New collections

Known host: Siganus fuscescens.

New localities: Tomari fish market (26°13′N, 127°40′E), Naha, Okinawa Prefecture, Japan.

Site in host: Intestine.

Deposition of specimens: 5 voucher specimens, MPM 25,532.

Molecular sequence data: cox1 mtDNA, PZ466217–18; ITS2 rDNA, PZ469408; 28S rDNA, PZ469389.

ZooBank LSID: urn:lsid:zoobank.org:act:3F2AEF6D-1B1B-49FA-8F52-6B74CFFD3759.

Remarks

We did not recover any new material of a suitable standard to allow useful morphological description; the worms were already dead and macerating when collected but did allow reliable sequencing. Egg measurements are given in Supplementary Table 1.

In our interpretation of this species, we accept as presently reliable only the original description by Goto and Ozaki (Reference Goto and Ozaki1929) from S. fuscescens from Japanese waters and the subsequent reports by Yamaguti (Reference Yamaguti1934) and Machida et al. (Reference Machida, Ichihara and Kamegai1970) from the same area and host. The many other reports from other localities and hosts (including several non-siganids) fall into three broad categories: (i) plausible but requiring molecular testing; (ii) inconsistent with what is understood for H. sigani; or (iii) too incompletely characterised to allow interpretation relative to what is known for morphology and host distribution of H. sigani. Multiple non-Japanese reports (Nagaty Reference Nagaty1954, Velasquez Reference Velasquez1961, Al-Yamani and Nahhas Reference Al-Yamani and Nahhas1981, Kardousha Reference Kardousha2003, Al-Zanbagi and Hassan Reference Al-Zanbagi and Hassan2013, Hassanine et al. Reference Hassanine, Al-Zahrani, Touliabah and Youssef2016, Khalifa et al. Reference Khalifa, Hassan, Mohamadain and Karar2018, Villahermosa et al. Reference Villahermosa, Estaño, Calagui and Paylangco2022, Abo Elhussien et al. Reference Abo Elhussien, Hussein, Rabie and Abuelwafa2025) are broadly plausible. The report of two specimens of H. sigani from an Acanthurus sp. from the Andaman and Nicobar Islands by Gupta and Miglani (Reference Gupta and Miglani1976) shows the vitelline follicles as exceptionally distant from the oesophageal bulb and an exceptionally small pharynx; it seems highly unlikely that these specimens represent H. sigani. The report of H. loossi, which was interpreted as a synonym of H. sigani by Velasquez (Reference Velasquez1961), is confused by the fact that the species was reported from both a siganid and a scarine labrid and that the figured samples appear significantly different without attribution to the two hosts. Numerous studies (Yamaguti Reference Yamaguti1953, Fischthal and Kuntz Reference Fischthal and Kuntz1964, Fischthal and Kuntz Reference Fischthal and Kuntz1965, Ku and Shen Reference Ku and Shen1965, Durio and Manter Reference Durio and Manter1968, Gupta and Miglani Reference Gupta and Miglani1976, Parukhin Reference Parukhin1976, Toman Reference Toman1977, Gupta and Tandon Reference Gupta and Tandon1983, Martens and Moens Reference Martens and Moens1995, Geets and Ollevier Reference Geets and Ollevier1996, Dzikowski et al. Reference Dzikowski, Paperna and Diamant2003, Sey et al. Reference Sey, Nahhas, Uch and Vang2003, El-Labadi et al. Reference El-Labadi, Ismail and Khalaf2006, Abd-Elmonsef Mahmoud et al. Reference Abd-Elmonsef Mahmoud, Osman and Abdel-Hakeem2020, El-Ella et al. Reference El-Ella, Youssef, Sallam, Hassanin and Abouelhassan2023, Motson et al. Reference Motson, Hutson and Hoey2023) are variously lacking in details, figures or report improbable (especially non-siganid) hosts and cannot presently be considered convincing. In our view, these results all require reconsideration, especially in the light of molecular data. At present, H. sigani is best considered an exceptionally large species of Hexangium seemingly restricted to S. fuscescens and perhaps other closely related species.

Hexangium affinum Tubangui & Masiluñgan, 1944 (Figures 2A–L)

Figure 2.

Gravid Hexangium affinum Tubangui & Masiluñgan, 1944. A–C, Heron Island: A, from S. corallinus; B, from S. punctatus; C, from S. doliatus (hologenophore). D–J, Lizard Island: D, from S. argenteus (hologenophore); E, from S. corallinus (hologenophore); F–H, from S. doliatus (H and I are hologenophores); J, from S. puellus. K–L, Ningaloo Reef: K, from S. doliatus (hologenophore); L, from S. virgatus (hologenophore). Scale bars = 1 mm.

Figure 2. Long description

A multi-panel set of 12 line drawings (A through L) illustrating the morphology of the flatworm Hexangium affinum. Each drawing shows an elongated, translucent body with a small circular oral sucker at the anterior tip.

* Panels A through F are arranged in the top row.

* Panels G through L are arranged in the bottom row.

Each specimen features a central digestive tract and a prominent uterus filled with dark, oval-shaped eggs concentrated in the mid-section. At the posterior end of each worm, there are two large, rounded testes and a smaller, stippled ovary.

* Specimen A is the longest and most complete, showing a clear distribution of eggs along the lateral margins of the midbody.

* Specimens C, D, I, K, and L are partial drawings, focusing on the anterior and mid-sections while the posterior tips are omitted.

* Specimen J is notably wider and more robust than the others.

Vertical scale bars representing 1 mm are placed to the left of each specimen to indicate relative size. The drawings highlight variations in body length, width, and egg density among specimens collected from Heron Island (A to C), Lizard Island (D to J), and Ningaloo Reef (K to L).

Synonym: Hexangium secundum Annereaux, 1947

Type-host: Siganus javus (Linnaeus), Streaked spinefoot (Siganidae).

Type-locality: Manila, the Philippines.

Records: 1. Tubangui and Masiluñgan (Reference Tubangui and Masiluñgan1944); 2. Annereaux (Reference Annereaux1947).

Hosts: Siganidae – Siganus guttatus (Bloch), Orange-spotted spinefoot (2); Siganus javus (1).

Localities: Manila, the Philippines (1); near Mercedes, Samar, the Philippines (2).

ZooBank LSID: urn:lsid:zoobank.org:act:2E3A8F6C-533A-47B7-9E65-6FC7F1FE78FE.

New collections

New hosts: Siganidae – Siganus argenteus (Quoy & Gaimard), Streamlined spinefoot; Siganus corallinus (Valenciennes), Blue-spotted spinefoot; Siganus doliatus Guérin-Méneville, Barred spinefoot; Siganus puellus (Schlegel), Masked spinefoot; S. punctatissimus Fowler & Bean, Peppered spinefoot; Siganus punctatus (Schneider & Forster), Goldspotted spinefoot; Siganus trispilos Woodland & Allen, Threeblotched spinefoot; Siganus virgatus (Valenciennes), Barhead spinefoot. Zanclidae – Zanclus cornutus (Linnaeus), Moorish Idol.

New localities: Off Heron Island (23°27′S, 151°55′E), southern Great Barrier Reef; off Lizard Island (14°40′S, 145°27′E), northern Great Barrier Reef, Australia; Ningaloo Reef (22°42′S, 113°40′E), Western Australia; off Noumea (21°30′S, 165°30′E), New Caledonia.

Host/locality combinations: Heron Island – S. corallinus, S. doliatus and S. punctatus. Lizard Island – S. argenteus, S. corallinus, S. doliatus, S. punctatissimus, S. punctatus, S. puellus and Z. cornutus. Ningaloo Reef – S. argenteus, S. trispilos and S. virgatus. New Caledonia – S. argenteus, S. corallinus and S. doliatus.

Site in host: Intestine.

Deposition of specimens: 96 voucher specimens QM G242,643–738; 14 voucher specimens, WAM 13,794–807.

Molecular sequence data: cox1 mtDNA, PZ466219–39; ITS2 rDNA, PZ469409–16; 28S rDNA, PZ469384–88.

Description

[Based on all specimens identified as H. affinum. Measurements are of eight hologenophore and three paragenophore specimens. 98 voucher specimens (no molecular data) examined and identified but not incorporated in main measurements. Measurements for each host/locality combination in Supplementary Table 2. Egg measurements given in Supplementary Table 1.]

Body elongate, frequently conspicuously narrow, with margins ranging from almost parallel to clearly widest in middle or posterior third of body, 2,871–3,193 (3,065) × 405–670 (518). Length to anterior testis 1,867–2,910 (2220). Length to anterior testis: body width ratio 3.76–5.12 (4.33):1. Largest gravid specimens approximately twice length of smallest. Tegument armed with fine spines; spines most clearly visible at anterior extremity. Eye-spot pigment not detected. Pharynx typically slightly wider than long, 106–157 (130) × 115–168 (148). Oesophagus conspicuously elongate, 447–650 (531) in length. Oesophagus length: pharynx length ratio 3.60–4.64 (4.10):1. Oesophageal bulb distinct. Intestinal caeca terminate close to anterior margin of anterior testis, highly variable in width.

Testes two, entire, contiguous, occasionally partly flanked by intestinal caeca, usually somewhat oblique but sometimes tandem or opposite. Seminal vesicle winding from well posterior to intestinal bifurcation to genital pore, typically largely obscured by uterus. Poorly developed ‘cirrus-sac’ present. Genital pore median, midway between pharynx and intestinal bifurcation.

Ovary entire, usually ventral to and partially overlapping posterior testis. Uterus normally well-developed from about anterior margin of anterior testis to about intestinal bifurcation, in smaller specimens appears to follow helical path but broadly irregular in fully mature specimens, typically ventrally overlapping intestinal caeca when full of eggs. Ventral vitelline follicles in two lateral fields, ventral to intestinal caeca, typically extend anteriorly from just anterior to ends of intestinal caeca (occasionally reaching ends) to 256–555 (380) from intestinal bifurcation, rarely almost reaching it; dorsal vitelline follicles medially confluent, dorsal to intestinal caeca and uterus but usually largely obscured by uterus, sometimes include overall most anterior follicles. Eggs 62–89 (77) × 38–56 (49) (n = 137, all samples).

Excretory pore dorsally subterminal, close to posterior extremity of body. ‘Rosette organ’ represented only by a pair of distinct diverticula that unite and open at excretory pore. Excretory vesicle thin-walled, voluminous, usually extends to posterior margin of ovary.

Remarks

These specimens are considered identifiable as the known species H. affinum (see Discussion). Specimens interpreted as this species are moderately consistent in their morphology, most importantly in their possession of a characteristically long oesophagus. Some of the specimens, including hologenophores, are amongst the narrowest seen among all six species in this study (e.g. Figures 2H, I), but this character is not unique to the species or entirely reliable within it. The largest specimen interpreted as this species (Figure 2J) from S. puellus at Lizard Island was anomalous in having no eggs in the uterus. We suspect that it was senescent, and as such perhaps a good indicator of the maximum size reached by this species.

Hexangium urabeae n. sp. (Figures 3A–I)

Figure 3.

Gravid Hexangium urabeae n. sp. A–G, Lizard Island: A–D, from S. lineatus (C is a hologenophore); E, from S. corallinus (hologenophore); F, from S. doliatus (hologenophore); G, from S. punctatus (hologenophore). H, Heron Island, from S. lineatus. I, New Caledonia, from S. doliatus. Scale bars = 1 mm.

Figure 3. Long description

A multi-panel line drawing containing nine specimens labeled A through I. Each drawing depicts an elongated, worm-like body with a distinct oral sucker at the anterior end.

* Anatomical structure from anterior to posterior: A circular oral sucker leads to a narrow esophagus, which bifurcates into two long intestinal caeca extending down the sides of the body. The central region contains a convoluted tubular uterus. Dark, solid-black circular spots representing vitelline follicles are distributed along the lateral fields. Near the posterior end, two large oval testes are situated side-by-side, followed by a smaller ovary and a terminal excretory pore.

* Panel A: A full specimen from S. lineatus at Lizard Island.

* Panel B: A smaller, narrower specimen from S. lineatus at Lizard Island.

* Panel C: A close-up of the anterior and mid-section of a hologenophore specimen.

* Panel D: A complete specimen from S. lineatus at Lizard Island.

* Panel E: A specimen from S. corallinus at Lizard Island.

* Panel F: A specimen from S. doliatus at Lizard Island.

* Panel G: A specimen from S. punctatus at Lizard Island.

* Panel H: A specimen from S. lineatus at Heron Island.

* Panel I: The largest specimen shown, from S. doliatus at New Caledonia, featuring a more robust body and denser vitelline follicles.

Vertical scale bars representing 1 mm are placed to the left of each specimen.

Type-host: Siganus lineatus (Valenciennes), Golden-lined spinefoot (Siganidae).

Type-locality: off Lizard Island (14°40′S, 145°27′E), northern Great Barrier Reef, Australia.

Other hosts: Siganidae – Siganus argenteus (Quoy & Gaimard), Streamlined spinefoot; Siganus corallinus (Valenciennes), Blue-spotted spinefoot; Siganus doliatus Guérin-Méneville, Barred spinefoot; Siganus puellus (Schlegel), Masked spinefoot; Siganus punctatus (Schneider & Forster), Goldspotted spinefoot.

Other localities: off Heron Island (23°27′S, 151°55′E), southern Great Barrier Reef; off Noumea (21°30′S, 165°30′E), New Caledonia.

Host/locality combinations: Lizard Island – S. argenteus, S. corallinus, S. doliatus, S. lineatus, S. puellus and S. punctatus. Heron Island – S. lineatus. New Caledonia – S. doliatus.

Site in host: Intestine.

Deposition of specimens: Holotype, QM G242,739; 17 paratypes, QM G242,740–56; 72 voucher specimens, QM G242,757–828.

Molecular sequence data: cox1 mtDNA, PZ466203–16; ITS2 rDNA, PZ469405–07; 28S rDNA, PZ469393–94.

ZooBank LSID: urn:lsid:zoobank.org:act:93F818E4-46D7-43A7-92B8-8EB7DF12BAA3.

Etymology: This species is named for Dr Misako Urabe in recognition of her major contribution to the field.

Description

[Based on 10 hologenophore and four paragenophore specimens. 64 voucher specimens (no molecular data) examined and identified but not incorporated in main measurements. Measurements for each host/locality combination in Supplementary Table 2. Egg measurements given in Supplementary Table 1.]

Body relatively elongate, with margins ranging from rarely almost parallel to typically clearly widest in middle or posterior third of body, 2,895–3,788 (3,338) × 595–1,025 (770). Length to anterior testis 2,040–3,443 (2,667). Length to anterior testis: body width ratio 3.05–3.94 (3.48):1. Largest gravid specimens over twice length of smallest. Tegument armed with fine spines; spines most clearly visible at anterior extremity. Eye-spot pigment not detected in anterior body. Pharynx typically slightly wider than long, 175–289 (235) × 209–314 (265). Oesophagus well-developed but conspicuously short, 376–541 (450) in length. Oesophagus length: pharynx length ratio 1.55–2.19 (1.93):1. Oesophageal bulb distinct. Intestinal caeca terminate close to anterior margin of anterior testis, highly variable in width.

Testes two, entire, usually contiguous, occasionally partly flanked by intestinal caeca, always oblique with sinistral testis more anterior. Seminal vesicle winding from well posterior to intestinal bifurcation to genital pore, typically largely obscured by uterus. Poorly developed ‘cirrus-sac’ present. Genital pore median, midway between pharynx and intestinal bifurcation.

Ovary entire, usually ventral to and partially overlapping posterior testis. Uterus normally well-developed from about anterior margin of anterior testis to about intestinal bifurcation, in smaller specimens appears to follow helical path but broadly irregular in fully mature specimens, sometimes ventrally overlapping intestinal caeca when full of eggs. Ventral vitelline follicles in two lateral fields, ventral to intestinal caeca, typically extend anteriorly from just anterior to ends of intestinal caeca (occasionally reaching ends) to 215–764 (494) from intestinal bifurcation, rarely almost reaching it; dorsal vitelline follicles medially confluent, dorsal to intestinal caeca and uterus but usually largely obscured by uterus, only rarely appears to include most anterior vitelline follicles. Eggs 69–95 (81) × 43–63 (51) (n = 97, all samples).

Excretory pore dorsally subterminal close to posterior extremity of body. ‘Rosette organ’ represented only by a pair of distinct diverticula that unite and open at excretory pore. Excretory vesicle thin-walled, voluminous, usually occupies space between rosette organ and ovary.

Remarks

This species has been found at the three Pacific sites where suitable hosts, principally belonging to Siganus Clades 3.1 and 3.2, have been examined. Notably, it has not yet been detected at Ningaloo Reef in the total of 23 S. trispilos and S. virgatus examined there. It is partly recognisable morphologically by its vitelline follicles being well-separated from the intestinal bifurcation and by its relatively short oesophagus.

Hexangium kuramochii n. sp. (Figures 4A–H)

Figure 4.

Gravid Hexangium kuramochii n. sp. A–B, Moreton Bay from S. fuscescens. C–F, Lizard Island: C, from S. argenteus; D, from S. fuscescens; E, from S. lineatus; F, from S. spinus. G–H, Ningaloo Reef: G, from S. fuscescens; H, from S. trispilos. Scale bars = 1 mm.

Figure 4. Long description

The multi-panel illustration consists of eight line drawings (A-H) of the trematode Hexangium kuramochii n. sp. Each drawing features a vertical scale bar representing 1 mm.

* Panel A: A small, slender specimen from S. fuscescens in Moreton Bay.

* Panel B: A larger specimen from the same host and location as A, showing a more developed internal structure.

* Panel C: An anterior portion of a specimen from S. argenteus at Lizard Island.

* Panel D: A specimen from S. fuscescens at Lizard Island with a distinct indentation on the right side of the body wall.

* Panel E: A shorter specimen from S. lineatus at Lizard Island.

* Panel F: A specimen from S. spinus at Lizard Island.

* Panel G: A large, robust specimen from S. fuscescens at Ningaloo Reef.

* Panel H: A specimen from S. trispilos at Ningaloo Reef with a constricted mid-body.

Common anatomical features across all specimens include:

* An anterior oral sucker at the top.

* A long esophagus leading to a bifurcated intestine.

* Numerous dark, rounded vitelline follicles distributed along the lateral margins of the mid-body.

* A central, coiled uterus filled with eggs.

* Large, rounded gonads (testes and ovary) located in the posterior third of the body.

Type-host: Siganus fuscescens (Houttuyn), Mottled spinefoot (Siganidae).

Type-locality: Moreton Bay (27°22′S, 153°13′E), southeastern Queensland, Australia.

Other hosts: Acanthuridae – Acanthurus triostegus (Linnaeus), Convict surgeonfish; Prionurus maculatus Ogilby, Yellowspotted sawtail. Siganidae – Siganus argenteus (Quoy & Gaimard), Streamlined spinefoot; Siganus lineatus (Valenciennes), Golden-lined spinefoot; S. spinus (Linnaeus), Little spinefoot; S. trispilos Woodland & Allen, Threeblotched spinefoot.

Other localities: Hell’s Gates, Sunshine Coast, southern Queensland (26°22′S, 153°07′E); off Lizard Island (14°40′S, 145°27′E), northern Great Barrier Reef, Australia; Ningaloo Reef (22°42′S, 113°40′E), Western Australia.

Host/locality combinations: Moreton Bay – A. triostegus, P. maculatus and S. fuscescens. Hell’s Gates – S. fuscescens. Lizard Island – S. argenteus, S. fuscescens, S. lineatus and S. spinus. Ningaloo Reef – S. argenteus, S. fuscescens and S. trispilos.

Site in host: Intestine.

Deposition of specimens: Holotype, QM 242,829; 10 paratypes, QM G242,830–9; 50 voucher specimens, QM G242,840–89; 18 paratypes, WAM 13,808–25.

Molecular sequence data: cox1 mtDNA, PZ466240–69; ITS2 rDNA, PZ469397–404; 28S rDNA, PZ469390–92.

ZooBank LSID: urn:lsid:zoobank.org: act:1BE021AC-7B5D-4A39-BA79-ECBCFC072708.

Etymology: This species is named for Dr Toshiaki Kuramochi in recognition of his significant contribution to this field.

Description

[Based on 24 hologenophore and four paragenophore specimens. 48 voucher specimens (no molecular data) examined and identified but not incorporated in main measurements. Measurements for each host/locality combination in Supplementary Table 2. Egg measurements given in Supplementary Table 1.]

Body relatively elongate, with margins ranging from rarely almost parallel to typically clearly widest in middle or anterior third of body, 3,470–6,839 (4885) × 698–1,288 (928). Length to anterior testis 2,463–5,144 (3,628). Length to anterior testis: body width ratio 3.16–4.45 (3.91):1. Largest gravid specimens over 2.5 times length of smallest. Tegument armed with fine spines; spines most clearly visible at anterior extremity. Eye-spot pigment not detected. Pharynx typically slightly wider than long, 202–364 (284) × 199–385 (307). Oesophagus well-developed but conspicuously short, 433–756 (588) in length. Oesophagus length: pharynx length ratio 1.79–2.65 (2.09):1. Oesophageal bulb distinct. Intestinal caeca terminate close to anterior margin of anterior testis, highly variable in width.

Testes two, entire, contiguous, occasionally partly flanked by intestinal caeca, usually oblique but ranging from almost opposite to tandem. Seminal vesicle winding from well posterior to intestinal bifurcation to genital pore, typically largely obscured by uterus. Poorly developed ‘cirrus-sac’ present. Genital pore median, midway between pharynx and intestinal bifurcation.

Ovary entire, usually ventral to and partially overlapping posterior testis. Uterus normally well-developed (full of eggs) from about anterior margin of anterior testis to about intestinal bifurcation, although sometimes not extensive as far anteriorly; in smaller specimens appears to follow helical path but broadly irregular in fully mature specimens, sometimes partially overlapping intestinal caeca ventrally when full of eggs. Ventral vitelline follicles in two lateral fields, ventral to intestinal caeca, typically extend anteriorly from just anterior to ends of intestinal caeca (occasionally reaching ends) to 269–868 (545) from intestinal bifurcation, rarely almost reaching it; dorsal vitelline follicles medially confluent, dorsal to intestinal caeca and uterus but usually largely obscured by uterus, appears to include most anterior follicles only rarely. Eggs 73–95 (84) × 43–65 (53) (n = 120, all samples).

Excretory pore dorsally subterminal close to posterior extremity of body. ‘Rosette organ’ represented only by pair of distinct diverticula that unite and open at excretory pore. Excretory vesicle thin-walled, voluminous, usually occupies space between rosette organ and ovary.

Remarks

This species has a seemingly disjunct distribution, known from southern Queensland and Ningaloo Reef, but not from suitable hosts at Heron Island on the Great Barrier Reef.

Hexangium robustum n. sp. (Figures 5A–I)

Figure 5.

Gravid Hexangium robustum n. sp. A–C, Lizard Island: A, from S. vulpinus (hologenophore); B–C, from S. punctatus (Hologenophores). D–I, Heron Island: D, from S. corallinus; E–F, from S. doliatus; G, from S. lineatus; H, from S. puellus; I, from S. vulpinus. Scale bars = 1 mm.

Figure 5. Long description

A multi-panel set of nine line drawings labeled A through I. Each drawing depicts the elongated, spindle-shaped body of a gravid fluke.

Anatomical features common to all specimens from anterior to posterior include:

* A small oral sucker at the very top tip.

* A long, thin esophagus leading to a bifurcated intestine that runs down the lateral sides of the body.

* Numerous dark, solid black circular vitelline follicles arranged in two longitudinal rows along the intestinal caeca.

* A central uterus containing coiled structures.

* Two large, stippled oval testes located in the posterior third of the body, usually arranged in a tandem or slightly diagonal orientation.

* A small, circular ovary with distinct internal spotting located just posterior to the testes.

* A small excretory vesicle at the posterior-most tip.

Variations across panels:

* Panels A, B, and C show specimens from Lizard Island. B and C are noticeably wider than A.

* Panels D through I show specimens from Heron Island.

* Panel H is the largest specimen, showing a much higher density and larger size of black vitelline follicles and more prominent posterior organs.

* Panel I is the smallest and most slender specimen.

* Vertical scale bars representing 1 mm are placed to the left of each specimen to indicate relative size differences.

Type-host: Siganus vulpinus (Schlegel & Müller), Foxface (Siganidae).

Type-locality: off Lizard Island (14°40′S, 145°27′E), northern Great Barrier Reef, Australia.

Other hosts: Labridae – Chlorurus microrhinos (Bleeker), Steephead parrotfish. Siganidae – Siganus doliatus Guérin-Méneville, Barred spinefoot; Siganus lineatus (Valenciennes), Golden-lined spinefoot; Siganus puellus (Schlegel), Masked spinefoot; Siganus punctatus (Bloch & Schneider), Goldspotted spinefoot; Siganus punctatissimus Fowler & Bean, Peppered spinefoot.

Other localities: Yonge Reef (14°35′S, 145°37′E), northern Great Barrier Reef; off Heron Island (23°27′S, 151°55′E); southern Great Barrier Reef, Australia.

Host/locality combinations: Heron Island – C. microrhinos, S. doliatus, S. lineatus, S. puellus, S. punctatus and S. vulpinus. Lizard Island – S. puellus, S. punctatissimus and S. vulpinus. Yonge Reef – S. punctatus.

Site in host: Intestine.

Deposition of specimens: Holotype, QM G242,890; 12 paratypes, QM G242,891–902; QM G 242,903–62.

Molecular sequence data: cox1 mtDNA, PZ466270–74; ITS2 rDNA, PZ469419–21; 28S rDNA, PZ469395.

ZooBank LSID: urn:lsid:zoobank.org:act:8EF3D6AA-F56D-40E7-A2B2-22530B475330.

Etymology: This species is named for its distinctly robust form.

Description

[Based on six hologenophore specimens and five whole specimens (all from the type-host/type-locality combination). 68 voucher specimens (no molecular data) examined and identified but not incorporated in main measurements. Measurements for each host/locality combination in Supplementary Table 2. Egg measurements given in Supplementary Table 1.]

Body relatively plump, rarely with almost parallel margins, in larger specimens frequently conspicuously widest at level of testes, 2,400–2,976 (2,658) × 612–1,267 (822). Length to anterior testis 1,716–2,700 (2,093). Length to anterior testis: body width ratio 2.04–3.18 (2.60):1. Largest gravid specimens over 3 times length of smallest. Tegument armed with fine spines; spines most clearly visible at anterior extremity. Eye-spot pigment not detected. Pharynx typically slightly wider than long, 138–228 (181) × 156–273 (210). Oesophagus well-developed, relatively long, 457–718 (603) in length. Oesophagus length: pharynx length ratio 3.05–4.04 (3.37):1. Oesophageal bulb distinct. Intestinal caeca terminate close to anterior margin of anterior testis, highly variable in width.

Testes two, entire, typically contiguous, occasionally partly flanked by intestinal caeca, usually oblique but also often nearly opposite. Seminal vesicle winding from well posterior to intestinal bifurcation to genital pore, typically largely obscured by uterus. Poorly developed ‘cirrus-sac’ present. Genital pore median, midway between pharynx and intestinal bifurcation.

Ovary entire, usually ventral to and partially overlapping posterior testis. Uterus normally well-developed from about anterior margin of anterior testis to about intestinal bifurcation, in smaller specimens appears to follow helical path but broadly irregular in fully mature specimens, sometimes partially overlapping intestinal caeca ventrally when full of eggs. Ventral vitelline follicles in two lateral fields, ventral to intestinal caeca, typically extend anteriorly from just anterior to ends of intestinal caeca (occasionally reaching ends) to as far as 39 anterior to or 247 posterior to intestinal bifurcation it (mean 76 posterior); dorsal vitelline follicles medially confluent, dorsal to intestinal caeca and uterus but usually largely obscured by uterus, never detected to include most anterior vitelline follicles. Eggs 65–97 (82) × 41–59 (51) (n = 63, all samples).

Excretory pore dorsally subterminal close to posterior extremity of body. ‘Rosette organ’ represented only by a pair of distinct diverticula that unite and open at excretory pore. Excretory vesicle thin-walled, voluminous, usually occupies space between rosette organ and ovary.

Remarks

This species has a distinctive host-specificity, being concentrated in Clade 3.1 siganids; no other species is found as commonly in these fishes. It is also relatively morphologically distinct in that the vitelline follicles normally extend to close to the intestinal bifurcation.

Hexangium reimeri n. sp. (Figures 6C–G)

Figure 6.

Hexangium reimeri n. sp. from Siganus fuscescens. A–D, Heron Island. E, Ningaloo Reef. Scale bars = 1 mm.

Figure 6. Long description

A multi-panel line drawing displays five specimens of the parasite Hexangium reimeri n. sp. Panels A, B, C, and D represent specimens from Heron Island, while panel E represents a specimen from Ningaloo Reef. Each specimen is elongated and cylindrical with a distinct oral sucker at the anterior tip.

* Specimen A is a large, mature individual. Below the oral sucker, a long esophagus leads to a bifurcated intestine. Large, black, circular vitelline follicles are densely packed along the lateral margins of the mid-body. Two large, stippled testes are located in the posterior third, followed by a smaller, spotted ovary and a terminal excretory pore.

* Specimen B is a significantly smaller, narrower juvenile form. It contains fewer vitelline follicles and less developed reproductive organs. A vertical scale bar is positioned to its left.

* Specimen C is an elongated mature form similar to A but with a more slender body. The vitelline follicles are more spaced out along the sides. A vertical scale bar is at the bottom left.

* Specimen D shows the anterior and mid-section of a specimen, focusing on the oral sucker, esophagus, and the beginning of the vitelline follicle rows. A vertical scale bar is at the bottom left.

* Specimen E is a large mature individual nearly identical in structure to A, showing the full complement of internal organs including the oral sucker, esophagus, lateral vitelline follicles, posterior testes, and ovary. A vertical scale bar is at the bottom left.

All scale bars represent 1 mm.

Type-host: Siganus fuscescens (Houttuyn), Mottled spinefoot (Siganidae).

Type-locality: off Heron Island (23°27′S, 151°55′E), southern Great Barrier Reef, Australia.

Other localities: Ningaloo Reef (22°42′S, 113°40′E), Western Australia.

Site in host: Intestine.

Deposition of specimens: Holotype, QM G242,963; 12 paratypes QM G242,964–75; 60 voucher specimens, QM G242,976–3,020; 1 paratype, WAM 13,826.

Molecular sequence data: cox1 mtDNA, PZ466275–79; ITS2 rDNA, PZ469417–18; 28S rDNA, PZ469396.

ZooBank LSID: urn:lsid:zoobank.org:act:FC893E09-947D-4D0F-8016-176D9BF4171F.

Etymology: This species is named for Dr James Reimer in recognition of his help in this study and his significant contribution to the understanding of Indo-Pacific biodiversity.

Description

[Based on two hologenophore specimens and 39 whole specimens (no molecular data) from the type-host/type-locality combination. 30 voucher specimens (no molecular data) examined and identified but not incorporated in main measurements. Measurements for each host/locality combination in Supplementary Table 2. Egg measurements given in Supplementary Table 1.]

Body relatively elongate, typically with almost parallel margins, widest point highly variable, 2,762–6,472 (3856) × 441–1,464 (696). Length to anterior testis 2,036–4,747 (2,842). Length to anterior testis; body width ratio 3.24–4.93 (4.16):1. Largest gravid specimen approximately 2.5 times length of smallest. Tegument armed with fine spines; spines most clearly visible at anterior extremity. Eye-spot pigment not detected in anterior body. Pharynx typically slightly wider than long, 154–365 (212) × 172–377 (236). Oesophagus well-developed, relatively elongate, 495–943 (660) in length. Oesophagus length: pharynx length ratio 2.37–3.85 (3.13):1. Oesophageal bulb distinct. Intestinal caeca terminate close to anterior margin of anterior testis, highly variable in width.

Testes entire, contiguous, occasionally partly flanked by intestinal caeca, usually oblique but ranging from almost opposite to directly tandem. Seminal vesicle winding from well posterior to intestinal bifurcation to genital pore, typically largely obscured by uterus. Poorly developed ‘cirrus-sac’ present. Genital pore median, midway between pharynx and intestinal bifurcation.

Ovary entire, usually ventral to and partially overlapping posterior testis. Uterus normally well-developed from about anterior margin of anterior testis to about intestinal bifurcation, in smaller specimens appears to follow helical path but broadly irregular in fully mature specimens, sometimes partially overlapping intestinal caeca ventrally when full of eggs. Ventral vitelline follicles in two lateral fields, ventral to intestinal caeca, typically extend anteriorly from just anterior to ends of intestinal caeca (occasionally reaching ends) to 111–476 (249) from intestinal bifurcation, rarely almost reaching it; dorsal vitelline follicles medially confluent, dorsal to intestinal caeca and uterus but usually largely obscured by uterus, never detected to include most anterior vitelline follicles. Eggs 71–89 (81) × 38–61 (49) (n = 70, all samples).

Excretory pore dorsally subterminal close to posterior extremity of body. ‘Rosette organ’ represented only by a pair of distinct diverticula that unite and open at excretory pore. Excretory vesicle thin-walled, voluminous, usually occupies space between ‘rosette organ’ and ovary.

Remarks

This species is morphologically relatively unspecialised and is difficult to distinguish from H. affinum with which it shares a relatively long oesophagus. It is, however, seemingly distinct from H. affinum in its host distribution.

Molecular analyses

Sequences for cox1 (n = 113) distinguish eight strongly supported clades that differ from each other at p-distances of 3.79–16.00 (18–76 bp) (Figure 7, Table 2). Each of the species recognised above relates to one clade except for H. affinum, which has three subclades. Sequence differences within each of the eight clades range from 0 to 9 bp with no or negligible host or locality correlation; these differences are interpreted as (almost entirely) uninformative intraspecific variation. The exception with minor structuring is H. urabeae, for which 18 of 19 sequences from Lizard Island differ from each other at just 0–4 bp and at 5–6 bp from the five sequences from Heron Island; this distinction is interpreted as representing weak intraspecific geographic variation. The level of distinction in p-distance between the eight clades ranges from 3.79–16.00 (18–76 bp) (Table 2). The three most similar clades involve p-distance differences of 3.79–5.47 (18–26 bp) and ultimately form the highly supported clade interpreted as relating to H. affinum.

Figure 7.

Phylogram from the unrooted neighbour-joining analysis of the cox1 mtDNA dataset. Bootstrap support values are shown at the nodes, with values of <85 not shown. The scale bar indicates the number of base differences.

Figure 7. Long description

The phylogram branches from a single root on the left into six distinct major clades. Each clade is labeled on the far right with the species name.

* The top clade is Hexangium urabeae, containing numerous specimens from Siganus lineatus and other hosts at Lizard Island and Heron Island.

* Below it is a smaller clade for H. sigani, primarily from Siganus fuscescens in Okinawa, with a bootstrap support value of 100.

* The third clade is Hexangium affinum, showing specimens from various Siganus species across Lizard Island, Heron Island, and Ningaloo Reef, with internal bootstrap values of 98 and 99.

* The fourth and largest clade is Hexangium kuramochii, containing a high density of specimens from Ningaloo Reef, Lizard Island, and Moreton Bay.

* The fifth clade is H. robustum, with a bootstrap value of 100, including specimens from Yonge Reef and Lizard Island.

* The bottom-most clade is H. reimeri, also with a bootstrap value of 100, featuring specimens from Heron Island and Ningaloo Reef.

A scale bar at the bottom left represents 10 base differences. Bootstrap support values are indicated at major nodes, including 98 at the primary divergence point for the top three clades.

Table 2.

Genetic differences between eight Hexangium cox1 clades. p-distances are shown above the diagonal, base pair differences below the diagonal, and intra-clade differences on the diagonal

Table 2. Long description

The table compares eight clades of Hexangium. The diagonal shows intra-clade differences. Values above the diagonal represent p-distances (percentages), and values below the diagonal represent base pair differences.

* Clade 1: H. urabeae (n=24). Intra-clade: 0–7. p-distances to clades 2-8 range from 4.84 to 14.32.

* Clade 2: H. sigani (n=7). Intra-clade: 0–1. Base pair diff to clade 1: 23–27. p-distances to clades 3-8 range from 5.68 to 14.95.

* Clade 3: H. affinum subclade 1 (n=12). Intra-clade: 0–9. Base pair diff to clades 1-2: 32–40. p-distances to clades 4-8 range from 3.79 to 14.74.

* Clade 4: H. affinum subclade 2 (n=7). Intra-clade: 0–5. Base pair diff to clades 1-3: 21–34. p-distances to clades 5-8 range from 3.79 to 15.16.

* Clade 5: H. affinum subclade 3 (n=7). Intra-clade: 0–5. Base pair diff to clades 1-4: 18–38. p-distances to clades 6-8 range from 11.16 to 16.00.

* Clade 6: H. kuramochii (n=43). Intra-clade: 0–8. Base pair diff to clades 1-5: 45–61. p-distances to clades 7-8 range from 12.63 to 15.16.

* Clade 7: H. robustum (n=7). Intra-clade: 0. Base pair diff to clades 1-6: 60–76. p-distance to clade 8: 15.16–15.37.

* Clade 8: H. reimeri (n=7). Intra-clade: 0–4. Base pair diff to clades 1-7: 61–73.

Sequences for ITS2 rDNA (n = 48) show a low overall level of distinction between samples; the maximum p-distance was 2.89 (11 bp) (Figure 8, Table 3). The distinctions suggested by these data are generally compatible with those from cox1 sequences with two exceptions. First, sequences corresponding to H. urabeae again form two clades (relating to Lizard and Heron Islands and differing at 4 bp) which do not, however, themselves form a clade. Second, ITS2 sequences corresponding to cox1 sequences for H. kuramochii show a consistent 1 bp difference between samples from Moreton Bay and those from Lizard Island and Ningaloo. Differences between corresponding cox1 sequences from samples from the two areas are negligible. Hexangium sigani differs consistently from all other species in ITS2, but by as few as 2 bp relative to H. urabeae and H. kuramochii. ITS2 sequences relating to the three cox1 variants of H. affinum form a clade differing from all other species at 3–11 bp and differ from each other at only 1–2 bp. Hexangium robustum differs at 7–11 bp from all other species. Hexangium reimeri differs at 3–9 bp from all other species. For H. sigani and H. robustum, there was no intraspecific variation. Hexangium affinum, H. kuramochii, H. reimeri and H. urabeae each showed low-level geographic variation.

Figure 8.

Phylogram from the unrooted neighbour-joining analysis of the ITS2 rDNA dataset. Bootstrap support values are shown at the nodes, with values of <85 not shown. The scale bar indicates the number of base differences.

Figure 8. Long description

The phylogram originates from a single root on the left, branching into several distinct clades.

* The uppermost clade contains Hexangium kuramochii specimens from Moreton Bay, Lizard Island, and Ningaloo Reef, primarily hosted by Siganus fuscescens and Siganus argenteus.

* Below this is a smaller group of Hexangium urabeae from Lizard Island.

* The middle section features a large clade of Hexangium sigani from Okinawa, followed by a diverse clade of Hexangium affinum from Ningaloo Reef, Lizard Island, and Heron Island.

* The lower section includes a separate cluster of Hexangium urabeae from Heron Island (bootstrap value 98), followed by Hexangium reimeri from Ningaloo Reef and Heron Island (bootstrap value 90 and 91).

* The bottom-most clade consists of Hexangium robustum from Yonge Reef, Lizard Island, and Heron Island (bootstrap value 99).

Terminal labels include the species name, host species, location, and accession number. A scale bar at the bottom left represents 1 base difference.

Table 3.

Genetic differences between six Hexangium species for ITS2 rDNA. p-distances are shown above the diagonal, base pair differences below the diagonal, and intra-clade differences on the diagonal. Two species (H. kuramochii and H. urabeae) are subdivided by locality. HI – Heron Island; LI – Lizard Island; MB – Moreton Bay; NR – Ningaloo Reef

Table 3. Long description

The table consists of 8 rows and 8 numbered columns representing species groups.

Row 1: H. urabeae L I. n equals 5. Diagonal: 0. P-distances to others: 0.87, 0.27 to 0.43, 0.43 to 0.65, 0.22 to 0.43, 0.22 to 0.43, 1.52 to 1.74, 0.65 to 1.30.

Row 2: H. urabeae H I. n equals 3. Base pair difference to group 1: 4. Diagonal: 0. P-distances to others: 1.09, 1.30 to 1.52, 1.09, 0.87 to 1.09, 2.39, 1.52 to 1.96.

Row 3: H. sigani. n equals 7. Base pair differences to groups 1 to 2: 1 to 2, 5. Diagonal: 0. P-distances to others: 0.65 to 1.09, 0.87, 0.43 to 0.65, 1.96, 1.09 to 1.52.

Row 4: H. affinum. n equals 9. Base pair differences to groups 1 to 3: 2 to 3, 6 to 7, 3 to 5. Diagonal: 0 to 2. P-distances to others: 0.87 to 1.09, 0.65 to 1.09, 2.17 to 2.39, 1.30 to 1.96.

Row 5: H. kuramochii M B. n equals 7. Base pair differences to groups 1 to 4: 1 to 2, 5, 4, 4 to 5. Diagonal: 0. P-distances to others: 0.22 to 0.43, 2.17, 1.30 to 1.74.

Row 6: H. kuramochii L I or N R. n equals 9. Base pair differences to groups 1 to 5: 1 to 2, 4 to 5, 2 to 3, 3 to 5, 1 to 2. Diagonal: 0 to 1. P-distances to others: 1.74 to 1.96, 0.87 to 1.52.

Row 7: H. robustum. n equals 5. Base pair differences to groups 1 to 6: 7 to 8, 11, 9, 10 to 11, 10, 8 to 9. Diagonal: 0. P-distance to group 8: 1.74 to 2.17.

Row 8: H. reimeri. n equals 3. Base pair differences to groups 1 to 7: 3 to 6, 7 to 9, 5 to 7, 6 to 9, 6 to 10, 4 to 7, 8 to 10. Diagonal: 0 to 2.

Sequences for 28S rDNA (n = 13) were generated for the eight clades recognised from cox1 sequences (Figure 9, Table 4). All combinations of clades recognised from cox1 data were distinguished by this marker except for two of those ultimately identified as H. affinum, which were identical for this marker. Notably, samples of H. urabeae from Heron and Lizard Islands form a clade (as for cox1), whereas they do not in the ITS2 analysis. Of the four species that show inferred intraspecific ITS2 variation (H. affinum, H. kuramochii, H. reimeri and H. urabeae), all but H. reimeri (sequenced only from Heron Island) varied at a single bp for 28S. All other combinations differed at p-distances of 0.30–1.20 (3–16 bp). Hexangium robustum and H. reimeri are by far the most distantly related species, differing from all other species at p-distances of 0.90–1.20 (12–16 bp).

Figure 9.

Phylogram from the unrooted neighbour-joining analysis of the 28S rDNA dataset. Bootstrap support values are shown at the nodes, with values of <85 not shown. The scale bar indicates the number of base differences.

Figure 9. Long description

A neighbor-joining phylogram with a scale bar at the bottom left representing 1 base difference. The tree originates from a single root on the left, splitting into two primary lineages.

Top Lineage:

- A major node with a bootstrap value of 99 leads to several sub-clades.

- The uppermost sub-clade contains five entries of Hexangium affinum from Siganus argenteus, Siganus virgatus, Siganus corallinus, and Siganus doliatus across Ningaloo Reef, Lizard Island, and Heron Island.

- Below this is a single branch for Hexangium sigani from Siganus fuscescens in Okinawa.

- A node with a bootstrap value of 85 leads to a clade of three Hexangium kuramochii samples from Prionurus maculatus and Siganus fuscescens in Moreton Bay and Ningaloo Reef.

- A node with a bootstrap value of 96 leads to two Hexangium urabeae samples from Siganus lineatus at Heron Island and Lizard Island.

Bottom Lineage:

- This lineage splits into two distinct branches.

- The upper branch leads to Hexangium robustum from Siganus punctatus at Yonge Reef.

- The lower branch, supported by a bootstrap value of 99, leads to two Hexangium reimeri samples from Siganus fuscescens at Heron Island.

Table 4.

Genetic differences between six Hexangium species for 28S rDNA. p-distances above the diagonal, base pair differences below the diagonal, and intra-clade differences on the diagonal. Two species (H. kuramochii and H. urabeae) are subdivided by locality. HI – Heron Island; LI – Lizard Island; MB – Moreton Bay; NR – Ningaloo Reef

Table 4. Long description

A matrix table comparing eight groups across six Hexangium species. The columns are numbered 1 through 8, corresponding to the rows: 1. H. urabeae (L I), 2. H. urabeae (H I), 3. H. sigani, 4. H. affinum, 5. H. kuramochii (M B), 6. H. kuramochii (L I, N R), 7. H. robustum, and 8. H. reimeri.

Values above the diagonal represent p-distances. Notable values include 0.07 between H. urabeae L I and H I, and higher distances reaching 1.20 between H. robustum and several other species.

Values below the diagonal represent base pair differences. For example, there is 1 base pair difference between the two H. urabeae localities, while H. robustum (row 7) shows 14 to 16 differences compared to species 1 through 6.

The diagonal represents intra-clade differences, with H. affinum and H. kuramochii (M B) showing a value of 0.

Sample sizes (n) for each row are: 1, 1, 1, 5, 2, 1, 1, 1 respectively. Locality abbreviations are L I for Lizard Island, H I for Heron Island, M B for Moreton Bay, and N R for Ningaloo Reef.

Morphology

The six species recognised above are morphologically broadly similar. Potential distinctions were sought by the examination, drawing and measurement of hologenophores, paragenophores and corresponding (mainly same host and same locality) intact specimens. Some specimens were identified despite representing host/locality combinations for which no sequences were obtained. Allocations from host/locality combinations for which molecular data indicate the presence of more than one species were challenging. We found two or three Hexangium species in nine of the 12 Siganus species relating to 12 host/locality combinations. Our case-by-case approach to delineation is detailed in Table 5. For none of the species was a single morphological character completely reliably diagnostic. However, combinations of several characters, in the light of molecular distinction and host distribution, support variably clear distinctions between the putative species. We recognise the following progressive but imperfect morphological and biological distinction between what is interpreted as six species.

Table 5.

Basis of distinction of multiple species of Hexangium from individual host/locality combinations. V – vouchers

Table 5. Long description

The table consists of four columns: Host, Locality, Species and material, and Voucher identification.

* S. argenteus at Lizard Island: Includes H. urabeae (1 c o x 1), H. affinum (4 c o x 1, 9 V), and H. kuramochii (5 c o x 1). Identified by vitelline follicle distribution and oesophagus to pharynx length ratio.

* S. corallinus at Lizard Island: Includes H. urabeae (1 c o x 1, 12 V) and H. affinum (3 c o x 1, 4 V). Identified by oesophagus to pharynx length ratio.

* S. doliatus at Lizard Island: Includes H. urabeae (4 c o x 1, 16 V) and H. affinum (9 c o x 1, 28 V). Identified by oesophagus to pharynx length ratio.

* S. doliatus at New Caledonia: Includes H. urabeae (2 V) and H. affinum (1 V). Identified by oesophagus to pharynx length ratio.

* S. fuscescens at Lizard Island: Includes H. sigani (1 c o x 1) and H. kuramochii (1 c o x 1, 8 V). Identified by rarity of H. sigani versus abundance of H. kuramochii.

* S. fuscescens at Ningaloo Reef: Includes H. kuramochii (18 c o x 1) and H. reimeri (1 c o x 1). No unsequenced vouchers.

* S lineatus at Lizard Island: Includes H. urabeae (12 c o x 1, 26 V) and H. kuramochii (1 c o x 1). Identified by probability from c o x 1 sequencing.

* S. puellus at Lizard Island: Includes H. urabeae (3 V), H. affinum (9 V), and H. robustum (1 c o x 1). Identified by vitelline follicle distribution and oesophagus to pharynx length ratio.

* S. punctatissimus at Lizard Island: Includes H. affinum (1 c o x 1, 7 V) and H. robustum (2 c o x 1). Identified by vitelline follicle distribution.

* S. punctatus at Heron Island: Includes H. affinum (7 V) and H. robustum (16 V). Identified by vitelline follicle distribution.

* S. punctatus at Lizard Island: Includes H. urabeae (1 c o x 1), H. affinum (15 V), and H. robustum (2 c o x 1). Identified by vitelline follicle distribution and oesophagus to pharynx length ratio.

* S. trispilos at Ningaloo Reef: Includes H. affinum (1 c o x 1, 2 V) and H. kuramochii (3 c o x 1). Identified by oesophagus to pharynx length ratio.

Hexangium sigani is recognised based on multiple sequences from S. fuscescens from Japan (from where we have one macerated hologenophore together with several intact but poor specimens). These sequences thus relate to the type-host and are close to the type-locality of H. sigani, and the specimens are broadly morphologically consistent with the original description. Measurements of specimens of H. sigani from the original descriptions of Goto and Ozaki (Reference Goto and Ozaki1929) and others from S. fuscescens from Japanese waters reported by Yamaguti (Reference Yamaguti1934) and Machida et al. (Reference Machida, Ichihara and Kamegai1970) are large; many are substantially longer than any of those of the other five species recognised here (Figure 10). The 25–75% quartiles of the length distribution of available H. sigani specimens do not overlap with those of any other species. Figure 11 shows a large (often the largest available) specimen of each of the six species, all at the same scale; the size distinction between H. sigani and the other species is striking. Although size is not informative for the identification of smaller specimens of H. sigani, this size skewing appears to support the recognition of H. sigani as both morphologically and genetically distinct from the other species in our collection. As originally described, H. sigani differs from H. robustum in having the vitelline follicles well-separated from the oesophageal bulb, and from H. affinum and H. reimeri by its relatively short oesophagus. These differences are also seen in the single available (but macerated) hologenophore specimen from Japan. Except by maximum size, H. sigani is not presently morphologically distinguishable from H. kuramochii and H. urabeae.

Figure 10.

Box and whisker plots of body lengths (in μm) of gravid specimens of six Hexangium species: 1, H. urabeae; 2, H. sigani; 3, H. affinum; 4, H. kuramochii; 5, H. robustum; 6, H. reimeri.

Figure 10. Long description

The Y axis represents body length in mu m, ranging from 1,000 to 11,000 in increments of 1,000. The X axis lists six species numbered 1 to 6. Each entry consists of a box plot showing the median, quartiles, and whiskers, with individual data points plotted vertically along the center line.

* Species 1 H. urabeae: Median is approximately 2,800. The box spans from 2,200 to 3,400. Whiskers reach from 1,200 to 4,500.

* Species 2 H. sigani: This species is the largest. Median is approximately 7,200. The box spans from 6,000 to 8,100. Whiskers reach from 3,900 to 10,300.

* Species 3 H. affinum: Median is approximately 2,600. The box spans from 2,300 to 3,500. Whiskers reach from 1,100 to 5,000.

* Species 4 H. kuramochii: Median is approximately 3,900. The box spans from 2,600 to 5,000. Whiskers reach from 1,500 to 6,900.

* Species 5 H. robustum: Median is approximately 2,700. The box spans from 2,100 to 3,200. Whiskers reach from 1,000 to 4,700.

* Species 6 H. reimeri: Median is approximately 3,200. The box spans from 2,700 to 4,000. Whiskers reach from 1,300 to 6,000, with one outlier near 6,500.

Figure 11.

Examples of largest or near-largest representatives of six putative Hexangium species. A, H. urabeae from S. corallinus, Lizard Island. B, Presumed holotype H. sigani from S. fuscescens, Japan. C, H. affinum from S. puellus, Lizard Island. D, H. kuramochii from S. fuscescens, Ningaloo Reef. E, H. robustum from S. puellus, Heron Island; F, H. reimeri from S. fuscescens, Ningaloo Reef. Scale bars = 1 mm.

Figure 11. Long description

A series of six anatomical line drawings labeled A through F, each representing a different Hexangium species. Each specimen is elongated and leaf-shaped with a small circular oral sucker at the anterior tip.

* Panel A shows H. urabeae, the largest specimen, with dense black vitelline follicles lining the lateral margins and two large stippled testes in the posterior third.

* Panel B shows H. sigani, a smaller specimen with fewer, more scattered follicles.

* Panel C shows H. affinum, characterized by a slender body and small, concentrated follicles.

* Panel D shows H. kuramochii, featuring a medium-sized body with follicles extending from the mid-body to the posterior.

* Panel E shows H. robustum, with a wider mid-section and large, prominent testes.

* Panel F shows H. reimeri, displaying large, dark follicles that are more circular in shape compared to the other specimens.

Common internal structures across all panels include a long, winding esophagus leading to a central intestinal cecum, and a small, stippled ovary located just posterior to the testes. At the very bottom of each specimen is a small Y-shaped excretory pore. Vertical scale bars representing 1 mm are positioned to the bottom-left of each drawing.

Hexangium robustum is distinguishable from the other five species in having the vitelline follicles reach close to the posterior margin of the oesophageal bulb, rarely slightly exceeding it anteriorly (Figure 12A); only one specimen of H. robustum (a hologenophore from S. punctatus, Lizard Island) is inconsistent with this measure. One specimen from S. corallinus (Heron Island), seven specimens from S. doliatus (Lizard Island, two individual fish), and two specimens from S. lineatus (one Heron Island, one Lizard Island), are morphologically identified as this species based largely on this character. These fishes belong to Siganus Clade 3.2 and are thus somewhat distant from the other identified hosts (S. puellus, S. punctatus, S. punctatissimus and S. vulpinus), which all belong to Clade 3.1 and harboured most of the specimens identified as this species. Unfortunately, no relevant material is available for the molecular testing of the identity of the S. corallinus, S. doliatus and S. lineatus samples. Specimens of H. affinum and H. kuramochii have small numbers of specimens in which the vitelline follicles approach relatively close to the oesophageal bulb. One specimen identified as H. affinum is exceptional in this respect, but the length of the oesophagus and the body shape suggest identity as H. affinum. Similarly, two specimens from S. fuscescens from off Noosa (southern Queensland) have vitelline follicles reaching close to the oesophageal bulb. In this case, the host (S. fuscescens) and the shortness of the oesophagus both suggest identity as H. kuramochii. The anterior distribution of the vitellarium for specimens of H. reimeri is adjacent to that for H. robustum. In large specimens, the body shape of H. robustum is typically distinctly more rounded than for all the other Hexangium species (Figure 12B), but this feature is not completely reliable and is not pronounced in smaller specimens. Hexangium robustum differs marginally from H. affinum by its relatively larger pharynx, and from H. kuramochii, H, reimeri and H. urabeae, by its proportionally more elongate oesophagus. None of these latter distinctions are completely reliable. Overall, although there are some issues with character overlap, H. robustum is the most distinctive of the six Hexangium species considered here.

Figure 12.

Morphometric separation of five putative Hexangium species. A, Vitelline follicles separation from oesophageal bulb vs body length to anterior testis. B, Body width vs body length to anterior testis. C, Oesophagus length: pharynx length ratio vs body length to anterior testis (H. urabeae, H. affinum, H. kuramochii, and H. reimeri). D, Vitelline follicles separation from oesophageal bulb vs body length to anterior testis (H. affinum and H. reimeri). E, Oesophagus length: pharynx length ratio vs body length (H. affinum and H. reimeri). F, Vitelline follicles separation from oesophageal bulb vs body length to anterior testis (H. urabeae and H. kuramochii). H. urabeae +; H. affinum Δ; H. kuramochii □; H. robustum ♦; H. reimeri.

Figure 12. Long description

A six-panel scatter plot series labeled A through F.

Panel A: X-axis ranges from 1000 to 5500. Y-axis ranges from 0 to 1000. It shows vitelline follicles separation from oesophageal bulb versus body length to anterior testis. H. robustum (solid diamonds) clusters at the bottom left. H. kuramochii (open squares) is widely dispersed at the top right. H. urabeae (plus signs), H. affinum (open triangles), and H. reimeri (open circles) overlap in the center.

Panel B: X-axis 1000 to 5500. Y-axis 250 to 1450. Body width versus body length to anterior testis. H. robustum shows a steep linear increase. Other species show a more gradual linear trend with H. kuramochii at the highest body length values.

Panel C: X-axis 1000 to 5500. Y-axis 2.00 to 6.00. Oesophagus length to pharynx length ratio versus body length to anterior testis. H. affinum (triangles) occupies the highest ratio values (3.0 to 5.5), while H. urabeae (plus signs) and H. kuramochii (squares) cluster at the bottom (2.0 to 3.0).

Panel D: X-axis 1000 to 5500. Y-axis 0 to 1000. Focuses on H. affinum (triangles) and H. reimeri (circles). H. affinum generally shows higher vitelline follicle separation values than H. reimeri across similar body lengths.

Panel E: X-axis 1000 to 5000. Y-axis 2.00 to 6.00. Oesophagus length to pharynx length ratio versus body length for H. affinum and H. reimeri. H. affinum maintains a higher ratio (3.0 to 5.5) compared to H. reimeri (2.5 to 3.5).

Panel F: X-axis 1000 to 5500. Y-axis 0 to 1000. Vitelline follicles separation versus body length for H. urabeae (plus signs) and H. kuramochii (circles). H. urabeae shows higher separation values (200 to 800) compared to H. kuramochii (100 to 450).

Comparison of the ratio of oesophagus length to pharynx length for specimens of H. affinum, H. kuramochii, H. reimeri and H. urabeae (Figure 12C) generates a compelling case for distinguishing two groups of species. In H. urabeae and H. kuramochii the ratios are relatively low at 1.55–2.90:1 and 1.74–2.76:1, respectively, whereas in H. affinum and H. reimeri they are substantially higher, at 2.87–5.27:1 and 2.37–3.85:1, respectively; there is only marginal overlap between the two pairs of species. This distinction is driven mainly by the length of the oesophagus so that H. urabeae and H. kuramochii are ‘short oesophagus’ forms and H. affinum and H. reimeri are ‘long oesophagus’ forms. Hexangium robustum also has a relatively high ratio range (2.53–4.04:1) but is distinguished from H. affinum and H. reimeri by the distribution of its vitelline follicles.

Hexangium affinum and H. reimeri (‘long oesophagus’ forms) have completely non-overlapping host distributions. Specimens of H. affinum have been sequenced from S. argenteus (n = 6 cox1 sequences), S. corallinus (n = 4), S. doliatus (n = 10), S. punctatissimus (n = 1), S. trispilos (n = 1) and S. virgatus (n = 4). Morphology-only voucher specimens have also been identified from S. puellus, S. punctatus and Zanclus cornutus. In contrast, specimens of H. reimeri have been sequenced only from S. fuscescens (n = 7), and no other hosts are suspected from voucher specimens. Hexangium affinum has been found at Heron Island, Lizard Island and Ningaloo Reef, whereas H. reimeri has been predominantly found at Heron Island and is represented by a single hologenophore from Ningaloo Reef. Morphologically, these two forms are partly distinguished by how closely the vitelline follicles approach the posterior margin of the oesophageal bulb (Figure 12D), the oesophagus to pharynx length ratio (Figure 12E), and the pharynx length itself. None of these distinctions is absolute; multiple specimens overlap in their ranges for all three characters. As such, these distinctions serve to suggest a biological distinction between the populations of these two species, but individuals of the two may be indistinguishable by morphology.

Samples identified as H. affinum form three clearly distinct cox1 clades and show some division in ITS2 and 28S sequences. Samples were found from 12 host/locality combinations, more than for any other species. Overall, specimens of this putative species are moderately distinctive, especially in their long oesophagus and strong separation of the vitellarium from the oesophageal bulb, and partly by their relatively narrow bodies. They are surprisingly variable in the ratio of oesophagus length to pharynx length (2.87–5.27:1), including between specimens of similar overall size from the same host and locality. The large range in this ratio arises from variability in the length of both the pharynx and the oesophagus. Variation of this character does not align with the three cox1 clades; we found no morphological variation that correlated with the three clades. All three subtypes are concentrated in siganids of Clade 3.1.

Hexangium urabeae and H. kuramochii (‘short oesophagus’ species) co-occur at Lizard Island; both are also found elsewhere but so far not together. They have been identified in largely non-overlapping ranges of hosts. Specimens of H. urabeae have been sequenced from S. argenteus (n = 1 cox1 sequence), S. corallinus (n = 1), S. doliatus (n = 4), S. lineatus (n = 13) and S. punctatus (n = 1); unsequenced vouchers have been identified from S. corallinus (n = 12), S. doliatus (n = 19) and S. lineatus (n = 31). In contrast, specimens of H. kuramochii have been sequenced from S. argenteus (n = 6 cox1 sequences), S. fuscescens (n = 26), S. lineatus (n = 1), S. spinus (n = 9), S. trispilos (n = 3) and the acanthurid Prionurus maculatus (n = 1); unsequenced identified vouchers are from Acanthurus triostegus (n = 1), S. fuscescens (n = 50) and S. spinus (n = 2). Morphologically, these two forms are highly similar. Notably, based on cox1 sequence data, S. argenteus and S. lineatus may be infected by both of these species. All specimens from S. argenteus were sequenced and are thus not ambiguous. Unsequenced specimens from S. lineatus were interpreted as representing H. urabeae because 13 of 14 cox1 sequences from that host related to H. urabeae and only one related to H. kuramochii. We found no reliable morphological distinction between these two species. Hexangium kuramochii has a noticeably larger size range [1,665–5,144 (3,392)] than H. urabeae [1,493–3,443 (2,377)], but although perhaps biologically significant, this distinction is of no value for the identification of smaller specimens. The only morphological distinction that we found is an imperfect tendency for the vitelline follicles to approach closer to the posterior margin of the oesophageal bulb in H. kuramochii than in H. urabeae (Figure 12F). Although the overall pattern of this character again suggests a distinction between these two species, many specimens overlap with regards to it.

Specimens of H. urabeae are sharply divided into distinct locality-based clades (Heron and Lizard Islands) based on ITS2 data (4 bp distinction) but show only slight distinction based on the cox1 (2–7 bp) and 28S (1 bp) datasets. Both clades overwhelmingly infect S. lineatus. We detected no morphological distinctions between them. Specimens of H. kuramochii from Moreton Bay and Lizard Islands differ consistently at a single bp in both the ITS2 and 28S datasets. Both clades have been found predominantly in S. fuscescens. We detected no morphological differences between samples from the two localities.

Eggs

Egg size was considered as a possible independent marker of distinction between species after initial distinction on the basis of molecular, host and other morphological data. We measured the length and width of 547 eggs selected on the basis of the extent to which they were orientated to allow accurate length measurement and still in a realistic oval shape. Means of length and width for the six species are charted in Figure 13A. The greatest separation of means was between those for H. sigani (87.3 × 51.9) and H. affinum (76.8 × 49.0); the most similar means, those for H. urabeae and H. robustum, differed by <1 μm in each dimension. When all the data points were plotted for pairwise combinations of Hexangium species, only those for H. sigani and H. affinum were convincingly separated, although still with a few overlapping data points (Figure 13B). Three other pairs of species (H. affinum + H. kuramochii; H. sigani + H. reimeri; and H. kuramochii + H. reimeri) show moderate separation but with considerable overlap. For all other pairs, the overlap between ranges largely or completely obscures any signal of distinction. These data fail to resolve three pairs of species that are otherwise the most difficult to distinguish by morphology: H. urabeae + H. sigani, H. affinum + H. reimeri, and H. urabeae + H. kuramochii. Thus, egg size provides support for some of the distinctions recognised here, but in no case facilitates completely unambiguous species delineation.

Figure 13.

Egg size of six Hexangium species. A, Mean length and width of eggs of six Hexangium OTUs. B, Lengths and widths of individual eggs. H. urabeae ▲; H. sigani □; H. affinum ●; H. kuramochii ♦; H. robustum +; H. reimeri ■.

Figure 13. Long description

A two-panel scatter plot.

Panel A is titled with a capital A in the top left. The Y axis is labeled Width and ranges from 48 to 53. The X axis is labeled Length and ranges from 76 to 88. Six distinct symbols represent mean values. A solid circle is at approximately length 77 width 49. A solid square is at length 81 width 49. A solid triangle is at length 81.5 width 51. A plus sign is slightly above and to the right of the triangle. A solid diamond is at length 84 width 52.8. An open square is at length 87.5 width 52.

Panel B is titled with a capital B in the top left. The Y axis ranges from 35 to 65 and the X axis ranges from 60 to 100. This plot shows a dense cluster of individual data points. Solid circles are concentrated between length 65 to 85 and width 40 to 55. Open squares are concentrated further to the right between length 80 to 95 and width 42 to 60. There is a significant overlap between the two groups in the center of the plot around length 82 width 52.

Host-specificity

Table 6 summarises the host distribution of the six Hexangium species relative to 12 examined Siganus species. This table distinguishes between identifications based on sequences and those made in the light of interpretation of morphological, host and geographical distributional data. Overall, the host-specificity of Hexangium species appears to be both variable and variably informative in this system. We observe the following key patterns. Two Hexangium species are restricted to just S. fuscescens, whereas the other four species have been found in at least five species of Siganus. No species has been found in species of all four siganid clades. No Hexangium species is restricted to S. argenteus, the sister to all other siganid species sampled; it shares its three species with the other Siganus clades. No species is found in and restricted to both the two Clade 2 siganids (S. fuscescens and S. spinus). Hexangium urabeae occurs predominantly in Clade 3.2 but also rarely in Clades 1 and 3.1. Hexangium robustum is primarily in Clade 3.1, but specimens from one S. corallinus, two S. doliatus and two S. lineatus (Clade 3.2) are identified as this species. Hexangium affinum and H. kuramochii (both in three Siganus clades) appear to have the lowest specificity, each infecting species of three siganid clades.

Table 6.

Host distribution of six Hexangium species in Indo-Pacific siganids. Filled circles represent species/host combinations confirmed from molecular data; empty circles represent combinations inferred without molecular data

Table 6. Long description

The table columns list Clade, Host Species, six Hexangium species (H. urabeae, H. sigani, H. affinum, H. kuramochii, H. robustum, H. reimeri), and a row-wise count. Filled circles indicate molecularly confirmed combinations; empty circles indicate inferred combinations.

* Clade 1: S. argenteus hosts H. urabeae (confirmed), H. affinum (confirmed), and H. kuramochii (confirmed). Count 3.

* Clade 2: S. fuscescens hosts H. sigani (confirmed), H. kuramochii (confirmed), and H. reimeri (confirmed). Count 3. S. spinus hosts H. kuramochii (confirmed). Count 1.

* Clade 3.1: S. puellus hosts H. urabeae (inferred), H. affinum (inferred), and H. robustum (confirmed). Count 3. S. punctatissimus hosts H. affinum (confirmed) and H. robustum (confirmed). Count 2. S. punctatus hosts H. urabeae (confirmed), H. affinum (inferred), and H. robustum (confirmed). Count 31. S. vulpinus hosts H. robustum (confirmed). Count 1.

* Clade 3.2: S. corallinus hosts H. urabeae (confirmed) and H. affinum (confirmed). Count 2. S. doliatus hosts H. urabeae (confirmed), H. affinum (confirmed), and H. robustum (inferred). Count 3. S. lineatus hosts H. urabeae (confirmed), H. kuramochii (confirmed), and H. robustum (inferred). Count 3. S. trispilos hosts H. affinum (confirmed) and H. kuramochii (confirmed). Count 2. S. virgatus hosts H. affinum (confirmed). Count 1.

Bottom row totals for parasite species: H. urabeae (6), H. sigani (1), H. affinum (8), H. kuramochii (5), H. robustum (6), H. reimeri (1).

Geographical distribution

Table 7 summarises the geographical distribution of the six putative Hexangium species relative to six collection localities, distinguishing between identifications based on cox1 sequences and those inferred only from morphological, host and geographical distributional data. French Polynesia, where moderate numbers of two siganids have been examined (S. argenteus, n = 17; S. spinus, n = 18), has not produced any infections. Eight Siganus species were examined at Palau without finding any infections, but numbers of individuals were low. Moreton Bay, where just one siganid species (S. fuscescens) has been sampled, has only a single species based on seven cox1 sequences. Japanese waters have a rich fauna of siganids (13 according to FishBase), but we have sampled only S. fuscescens, which, based on 6 sequences, has a single Hexangium species. Heron Island, Lizard Island and Ningaloo Reef are by far the best studied localities with 8, 10 and 4 siganid species sampled, respectively. All six Hexangium species have been found in at least two localities.

Table 7.

Geographical distribution of six Hexangium species. Filled circles indicate combinations confirmed by molecular data. Empty circles indicate combinations without molecular data but inferred from other indications. HI – Heron Island; LI – Lizard Island; NR – Ningaloo Reef

Table 7. Long description

The table consists of 8 columns and 8 rows. The columns are labeled: Species Name, L I, H I, New Cal. (New Caledonia), N R, Japan, S E Queensland, and n (total count of locations per species). Filled circles indicate molecularly confirmed data, while empty circles indicate inferred data.

* Hexangium urabeae: Filled circles at L I and H I; empty circle at New Cal.; n equals 3.

* Hexangium sigani: Filled circle at Japan; n equals 1.

* Hexangium affinum: Filled circles at L I, H I, and N R; empty circle at New Cal.; n equals 4.

* Hexangium kuramochii: Filled circles at L I, N R, and S E Queensland; n equals 3.

* Hexangium robustum: Filled circle at L I; empty circle at H I; n equals 2.

* Hexangium reimeri: Filled circles at H I and N R; n equals 2.

The bottom row provides a total species count per location: L I has 4; H I has 4; New Cal. has 2; N R has 3; Japan has 1; S E Queensland has 1.