helen Reef, W side, about 1.3 km south of channel entrance, 2.86308° N, 131.73012° E, wide variety of hard corals, some sea-fans and sea whips, Millepora , sponges, ascidians, reef slope (ca 75°) down from reef-top, 10–30 m, rotenone, 20080925, Field No. RW08-40, R. Winterbottom et al. ROM 83338, 2 (13.0–14.8), Hatohobei State, Helen Reef, about 600 m N of large wreck on W coast; outer reef, 2.941° N, 131.77068° E, huge variety of hard corals (esp. Acropora, with several tabular forms), Pocillopora , star and cactus, Fungia , some rock, a few small caves, reef top to below lip of drop-off slope (ca. 80°), 2.5–22 m, rotenone, 20080926, Field No. RW08-42, R. Winterbottom et al. ROM 83351, 60 (10.7–19.5), Hatohobei State, Helen Reef, NW outer reef about 1.5 km S of northern large ship wreck, NW of N tip of island in lagoon, 2.99385° N, 131.79958° E, huge variety of hard corals (Acropora, Pocillopora , cactus, star, lettuce corals), a few hydroids, soft corals, some encrusting algae, steep (80–85°) drop-off slope just past lip of reef slope, a few small caves (some with sand floor, most with rubble) and crevices, 7–33 m, rotenone, 20080927, Field No. RW08-45, R. Winterbottom et al. ROM 83374, 51 (9.7–18.2), Hatohobei State, Helen Reef, S coast at W side, 2.80222° N, 131.73137° E, lots of hard corals, some sea-fans, sponges, hydroids, almost vertical wall of drop-off, numerous small, sandy-floored caves, crevices, some coral rubble, 23–36 m, rotenone, 20080928, Field No. RW08-48, R. Winterbottom et al. (see also under “Tissues”). ROM 83398, 4 (15.1–17.5), Sonsoral State, Merir Island, E coast, a little south of northern tip, 4.3228° N, 132.31712° E, Acropora, Porites, star and lettuce corals, numerous sea-fans, Halimeda , steep (ca 75°) reef slope with moderate coral cover, limestone, a few small caves and crevices with coarse sand floors, 22–35 m, rotenone, 20080929, Field No. RW08-50, R. Winterbottom et al. ROM 101386, 2 (21.0–22.3), Sonsoral State, SW corner of Sonsoral Island just before southern tip, 5.315° N, 132.22639° E, lots of soft corals, hydroids in caves, good hard corals (including Acropora and Pocillopora ), vertically walled steep gully from flat reef-top at 16 m with numerous caves and undercuts with coarse white sand/shell fragments, 16–33 m, rotenone, 20080911, Field No. RW08-03, R. Winterbottom et al. ROM 101394, 4 (12.9–21.3), Sonsoral State, off middle of SE coast of Merir Island, 4.30552° N, 132.3154° E, lots of soft and hard corals, steep crevice and vertical wall with small caves and overhangs, 20–35 m, rotenone, 20080914, Field No. RW08-10, M. Westneat et al. ROM 101395, (19.8), Sonsoral State, off SW coast of Merir Island, about 3/4 of the way S down length of island, 4.30725° N, 132.30643° E, rich coral cover, lots of plate forms of Acropora, Pocillopora , Millepora , some Halimeda , hydroids in caves, gully running from surf zone, ending in sheer drop-off at 25 m, large cave at 20 m on S wall, steep slope to gully, some undercuts, 12–30 m, rotenone, 20080914, Field No. RW08-11, R. Winterbottom et al. ROM 101396, (19.4), Sonsoral State, off middle of E coast of Pulo Anna, 4.65195° N, 131.95375° E, lots of hard and soft corals (Acropora, Pocillopora ), many Acropora in plate form, Millepora , hydroids, sponges, Halimeda , vertical wall with numerous caves and undercuts, floors with algal covered 'pebbles' and coral debris, 20–31 m, rotenone, 20080913, Field No. RW08-08, R. Winterbottom et al. Tissues (non-types): ROM T20912, T20914, T20915, T20916 and T20917 (originally preserved in 80% ethyl alcohol), collectively catalogued under ROM 84868, 5 (13.8–17.1), ex-ROM 83374.

Other material. Australia; Great Barrier Reef: See under material listed as T. preclarum in Winterbottom & Hoese (2015:73) . Fiji: Paratypes of T. preclarum , as listed in Winterbottom (2006:63). Indonesia: ROM 101382, 2 (16.0-20.7), West Papua, Mapia Atoll, off southern tip of Pulau Pegun, 50 m, 00° 47.795' N, 134° 18.198' E, MVE- 16-068, 22 Oct., 2016, M.V. Erdmann. Niue: ROM 101364, 20.1 mm SL female, Snake Gully, 19° 07.584' S, 169° 54.997' E, 15 m., clove oil, 1 Sept., 2016, MVE-16–057, M.V. Erdmann. ROM 101381, 23.0 mm SL male, collected with ROM 101364. Solomon Is.: USNM 365581, 8 (14.0–18.9), Santa Cruz Is., Ngalo I., 10°16'30" S, 166°19'00" E, steep vertical wall with rubble at base, 0–35 m, rotenone, 19 Sept., 1998, Field No. SOL 98–06, J.T. Williams et al.

Diagnosis. A species of Trimma with scales absent from the cheeks and opercle, 8–9 scales in predorsal midline, a short second dorsal spine reaching to a mean of the base of the first ray of the second dorsal fin when abducted, 17–19 pectoral-fin rays with 5–10 branched rays in the middle of the fin, an unbranched 5th pelvic-fin ray that is 51–64% the length of 4th ray, 17–19 total gill rakers, 6 papillae in row c below the eye, and a U-shaped bony interorbital with a narrow slit-like postorbital trench which ends at the posteriormost papilla in row p. When freshly collected, the new species has a reddish head with 3–4 yellow bars on the cheek, a red iris with four yellow spots, a yellow body, lighter posteriorly, with scales pockets outlined with melanophores and scale rows above and below midlateral row much lighter posterior to second dorsal fin origin (giving appearance of three yellow stripes on body); preserved specimens with fairly evenly distributed melanophores over the dorsal surface of the snout.

Description. The description is based on the holotype and 19 paratypes. Dorsal fin VI + I 9, second spine may be slightly elongated, reaching from interspace between dorsal fins to base of 3rd ray (between bases of 1 st and 2nd rays), third dorsal spine reaching posteriorly to between fin interspace and base of 1st or 2nd dorsal ray (base of spine of second dorsal fin), first ray of second dorsal fin branched (n = 10) or unbranched (n = 10), remaining fin rays branched except for posterior element of last ray, which reaches posteriorly 42– 44 –56% (48.5%) distance between its base and first exposed dorsal procurrent caudal-fin ray; anal fin I 8– 9 (8.9), first ray unbranched, last ray reaching posteriorly 39– 41 –51% (44.8%) distance between its base and first exposed ventral procurrent caudal-fin ray; pectoral fin 17– 18 –19 (17.9) with 4– 5 –6 (4.7) dorsal and 4– 5 –8 (6.3) unbranched rays and 5– 8 –10 (7.0) branched rays in central portion of fin, fin reaching posteriorly to region above urogenital papilla to anal spine; pelvic fin I 5, fifth ray unbranched and 51– 54 –64% (55.0%) length of fourth ray, which reaches posteriorly to between bases of first to fourth anal rays, pelvic rays 1–4 with a single sequential branch point; basal membrane apparently absent or forming fold across midline above last pre-pelvic scale; no fraenum. Lateral scales 23 –24 (23.1); anterior transverse scales 9 –10 (9.1); posterior transverse scales 8 –9 (8.1); cheek and opercle scaleless, midline of predorsal with 8– 9 (8.2) scales, anterior few rows may be cycloid, otherwise ctenoid; anteriormost scales on sides of nape (may be ctenoid or cycloid) and top of nape reaching to one scale width posterior to margin of eye; 3 vertical rows of cycloid scales on pectoral fin base with 1– 2 in anteriormost row, 2– 3 in second row and 4 in outer row (n = 18); 6– 7 –8 (7.0) cycloid scales in midline anterior to pelvic fin base; area between pelvic spine and ventral margin of pectoral fin base with cycloid (smaller specimens) or ctenoid (larger specimens) scales; anterior few rows of scales in midline of belly cycloid; sometimes anteriormost row of body scales beneath axil of pectoral fin base with cycloid scales, these scales usually ctenoid. Circumpeduncular scales 12, scales rows in midline between base of last anal ray and first ventral procurrent caudal-fin ray 8. Description of jaw teeth based on ROM 1971CS (10.3–19.5 mm SL, collected with ROM 83314). Upper jaw with outer row of 5–6 spaced, enlarged curved canines to just beyond bend of premaxilla, then much smaller (1/3rd height) similar teeth to end of premaxilla; 2–3 irregular rows of small conical teeth behind outer teeth, gradually decreasing in size posteriorly to middle of length of premaxilla, becoming reduced to single row at bend of premaxilla and continuing posteriorly to distal tip; innermost few teeth near symphysis slightly larger and posteriorly oriented. Lower jaw with 3–5 enlarged, spaced, curved canines in outer row to bend of dentary; 2–3 irregular rows of slightly curved small teeth at symphysis (innermost 1.5 times preceding teeth) and reaching posteriorly to dorsal rim of coronoid process, outer row smaller and ending at base of coronoid process Tongue relatively narrow with spatulate tip. Gill opening extending anteroventrally to below mid-pupil; gill rakers 3– 5 + 13– 14 –15 = 17– 19 (4.0 + 14.3 = 18.3). Anterior naris in short tube reaching anteriorly to above anterior margin of upper lip, posterior opening pore-like with raised rim, separated from bony front of orbit by 1 –3 times its diameter (1.8), nasal sac raised and on anterior one-third of snout. Bony interorbital width 30– 37 –43% (36.6) pupil diameter; moderate U-shaped interorbital trench and narrow, groove-like postorbital slit ending at last papilla of row p ( Fig. 1 B View FIGURE 1 ); epaxialis reaching anteriorly in midline to vertical above posterior region of eye; no narrow ridge of skin in midline of nape extending anteriorly from origin of first dorsal fin. Caudal peduncle depth as percentage caudal peduncle length 40.8– 49.1 (44.3); head length as percentage SL 29.2– 30.3 –32.8 (31.5); as percentage head length: horizontal eye diameter 31.2– 37.2 –38.5 (36.2); snout length 20.1– 25.0 –27.9 (24.0); cheek depth 17.7– 25.1 (21.3). Cephalic sensory papillae as in Fig. 1 View FIGURE 1 . Number of papillae in each row: a = 6; b = 5– 6 –7 (6.0); c = 6; cp = 1; d = 5– 8 (6.3); dʹ = 6 –7 (6.6); e-anterior = 10– 14 –15 (12.4); e-posterior = 12– 13 –16 (12.7); i-anterior = 7– 8 –9 (7.9); i-posterior = 7– 8 –9 (8.0); p = 6; r = 2; f = 3 –4 (3.1); cs" = 3; g = 3–8 (5.2, n = 12, no count for holotype); n = 1; x = 5– 6 –7 (5.6); u = 4– 5 –6 (4.8); z = 5 –7 (5.4, n = 19); ot = 11– 12 –15 (12.8); os = 5– 7 (5.9, n = 17); oi = 3 –5 (3.6, n = 19). Abdominal/caudal vertebral transition Type B (based on ROM 1971CS; for definitions of Types A & B see Winterbottom, 2011:130).

Colour pattern. Freshly collected, based on 35 mm colour slide of 21.3 mm SL male, ROM 101394, from Merir I, Sonsoral State ( Fig. 2 View FIGURE 2 ). Snout, jaws and cheeks reddish with diffuse yellow bars from anterior margin of eye to premaxilla one third of its length from symphysis, from anteroventral eye to distal tip of premaxilla, below posterior margin of pupil across cheek, and over posterior margin of vertical limb of preopercle. Posterodorsal cheek, operculum and suboperculum with numerous dark melanophores on red to yellowish background. Pectoral fin base yellowish brown and more or less uniformly sprinkled with brown melanophores. Iris red, with two onequarter pupil-diameter yellow spots aligned diagonally across top of pupil, and two more across ventral portion of iris, posteroventral of these spots twice size of other three. Body mostly yellow, with margins of scales heavily invested with melanophores, especially on nape and anterodorsal part of trunk, centres of scales with scattered brown melanophores. Scale rows immediately above and below midlateral scales lighter, gradually merging with remainder of body colour anteriorly above anal fin base to give impression of three yellow stripes along posterior half of body, one dorsal, one along middle and one ventral. One-quarter pupil diameter dark basal stripe, followed by a similar yellow stripe in dorsal fins, remainder of fin membranes with faint yellow wash, fin rays reddishbrown; anal fin similar but yellow stripe less well defined and yellow wash more intense and filling most of rest of fin. Caudal fin yellow, fading distally, middle rays tinged with black at tips. Pelvic fins yellow, pectoral fin membranes hyaline, rays reddish.

Preserved. Holotype with fairly densely scattered brown melanophores on snout and anterior half of jaws, cheek with fairly even but well-spaced slightly larger brown melanophores which become somewhat smaller but more densely concentrated on posterodorsal part of cheek, over operculum and suboperculum, and then larger but still densely distributed on pectoral fin base ( Fig. 3 View FIGURE3 ). Yellow bars below posterior eye and on preopercular margin in fresh specimens faintly discernible as lighter areas. Top of snout between nasal capsules and posteriorly to posterior papilla of row r fairly densely and evenly covered with dark brown melanophores ( Fig. 4 A View FIGURE 4 ). Ventral surface between jaws, and sides of sternohyoideus muscle beneath gill membranes with scattered brown melanophores. Margins of nape and anterior body scales of dorsum faintly outlined with dark melanophores, scale pockets densely invested with large amorphous brown melanophores ( Fig. 3 View FIGURE3 ). Belly with only a few widely scattered melanophores. Pigmentation on body gradually decreasing posteriorly. Dark basal stripe in dorsal fins, rest of fin membranes with some scattered melanophores (especially first dorsal fin). Dark basal stripe of anal fin gradually increasing in width posteriorly, reaching almost to ray tips on last two rays; distal margins with scattered small melanophores. Membranes between dorsal and ventral procurrent and unbranched caudal-fin rays with some melanophores, rest of fin hyaline. Some specimens, usually but not always <15 mm SL, with variable number of up to four small diffuse dark saddles across dorsum, first between dorsal fins, second at base of third ray of second dorsal fin, third at or just behind base of last ray, and fourth 2–3 scales anterior to first dorsal procurrent ray. The latter two are present more frequently than the former two.

Etymology. From the Greek word “hamartia” (ἁµαρτία) meaning “...a mistake or error in judgment...” (see, for example, https://literarydevices.net/hamartia/), in allusion to the author’s error in not recognizing the differences between this species and T. preclarum when describing the latter. The term encompasses “wrongdoings” which may be either accidental or deliberate.

Distribution. Currently recorded with certainty from the South-West Islands of Palau. Specimens from other areas which may well be this species, but for which genetic material to test for congruence is not available, include Fiji (paratypes of Trimma preclarum ), Niue, the Solomon Islands, Indonesia (Mapia Atoll), Australia (Great Barrier Reef), and Kiribati (Abaiang Atoll)—see Discussion. The overall distribution (as given in Fig. 5 View FIGURE 5 ) assumes these specimens are in fact T. hamartium .

Comparisons. Trimma hamartium is part of an artificial grouping of three other valid species that share a narrow bony interorbital with interorbital and postorbital trenches, lack scales on the cheek and operculum but have scales present in the predorsal midline, have at least some branched ray present in the pectoral fin, and have an unbranched fifth pelvic-fin ray. These are: T. anthrenum Winterbottom, 2006 , T. lutea Viviani et al., 2016 , and T. squamicana Winterbottom, 2004 . Certain aspects of morphology differ between T. hamartium and the above species. Trimma anthrenum has fewer anterior and posterior transverse scale rows (6–8 and 6–7 vs. 9–10 and 8–9 respectively), 5 papillae in row c (vs. 6) and a freshly collected colour pattern consisting of an all yellow body with basal black stripes in the dorsal and anal fins, and a yellow iris with a black (dark blue in life) diagonal bar across the pupil which narrows distally both anteriorly and posteriorly (vs. not as above). In T. lutea , there are 10 dorsalfin rays (vs. 9), 7–8 and 6–7 anterior and posterior transverse scale rows (vs. 9–10 and 8–9 respectively), 4–7 scales in the predorsal midline (vs. 8–9), and 4–6 (vs. 6–8) prepelvic scales, and the freshly collected colour patterns consist of alternating bars of yellow and reddish-brown (vs. body plain yellow). In T. squamicana , the postorbital trench is generally poorly-developed or absent (vs. well developed) and the freshly collected and preserved colour patterns are very different (pale body with red blotches when fresh, completely pale in preservative vs. yellow with red head when fresh, head and body heavily pigmented with melanophores when preserved).

The new species can most easily be confused with T. preclarum , as demonstrated by Winterbottom’s (2006) erroneous inclusion of what appears to be T. hamartium among the type specimens of T. preclarum (the specimens from Fiji). However, it differs from T. preclarum in having 6 (vs. 5) papillae in papillae row c below the eye (cf. Fig. 1 A View FIGURE 1 with Fig. 9 A View FIGURE 9 herein and with Fig. 7 A View FIGURE 7 of Winterbottom, 2006 —note that row d is incorrectly labelled as row c, and that row dʹ is labelled as row d in that paper), a generally shorter second dorsal spine (reaching posteriorly when abducted to the interspace between the fins to the base of 3rd dorsal ray, mean = base of 1 st ray , n = 431) vs. between base of spine and 7th ray (mean = base of 4th ray, n = 91), an unbranched (vs. branched) fifth pelvic-fin ray, and the dorsal surface of the snout between the anterior nares and the 3rd papillae of row p is relatively evenly sprinkled with melanophores (vs. melanophores in irregular agglomerations with clear interspaces—cf. Fig. 4 A View FIGURE 4 vs. B and Fig. 9 B View FIGURE 9 ). Trimma cheni may also exhibit three yellow lines on the body and light bars on the cheek and opercle. It differs from T. hamartium in having 5 papillae in row c, a branched 5th pelvic-fin ray, and scales present on the upper region of the operculum, at least in larger (> 17 mm SL) specimens. This species also differs from T. preclarum in having opercular scales (vs. absent) and the top of the snout has evenly scattered melanophores (vs. melanophores in groupings separated by unpigmented areas).

Discussion. During the South-West Islands expedition, specimens were separated for an analysis of age and growth in Trimma benjamini Winterbottom, 1996 (see Winterbottom et al, 2011). These specimens were initially preserved in 80% ethyl alcohol to preserve otolith structure. Some specimens of other Trimma species were inadvertently included among this material, including five specimens of T. hamartium . Of these five specimens potentially available for DNA analysis, one failed to amplify adequately for inclusion, and the other four returned>200 but <400 base pairs. A Neighbour-Joining analysis of the somewhat limited COI data for these specimens suggests that they are phenetically closest to an undescribed species of Trimma which is currently known only from the single 11.0 mm SL tissue specimen from Olorua, Lau Is., Fiji ( Fig. 6 View FIGURE 6 , as “ Trimma species”). These two haplogroups are separated by a minimum of 7.4% of the CO1 base pairs sequenced. Together, they are phenetically closest to the two haplogroups of T. preclarum , separated by a minimum of 12.4% for the two Palau samples (type locality) and 12.1% for the two specimens from Cendrawasih Bay (identified in Fig. as T. preclarum cf). The two haplogroups currently identified as T. preclarum are separated by 10.3% of the CO1 base pairs.

As mentioned above (under Distribution), specimens that are morphologically very close or identical to T. harmartium are known from Fiji, Niue, the Great Barrier Reef of Australia, the Solomon Islands, NE Indonesia and Abaiang Atoll, Kiribati. No differences were noted between the material from Fiji (paratypes of T. preclarum , see Winterbottom, 2006, for a list of this material), and those from the South-West Islands of Palau (see Fig. 7 A View FIGURE 7 for colour of a freshly collected Fiji specimen). The colouration of specimens from these two localities appears almost identical, with three yellow stripes along the posterior body, a dark basal band followed by a yellow stripe in the dorsal and anal fins, and a complex pattern of yellow spots on a red background on the iris. Two specimens from Niue are similar, although the colouration of a freshly collected specimen is more intense (see Fig. 7 B View FIGURE 7 ). The 20.1 mm SL female from Niue has 3 ctenoid scales (of 7) in the prepelvic midline, and a few other ctenoid scales laterally; these scales are all cycloid in the 23.0 mm SL male specimen. The photographed specimen from Mapia Atoll, Pulau Pegun ( Indonesia, Fig. 7 C View FIGURE 7 ) is also more intensely coloured than the Fiji specimens (although some of this may be due to deterioration of the 35 mm colour slides of the latter, which were taken in 1983). Winterbottom and Hoese (2015) provided a re-description of what they identified as T. preclarum based on specimens from the Great Barrier Reef, Australia. Those specimens have an unbranched 5th pelvic-fin ray, and the adpressed second dorsal spine of the fist dorsal fin reaches to between the bases of the dorsal spine and 3rd ray of the second dorsal fin (mean = base of 2nd ray, n = 7). A re-examination of those specimens revealed that they possess 6 papillae in row c beneath the eye (Hoese, in litt.). No images are available for these specimens, but field notes of the colouration of freshly collected material from Ashmore Reef made by H. Larson are congruent with both T. preclarum and T. hamartium . Note that the image provided by Winterbottom and Hoese (2015, Fig. 41) under the name of T. preclarum is of a specimen of T. hamartium from Merir I., SW Islands of Palau ( Fig. 2 View FIGURE 2 herein). Because both species were present in that collection, the photographed specimen has now been re-catalogued as ROM 101394 (previously ROM 83037), and it is a male (not female as stated in Winterbottom & Hoese, 2015, figure legend for Fig. 41). The specimens cited in Winterbottom & Hoese as being T. preclarum from Abaiang Atoll, Kiribati are probably T. hamartium or a closely allied undescribed species. Re-examination of that material by Hoese (in litt.) found that 30 of the 43 specimens had 6 papillae in row c and an unbranched 5th pelvic-fin ray (condition in the remaining specimens undeterminable). A single specimen from American Samoa (AMS I34738-001) also has 6 papillae in row c and an unbranched 5th pelvic-fin ray (Hoese, in litt.), and probably represents this species. However, I have not added it to the distribution map ( Fig. 5 View FIGURE 5 ) pending collection of further material. Two specimens from Indonesia’s Pulau Pegan (ROM 101382, Mapia Atoll, about 200 kms N of the western arm of Cendrawasih Bay) are congruent in all the relevant characters with T. hamartium (see Fig. 7 C View FIGURE 7 ). This site lies some 350 kms SE of Helen Reef, the type locality.

Truimma preclarum is apparently found at Saipan ( Fig. 6 View FIGURE 6 in Winterbottom, 2006 —based on the length of the second dorsal spine, which reaches to between the bases of the 4th and 5th fin rays when adpressed). This species also seems to occur at Cendrawasih Bay, Indonesia ( Fig. 8 C View FIGURE 8 ; but note that second dorsal spine appears to be truncated, and that the posteroventral yellow iris spot is the same size as the other spots), as well as at Madang and Kimbe Bay, Papua New Guinea. Additionally there are images of what appears to be this species from Milne Bay, PNG, but I was unable to decide whether they are of T. preclarum or of T. hamartium . Genetic samples from Cendrawasih Bay differ from those from the main islands of Palau by 10.3% of the CO1 gene (see above), suggesting that caution be exercised in assigning these specimens to T. preclarum . The morphological characters used here to diagnose T. preclarum are all present in this material: 5 papillae in row c (see Fig. 9 A View FIGURE 9 ), a branched 5th pelvic-fin ray, 2nd dorsal spines reaching posteriorly to bases of 2nd to 5th fin rays of second dorsal fin, and top of snout with melanophores in agglomerations with clear interspaces ( Fig. 9 B View FIGURE 9 ).

A single specimen identified as T. hamartium (19.2 mm SL male, ROM 83073) from Tobi Island had 5 papillae in row c (bilaterally). This specimen has the 5th pelvic-fin rays unbranched, and the pigmentation on the dorsal surface of the snout is more or less evenly distributed (not concentrated in discreet agglomerations). The second dorsal spine is broken in this specimen, but the third dorsal spine reaches to the base of the spine of the second dorsal fin (3nd spine mean length to base of second ray of second dorsal fin in T. preclarum , n = 101 and base of first ray in T. hamartium , n = 437). It thus exhibits two, if not three of the four quantifiable defining characters of T. hamartium . All other 50 specimens in this lot had 6 papillae in row c. Only two collections were made at Tobi, neither of which definitively contains specimens of T. preclarum , although that species has been recorded from the South-West Islands of Palau at Merir and Sonsoral Islands (195 km NE and 285 km NNE of Tobi, respectively). There are currently no records of T. preclarum from either Tobi or the neighbouring Helen Reef (70 km to the ESE), and therefore I identify this specimen as an aberrant T. hamartium , although the possibility of it being a hybrid between the two species is not rejected.

ROM

Royal Ontario Museum

Kingdom

Animalia

Phylum

Chordata

Class

Actinopterygii

Order

Perciformes

Family

Gobiidae

Genus

Trimma

Loc

Trimma hamartium

Winterbottom, Richard 2018
2018
GBIF Dataset (for parent article) Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF