Siphamia guttulata, : Allen et al. 2006:, 1108

Pogonoski, John J., Gon, Ofer & Appleyard, Sharon A., 2020, Redescription and distributional range extension of the Speckled Siphonfish, Siphamia guttulata (Pisces: Apogonidae), Zootaxa 4766 (2), pp. 377-388 : 379-385

Siphamia guttulata ­(Alleyne­&­Macleay­1877)

Figures 1–2 , 3 A–B, 4–5, Tables 1–3

Apogon guttulatus Alleyne & Macleay 1877: 267 , plate 5, fig.1 (type locality, Darnley Island, Torres Strait, Queensland); Stanbury, 1969: 206.

Siphamia guttulata: Allen et al. 2006:1108 ; Gon & Allen 2012:48; Fraser & Prokofiev 2016: 229; Kuiter & Kozawa 2019: 154.

Siphamia guttulatus: Munro 1960: 139 ; Paxton et al. 1989: 557; Allen 1999: 2608; Mabuchi et al. 2014: 201.

Lectotype.­ AMS IA.8103, 22.5 mm, Torres Strait, Darnley Island , Queensland, 9°35’S, 143°46’E, 1875.

Paralectotypes.­ AMS I.16306-001 [ex MAMU F402], 31 specimens, 19.0–28.0 mm, collected with lectotype.

Non-type­specimens.­ 31 specimens, 16.6–27.6 mm SL, all from Queensland, unless otherwise indicated. AMS I.49030-001, 20.3 mm SL, Cumberland Islands, east of Repulse Bay, 20°34.79’S, 149°08.95’E, 27 m, 11 Dec. 2003; CSIRO H 6648-02, 25 mm SL, north of Edgecumbe Bay, 19°48.59’S, 148°15.87’E, 35 m, 30 Apr. 2004; CSIRO H 6685-03, 23 mm SL, CSIRO H 6685-04, 3: 21.55–23.8 mm SL, north-east of Innisfail, 17°13.50’S, 146°21.57’E, 44 m, 19 Oct. 2004; CSIRO H 6734-02, 5: 21.3–24.55 mm SL, south-east of Cooktown, 15°42.67’S, 145°37.36’E, 35 m, 18 Nov. 2003; CSIRO H 6930-02, 25 mm SL, CSIRO H 6930-03, 2: 21.5–24.2 mm SL, south-east of Cairns, 17°06.36’S, 146°00.30’E, 20 m, 18 Oct. 2004; CSIRO H 7026-02, 24.45 mm SL, north-east of Cairns, 16°45.93’S, 146°04.62’E, 40 m, 11 Oct. 2004; CSIRO H 7457-01, 27 mm SL, CSIRO H 7457-03, 26.0 mm SL, north of Cape Weymouth, 12°34.72’S, 143°28.70’E, 33 m, 3 Oct. 2004; CSIRO H 7486-03, 2: 16.6–24.65 mm SL, south of Cape Flattery, 15°02.64’S, 145°20.17’E, 20 m, 13 Oct. 2004. CSIRO H 7487-03, 25.2 mm SL, east of Port Douglas, 16°28.65’S, 145°41.20’E, 31 m, 22 Nov. 2003; CSIRO H 8479-01, 24.4 mm SL, Torres Strait, north of Darnley Island, 9°31.23’S, 143°45.68’E, 37 m, 28 Jan. 2004; CSIRO H 8480-01, 2: 21.05–23.85 mm SL, north-west of Lizard Island, 14°33.24’S, 145°19.28’E, 26 m, 20 Nov. 2003; CSIRO H 8481-01, 19.9 mm SL, north of Port Douglas, 16°21.77’S, 145°30.38’E, 16 m, 18 Nov. 2003; CSIRO H 8483-01, 21.8 mm SL, Torres Strait, east of Coconut Island, 10°02.11’S, 143°11.47’E, 25 m, 29 Jan. 2004; CSIRO H 8484-01, 22.0 mm SL, south-east of South Island, Northumberland Islands Group, 21°53.14’S, 150°28.56’E, 41 m, 14 Nov. 2005; CSIRO H 8489-01, 27.6 mm SL, north of Princess Charlotte Bay, 13°57.81’S, 143°58.72’E, 36 m, 6 Oct. 2004; CSIRO H 8490-01, 22.6 mm SL, south-east of Cooktown, 15°42.21’S, 145°42.49’E, 46 m, 18–19 Nov. 2003; NTM S.18369-001, 23.7 mm SL, north-east of Orford Bay, 11°09.69’S, 142°57.34’E, 28 m, 30 Sep. 2004; QM I.40983, 20.3 mm SL, Torres Strait, south-east of Dalrymple Island, 9°46.09’S, 143°32.24’E, 41 m, 29 Jan. 2004; WAM P.35009.001, 24.2 mm SL, north-east of Great Palm Island, 18°34.23’S, 146°48.84’E, 43 m, 15 Dec. 2003.

Diagnosis.­ A species of the Siphamia tubifer group with 2 supraneurals, 0–1 median predorsal scales and 13–25 preopercular serrations.

Description.­ Based on the material listed above. Dorsal rays VII+I,9; anal rays II,8; pectoral rays 14–16 (modally 15); median predorsal scales 0–1; developed gill rakers 1 + 7–8; supraneurals 2. Body depth 2.2–2.8 in SL and body width 1.9–2.25 in the depth; eye diameter 2.7–3.2 in head length; first dorsal spine 2.1–2.9 in second spine; second dorsal spine 3.7–4.9, spine of second dorsal fin 4.0–5.0, and second anal spine 4.0–5.1, all in head length; pectoral-fin length 4.1–5.5 and pelvic-fin length 3.7–4.6 in SL; caudal-peduncle length 1.4–2.1 in distance between pelvic spine insertion and anal-fin origin; predorsal distance 2.0–2.2, preanal distance 1.3–1.4 and prepelvic distance 8.8–11.35, all in SL. Preopercular edge with 13–25 serrations around angle and ventral edge; scales spinoid; light organ ending over to 1 st to 6 th procurrent caudal rays in specimens 20–26 mm SL

Colour (from Figure 1 ; photograph taken after specimen stored frozen for eight years). Head and body with varying shades of brown, darker on nape, becoming lighter posteriorly and peppered with black dots of different sizes. Iris dark brown. Jaws creamy white, peppered with small dark dots. Abdominal area, gill cover, and area between eye and upper jaw silvery. Light organ silvery, with dark vertical striations. A line of large, black melanophores along midline of body from upper posterior edge of preopercle to caudal-fin base. Fins pale; first dorsal-fin spines, pelvic spine and posteriormost second dorsal-fin rays with small black dots.

Comparisons.­ Species of Siphamia fall into two groups easily distinguished by the colour pattern of the light organ made of a silvery region extending along the ventral margin of the body from the isthmus to the caudal peduncle. Siphamia guttulata belongs to the S. tubifer species group which has a striated light organ ( Fig. 1 ). Species of the second group have a dotted light organ (see images of both groups in Gon & Allen 2012). Within the S. tubifer group S. guttulata ( Fig. 2 ) and S. argentea Lachner 1953 share two supraneurals; all other species of this group have one supraneural. Most members of the S. tubifer group, including S. guttulata , share a similar colour pattern of, usually, three dark stripes on a pale background. The exceptions are S. jebbi Allen 1993 and S. stenotes Gon & Allen 2012 , the former with no stripes and the latter with two stripes on the upper part of the body ( Gon & Allen 2012). These two species as well as S. argentea differ from S. guttulata and the remaining member of the S. tubifer group in having 13 pectoral-fin rays [vs. 14–16 (usually 15) rays]. In addition, S. guttulata differs from all other members of the S. tubifer group in having the scales in its predorsal area arranged irregularly resulting in none or one proper predorsal scale ( Fig. 3 A–B). The median predorsal scales of S. guttulata requires further clarification. Predorsal scales are present either side of the dorsal midline on the type specimens, with some appearing to have a very small median predorsal scale, usually just prior to the first dorsal fin spine ( Fig. 3A ). For the non-type specimens most scales were lost in the trawls, but in at least one specimen a small median predorsal scale was present ( Fig. 3B ). In the other Queensland survey specimens, it was impossible to reliably determine the presence/absence of this small scale. Further examination of this character on specimens in better condition will be required to resolve this issue. In comparison, the median predorsal scales of S. tubifer are larger and more obvious ( Fig. 3C ) and range from 3–5 in number ( Gon & Allen 2012).

DNA­barcoding. Twenty-six specimens were successfully barcoded at the COI gene. Following alignment of the consensus sequences and the Pterapogon kauderni sequence from GenBank, 633 base pairs of the gene were analysed. The ratios of nucleotides in the COI gene across the 26 Siphamia specimens ranged from 19.5% (G) to 29.6% (T), with 211 of the 633 representing parsimony informative sites. The within species variation was very low (reflected by small genetic distances among individuals of each species) - Siphamia cuneiceps (d = 0.006); S. roseigaster (d = 0.003); S. tubulata (d = 0.006); S. tubifer (d = 0.003) and S. guttulata (d = 0.004). Pairwise distances (based on the number of base substitutions) between the species was much larger and is shown in Table 3 . The largest divergence between species was S. guttulata and S. roseigaster / S. tubulata .

The within group distance for the individuals from each of the species in the ‘ S. tubulata group’ was higher (0.170, s.e ±0.010) than the distances among individuals of each of the species in the ‘ S. tubifer group’ (0.070, s.e ±0.010). The topology of the NJ and ML trees of the COI gene was identical and clearly demonstrates the among species genetic distances with the separation of each species strongly supported. The maximum likelihood tree is shown in Figure 5 . An attempt was made to extract DNA out of 3 of the 31 paralectotypes of Siphamia guttulata (AMS I.16306-001, 21.9–24.8 mm SL) although the early fixation history of these specimens was unknown. However, low yields and poor-quality DNA from these specimens prevented the acquisition of adequate sequence lengths to be informative.

Remarks.­ Alleyne & Macleay (1877) described Apogon guttulatus as having three dark stripes on the body, practically invisible in their small black and white illustration, the lowest of which marks “the limits of a very silvery belly.” A close examination of this drawing (also reproduced by Kuiter & Kozawa 2019: 154), enlarged on a computer screen, reveals a darkish area extending backwards from the lower part of the gill cover along the ventral contour of the body. It would be difficult for an eye not trained for the light organ of Siphamia to relate this species to Weber’s (1909) genus. Indeed, McCulloch (1929) classified it in Apogonichthys Bleeker 1854 . It is therefore not surprising that 51 years went by before Munro (1960) made the correct generic association.

The proportional measurements and counts of the studied specimens ( Table 2 ) are compared with those of Gon & Allen (2012). In conjunction with the striated light organ, colour pattern and molecular data, these data helped confirm the identification of the specimens as Siphamia guttulata . The Queensland specimens we examined lost most of their scales due to abrasion during capture in prawn trawl (the types of the species at AMS were dip-netted and are in better condition). The largest specimen in this study was 27.6 mm SL, a similar size to the maximum size of 28.0 SL mm given by Gon & Allen (2012). The range of the spines’ proportions in this study ( Table 2 ) indicates relatively high character variability and this was also evident in other species of the Siphamia tubifer group such as S. goreni Gon & Allen 2012 , S. randalli Gon & Allen 2012 , and S. tubifer , as noted by Gon & Allen (2012). Twenty-three recently collected specimens were counted for pectoral-fin rays. The majority (14) of these fish had 15 rays in both pectoral fins, and another two had 15 on the left fin while the right one was damaged; three fish had 16 rays; one fish had 14 rays on the left pectoral-fin and three had this count on the right fin. Gill rakers were counted in 18 specimens and all had a total of 8 developed rakers and 7 rakers on the ceratobranchial. The light organ of the smallest specimen ( CSIRO H 7486-03, 16.6 mm SL) ended at about two thirds of the caudal peduncle length; in a 19.9 mm SL specimen ( CSIRO H 8481-01) it ended in front of the first procurrent caudal-fin ray; and in the largest non-type specimen ( CSIRO H 8489-01, 27.6 mm SL) it ended over the 6 th procurrent caudal-fin ray. The Queensland specimens differed slightly in morphometrics from the type specimens as follows: 1.9–2.25 body width in the former (vs. 1.8 in latter; Gon & Allen 2012), because most of the specimens examined in this study were frozen for 8–10 years before fixation.

Siphamia guttulata is a common resident in the Great Barrier Reef (GBR) ecosystem. It was collected in many areas from Darnley Island, Torres Strait, (09°35’S, 143°46’E) southwards to Northumberland Islands Group, Queensland (21°53.14’S, 150°28.56’E), encompassing most of the GBR ( Fig. 4 ), at a depth range of 11– 46 m. Its absence from over 100 demersal trawl stations, including night-time hauls, during the same survey undertaken south of 22°S suggests it is unlikely to be present in the extreme southern GBR. Although the exact depth of collection of the type specimens at Darnley Island was not originally stated, Alleyne & Macleay (1877) noted that it was “very numerous…in holes in the rocks at low water,” implying that the collectors were able to see the fish while wading in the water, possibly with the aid of a viewing box or primitive snorkelling equipment. The species is likely to occur in nearby Papua New Guinea waters.

Comparative­material: Siphamia cuneiceps (6 specimens, 27–32 mm SL, all from Queensland): CSIRO H 6791-04, 27 mm SL, Torres Strait, west of Banks (Moa) Island, 10°15.52’S, 141°51.06’E, 13 m, 17 Jan. 2004; CSIRO H 6901-13, 27 mm SL, Torres Strait, south of Dungeness (Zagai) Island, 9°57.01’S, 142°53.76’E, 10 m, 20–21 Jan. 2004; CSIRO H 7679-03, 29 mm SL, north of Broad Sound, 21°43.81’S, 149°36.46’E, 14 m, 12 Nov. 2005; CSIRO H 8482-01, 32 mm SL, CSIRO H 8482-02, 32 mm SL, CSIRO H 8482-03, 30 mm SL, south-east of Mackay, 22°07.49’S, 150°19.48’E, 26 m, 14 Nov. 2005. Siphamia roseigaster (21 specimens: 32–48 mm SL, all from Queensland, unless otherwise indicated): CSIRO H 6440-02, 8: 35–41 mm SL, Torres Strait, south of Bristow (Bobo) Island, Papua New Guinea, 9°13.68’S, 143°19.23’E, 36 m, 25 Jan. 2004; CSIRO H 6912-05, 5: 37–48 mm SL, Torres Strait, east of Saibai Island, Papua New Guinea, 9°18.98’S, 142°55.43’E, 11 m, 23 Jan. 2004; CSIRO H 7452-01, 40 mm SL, CSIRO H 7452-02, 6: 32–48 mm SL, Broad Sound, 22°07.59’S, 149°36.95’E, 16 m, 27 Apr. 2004; CSIRO H 8478-01, 39 mm SL, Broad Sound, 22°07.75’S, 149°41.21’E, 18 m, 27 Apr. 2004. Siphamia tubifer (11 specimens: 24–35 mm SL, all from Queensland): CSIRO H 6516-02, 29 mm SL, north of Princess Charlotte Bay, 13°43.52’S, 144°01.49’E, 40 m, 05 Oct. 2004; CSIRO H 6535-04, 29 mm SL, south-east of Cairns, 17°08.66’S, 146°32.05’E, 71 m, 19 Oct. 2004; CSIRO H 6701-03, 28 mm SL, north-east of Rockingham Bay, 17°43.21’S, 146°43.68’E, 67 m, 13 Dec. 2005; CSIRO H 6722-08, 27 mm SL, Torres Strait, north-east of Cape York Peninsula, 10°31.84’S, 143°51.19’E, 27 m, 10 Jan. 2004; CSIRO H 6752-02, 31 mm SL, CSIRO H 6752-03, 2: 33.0– 33.6 mm SL, north-east of Hinchinbrook Island, 17°50’S, 146°45’E, 72 m, 28 Nov. 2003; CSIRO H 7026- 03, 35 mm SL, north-east of Cairns, 16°45.93’S, 146°04.62’E, 40 m, 11 Oct. 2004; CSIRO H 7460-01, 32 mm SL, north-east of Townsville, 18°31.23’S, 147°35.45’E, 76 m, 14 Dec. 2003; CSIRO H 7461-01, 26 mm SL, east of Hinchinbrook Island, 18°22.63’S, 146°44.59’E, 51 m, 1 May 2004; CSIRO H 7463-01, 24 mm SL, north-east of Mackay, 20°06.24’S, 150°13.48’E, 74 m, 5 Dec. 2005. Siphamia tubulata (10 specimens: 17–32 mm SL, all from Queensland): CSIRO H 6145-08, 17 mm, Torres Strait, north-west of Prince of Wales (Muralug) Island, 10°33.57’S, 141°37.14’E, 17 m, 16 Jan. 2004; CSIRO H 6442-05, 27 mm SL; north-east of Whitsunday Island group, 19°40.31’S, 150°04.95’E, 73 m, 5 Dec. 2003; CSIRO H 6713-02, 30 mm SL, Shoalwater Bay, 22°15.09’S, 150°11.73’E, 10 m, 14 Nov. 2005; CSIRO H 7453-01, 31 mm SL, north-east of Northumberland Islands, 20°56.73’S, 150°20.35’E, 43 m, 4 Dec. 2005; CSIRO H 7660-05, 28 mm SL, east of Northumberland Islands, 21°37.92’S, 150°07.89’E, 34 m, 28 Nov. 2005; CSIRO H 7894-02, 32 mm SL, Torres Strait, Seven Reefs, 10°19.14’S, 143°44.99’E, 54 m, 31 Jan. 2004; CSIRO H 8031-02, 25.4 mm SL, east of Bowling Green Bay, 19°09.45’S, 148°07.06’E, 54 m, 12 Dec. 2003; CSIRO H 8487-01, 29 mm SL, east of Cumberland Islands, 20°36.36’S, 150°15.98’E, 45 m, 4 Dec. 2005; CSIRO H 8488-01, 2: 31–32 mm SL, north of Whitsunday Islands, 19°39.87’S, 149°09.91’E, 49 m, 8 Dec. 2005.

Figures

FIGURE­1. Lateral view of Siphamia guttulata (CSIRO H 6930-02, 25 mm SL). Scale bar = 5 mm.

FIGURE­2. Digital radiograph of Siphamia guttulata (CSIRO H 6648-02, 25 mm SL). Blue circle highlighting two supraneurals. Scale bar = 0.5 mm.

FIGURE­3.­Median predorsal scales of Siphamia guttulata (A—paralectotype, AMS I.16306-001, 25.7 mm SL; B—CSIRO H 7457-01, 27 mm SL) and Siphamia tubifer (C—CSIRO H 6752-02, 31 mm SL). Scale bar = 0.5 mm. Yellow arrows point to first dorsal spine, numbers or pink arrows indicate median predorsal scales. Fifth median predorsal scale of Fig. C is a scale pocket as the scale has been lost.

FIGURE­5. Maximum likelihood tree based on the Kimura 2-parameter model of the COI sequences from the Siphamia species with P. kauderni as the outgroup. Tree shown here has the highest log likelihood following 10 000 replications. The percentage of trees in which the associated taxa clustered together is shown next to the branches, branch lengths are measured in the number of substitutions per site and all positions containing gaps and missing data have been eliminated.

FIGURE­4.­Map of selected records of Siphamia species recorded in Queensland and Papua New Guinea waters in 2003–2005, including the type locality of Siphamia guttulata. Some symbols represent more than one specimen.

Tables

TABLE­3. Estimates of genetic divergence of sequence pairs between species groups; number of base substitutions from averaging over all sequence pairs between species are shown. Analyses are based on the Kimura 2-parameter distance model, genetic distances below the diagonal and standard error rates above.

	S. cuneiceps	S. roseigaster	S. tubulata	S. tubifer	S. guttulata
S. cuneiceps	-----	0.018	0.022	0.021	0.023
S. roseigaster	0.190	-----	0.021	0.022	0.025
S. tubulata	0.252	0.241	-----	0.023	0.024
S. tubifer	0.234	0.240	0.232	-----	0.021
S. guttulata	0.255	0.267	0.268	0.209	-----

TABLE­3. Estimates of genetic divergence of sequence pairs between species groups; number of base substitutions from averaging over all sequence pairs between species are shown. Analyses are based on the Kimura 2-parameter distance model, genetic distances below the diagonal and standard error rates above.

TABLE­2. Summary of morphometric and meristic data for Siphamia guttulata. Morphometrics expressed as a proportion of standard length unless followed by mm in parentheses.

Characters	This study (n=10)			Gon & Allen 2012 (n=5)
	Range	Mean ± std	Median	Range	Mean ± std	Median
Standard length (mm)	20.3–26.0			22.5–26.0
Head length (mm)	9.8–12.6			10.5–12.1
Head length	45.2–50.1	47.6 ± 1.7	47.6	45.6–50.7	47.1 ± 2.1	46.5
Body depth	35.9–44.9	39.9 ± 3.0	39.4	39.1–44.4	40.6 ± 2.2	39.6
Snout length	9.1–11.6	10.2 ± 0.7	10.1	8.3–9.6	8.9 ± 0.6	8.65
Eye diameter	14.4–17.6	15.9 ± 0.8	16	14.4–16.0	15.3 ± 0.6	15.4
Upper jaw length	22.7–25.5	24.2 ± 0.9	24.4	24.6–26.7	25.3 ± 0.8	25.2
Lower jaw length	23.0–29.2	26.2 ± 1.8	26.3	28.4–30.9	29.5 ± 1.0	29.6
Pectoral-fin length	18.1–24.2	22.4 ± 1.9	23	18.6–21.5	19.7 ± 1.2	19.9
Pelvic-fin length	21.8–27.2	23.7 ± 1.6	23.5	22.9–26.4	24.6 ± 1.2	24.4
Caudal peduncle depth	10.5–12.1	11.2 ± 0.6	10.9	10.8–11.4	11.1 ± 0.3	11.1
Caudal peduncle length	15.0–20.6	17.8 ± 1.7	17.6	16.4–18.5	17.2 ± 0.9	16.9
Body width	15.5–20.1	18.5 ± 1.5	18.9	21.3–24.2	22.3 ± 1.1	21.9
Predorsal distance	44.8–50.2	47.0 ± 1.7	46.9	44.0–48.0	45.3 ± 1.6	45
Preanal distance	71.2–76.4	73.4 ± 1.8	73.5	68.7–72.0	70.9 ± 1.3	71.3
Prepelvic distance	42.7–47.3	44.7 ± 1.7	43.9	41.7–46.2	43.5 ± 1.7	42.9
Distance from pelvic-fin insertion to anal-fin origin	29.2–35.8	32.2 ± 2.8	31.2	29.3–32.7	31.2 ± 1.4	31.75
Length of 1st dorsal-fin spine	3.7–5.3	4.6 ± 0.5	4.5	1.7–4.0	3.1 ± 0.9	2.9
Length of 2nd dorsal-fin spine	10.0–12.6	11.2 ± 0.8	11.1	8.1–10.9	9.4 ± 1.3	9.3
Length of 2nd anal-fin spine	9.1–11.8	10.3 ± 1.0	9.9	7.3–8.7	7.8 ± 0.6	7.5
Maxilla width	5.5–6.8	6.2 ± 0.4	6.1	5.1–6.0	5.7 ± 0.3	5.7
Meristics	This study (n=18–23)			Gon & Allen 2012 (n=14–20)
Left pectoral- fin rays	14–16		15	15–16		15
Right pectoral- fin rays	15–16		15	15–16		15
Developed gill rakers	1 + 8			1 + 7–8
Ceratobranchial gill rakers	7			7
Supraneurals (n=31)	2			2
Serrations on preopercular edge	13–25			16–20

TABLE­2. Summary of morphometric and meristic data for Siphamia guttulata . Morphometrics expressed as a proportion of standard length unless followed by mm in parentheses.