Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Sep 2;109(4):2512–2515. doi: 10.1111/jfb.70644

First record of piscivory in ghost pipefishes (Teleostei, Solenostomidae) revealed by micro‐computed tomography

Graham Short 1,2,3,4,5,✉, Kerryn Parkinson 1, Thomas J Peachey 1, Richard Smith 5, David Harasti 5,6
PMCID: PMC13675491  PMID: 42686108

Abstract

Ghost pipefishes (Solenostomidae) have been regarded as specialist crustacean feeders since the 1960s, an inference resting on two primary dietary accounts. We report the first documented evidence of piscivory in the family based on micro‐computed tomography (micro‐CT) examination of voucher specimens of Solenostomus snuffleupagus and S. cyanopterus. Two head‐first teleost prey, both identified as gobiids, were recovered from the holotype of S. snuffleupagus, and additional ichthyic remains were recovered from an S. cyanopterus specimen collected in 1952, indicating a solenostomid diet broader than previously recognized.

Keywords: diet, gut contents, micro‐CT, piscivory, Solenostomidae, syngnathiform

Ghost pipefishes (family Solenostomidae) are small, cryptic syngnathiform fishes distributed throughout the tropical Indo‐Pacific and the sister group to the Syngnathidae (seahorses, pipefishes and sea dragons; Orr & Fritzsche, 1993; Hamilton et al., 2017; Stiller et al., 2022). Despite their popularity among recreational divers, solenostomids remain among the least‐studied members of the order Syngnathiformes (Ascione et al., 2026; Orr & Fritzsche, 1993; Wetzel & Wourms, 1995). The foundational dietary account rests on Fishelson's (1966) captive observations of five Solenostomus cyanopterus Bleeker, 1854 feeding on small benthic invertebrates and crustaceans. This account has been repeated in subsequent treatments (Fritzsche & Thiesfeld, 1999; Manning et al., 2019). A recent review (Ascione et al., 2026) found primary dietary records confirmed for only two species, S. cyanopterus and S. paegnius Jordan & Thompson, 1914, the record for the latter species resting on Agara (1967). Field observations by divers of solenostomids taking small crustaceans are widely reported in the aquarium and dive literature, but such accounts reach the literature as tertiary compilations and not as primary observations (Allen & Erdmann, 2012; Ascione et al., 2026; Kuiter, 2009). Piscivory has been asserted in the aquarium literature, where large solenostomids are said to take very small fishes in the wild and live guppies in captivity (Ross & Wandell, 2012), but no specimen, locality or observation is given in support. Solenostomids are reported to feed as ambush predators that capture prey by pivot feeding, a mechanism shared across Syngnathiformes (van Wassenbergh et al., 2014). Prey are engulfed whole rather than grasped and processed with oral jaw teeth, placing solenostomid feeding in the engulfer functional group recognized for piscivorous fishes (Mihalitsis & Bellwood, 2021). We report micro‐computed tomography (micro‐CT) evidence of teleost skeletal remains in the gut contents of preserved voucher specimens of two Solenostomus species, the first documented record of piscivory in the family.

Micro‐CT scanning was conducted on preserved voucher specimens of Solenostomus species. Material examined comprised the holotype and paratype of S. snuffleupagus Short & Harasti, 2026 (AMS I.33751–047, female, 33.6 mm SL, Portlock Reef, Coral Sea, Queensland, Australia, 29 January 1993; NTM S.13600–047, male, 21.3 mm SL, Ashmore Reef, Coral Sea, Queensland, Australia, 19 January 1993) and two specimens of S. cyanopterus (SAIAB 14841, female, collected by J.L.B. Smith and M. Smith at Shimoni, Kenya, 1952; AMS I.51717‐001, female, 85 mm SL, Port Stephens, New South Wales, Australia, 11 January 2025). Institutional acronyms follow Sabaj (2020).

Type specimens of S. snuffleupagus were scanned at the Australian Museum using a Nikon XT H 225 micro‐CT scanner at 10.0–17.8 μm voxel resolution, 70 kV and 210–220 μA. Specimens were sealed in polyethylene film and reconstructed using VG Studio Max 2023.3.1 (Volume Graphics). The S. cyanopterus specimen was scanned at the Karel F. Liem Bioimaging Facility, Friday Harbour Laboratories, on a Bruker Skyscan 1273 at 60 kV, 110 μA and 4.5 μm resolution. Volume rendering and digital segmentation were performed in 3D Slicer v5.6 (Fedorov et al., 2012). In the holotype of S. snuffleupagus, gut contents were isolated and false‐colourized through manual segmentation to discriminate prey osteological elements from host skeletal tissue using VG Studio Max 2023.3.1 (Volume Graphics). In SAIAB 14841, ichthyic skeletal remains were identifiable in unprocessed volume renderings without supplementary segmentation. Linear dimensions of the prey and of the host upper jaw were measured on the segmented volumes in 3D Slicer, with maximum prey body depth taken as the greatest dorsoventral dimension of the prey trunk, and premaxilla–maxilla length taken following Mihalitsis and Bellwood (2019), who recovered premaxilla–maxilla length as a strong correlate of both vertical and horizontal gape size and therefore as a proxy for gape in material where the jaws cannot be manipulated. An ingestion index was calculated as maximum prey body depth divided by host premaxilla–maxilla length and relative prey depth as maximum prey body depth divided by host standard length. Both indices were computed for the holotype of S. snuffleupagus and for its larger prey item only. The smaller prey item and the SAIAB 14841 remains were too incompletely preserved for reliable depth measurement. Premaxilla–maxilla length was measured with the jaws in the resting position, and the resulting index therefore refers to the resting oral aperture and not to the aperture attained during a feeding strike. Host total length was taken as standard length plus caudal fin length, the caudal fin of the holotype measuring 11.5 mm. Measurements are reported to the nearest 0.01 mm.

Micro‐CT imaging of the holotype of S. snuffleupagus (AMS I.33751–047, female, 33.6 mm SL, 45.1 mm TL) revealed two partially digested teleosts within the V‐shaped intestinal tract, both oriented head‐first, with osteological remains comprising neurocranial elements, vertebrae and caudal fin rays (Figure 1a,b). The larger (yellow) and smaller (violet) prey are estimated at approximately 10 and 8 mm total length, about 30% and 24% of the host's 33.6 mm standard length. Maximum body depth of the larger (yellow) prey item is 1.40 mm, and premaxilla–maxilla length of the holotype is 1.83 mm, or 5.4% of standard length, giving an ingestion index of 0.76 for this prey item and a relative prey depth of 4.2% of host standard length. Neither index could be calculated for the smaller (violet) item or for the SAIAB 14841 remains. Both prey conformed to the luminal contour of the gut and were partially digested, consistent with whole‐prey ingestion well before fixation rather than with incidental uptake during collection or net feeding. The larger (yellow) item preserves a basipterygium bearing a thin, ventrally directed process, together with a non‐osseous gap between the symplectic and preopercle in the suspensorium, a characteristic feature of gobiid suspensoria (Tornabene et al., 2018; Vaz et al., 2025). Preservation is less complete in the smaller (violet) item, which nonetheless shows comparable cranial and pelvic morphology. Both prey are accordingly identified as gobiids (Gobiidae) of indeterminate species (Winterbottom, 2003; Vaz et al., 2025; D. F. B. Vaz, personal communication, 2026). The paratype (NTM S.13600–047, male, 21.3 mm SL) contained gut material that could not be confirmed as ichthyic at the available voxel resolution. Lacking diagnostic skeletal elements, it is scored as indeterminate rather than as evidence of absence. Micro‐CT examination of S. cyanopterus SAIAB 14841 revealed comparable ichthyic skeletal material, also oriented head‐first, including vertebral column fragments and lepidotrichial elements (Figure 1c,d).

FIGURE 1.

FIGURE 1

Open in a new tab

Micro‐CT evidence of piscivory in Solenostomus. (a) S. snuffleupagus holotype (AMS I.33751–047, female) with two head‐first prey items false‐colourized in yellow and violet within the intestinal tract; host skeleton in grey. (b) Isolated view of the two segmented prey items. (c) S. cyanopterus (SAIAB 14841, female) in lateral view; arrow indicates prey location within the gut. (d) Magnified view of the gut region in (c); arrows indicate the head‐first oriented ichthyic skeletal remains, including vertebral column fragments and lepidotrichial elements of a partially digested teleost.

The recovery of teleost osteological material from S. snuffleupagus and S. cyanopterus extends the record of piscivory to a second species and a different ocean basin (the western Indian Ocean) and requires the family's assumed dependence on crustacean prey to be reassessed. In both specimens the prey were oriented head‐first, which is consistent with capture by suction feeding. Captive maintenance at public aquariums relies exclusively on live micro‐crustacean diets, with feeder fish declined on nutritional grounds rather than refused by the fish (Ross & Wandell, 2012), a practical constraint rather than evidence of natural feeding. In the wild, S. cyanopterus has been observed feeding diurnally on small crustaceans at Nelson Bay, Port Stephens, New South Wales, Australia (D.H., personal observation). At Anilao, Batangas, Philippines, S. paradoxus (Pallas, 1770) has been observed feeding nocturnally on similar prey (G. Paras, personal communication, 2026). A further field observation from Wakatobi, Indonesia, records a S. halimeda Orr, Fritzsche & Randall, 2002 directing a feeding attempt at a pygmy seahorse, Hippocampus pontohi Lourie & Kuiter, 2008 (R.S., personal observation; Figure S1). The pipefish abandoned the attempt and swam away, so the observation records an unsuccessful strike rather than a completed predation event, but it nonetheless extends the prey recognized as targets by S. halimeda to small syngnathids.

The remains reported here show that solenostomids take teleosts, which are available in the macroalgal and rubble substrata these species occupy. Suction generated by the pivot‐feeding apparatus is sufficient to engulf prey of this size in seahorses (van Wassenbergh et al., 2014). The larger prey item is 30% of host standard length and 22% of host total length, the latter below the mean relative prey size of 37% reported for engulfing piscivores (Mihalitsis & Bellwood, 2021). Its body depth is only 4.2% of host standard length, well below the maximum relative prey depths recorded for benthic reef piscivores that ingest prey whole (20–27% of standard length; Mihalitsis & Bellwood, 2017). Depth and not length is the dimension that must pass the gape, and the slender gobiid body form of the prey is consistent with passage through the solenostomid oral aperture. Syngnathids are not strictly limited by gape, since Hippocampus guttulatus Cuvier, 1829 ingests prey of 1–2 cm by fragmenting it (Ape et al., 2019), whereas the solenostomid prey reported here were entire. An ingestion index of 0.76 places the ingested prey close to the aperture implied by the upper jaw at rest. Should the solenostomid snout prove capable of expansion during the strike, as in other syngnathiforms, the aperture attained would exceed the resting value, so 0.76 is an upper bound on the fraction of the functional aperture occupied by this prey item. On either reading, prey depth appears to set the ecomorphological constraint on solenostomid piscivory, which predicts that elongate, laterally compressed teleosts are the prey most accessible to a tubular‐snouted engulfer at this body size.

Snout proportion constrains which prey are encountered as well as which can be swallowed, and across syngnathids, snout length relative to head length predicts prey mobility more closely than it predicts prey size. Species whose snouts exceed about half of head length take predominantly mobile water‐column prey, whereas shorter‐snouted species take slow‐moving epibenthic prey (Kendrick & Hyndes, 2005). With a snout 63.5% of head length in the holotype (Short & Harasti, 2026), S. snuffleupagus falls in the long‐snouted group, consistent with capture of an actively swimming gobiid.

If piscivory proves consistent across the genus, solenostomids may exert previously unrecognized predation pressure on small reef‐associated teleosts. Teleost prey has already been recorded in the sister family Syngnathidae, where tripterygiid remains occurred in 5.1% of the guts of wild Hippocampus abdominalis Lesson, 1827 examined by Woods (2002), and fish are listed among the prey of the long‐snouted pipefish Syngnathus typhle Linnaeus, 1758 (Kendrick & Hyndes, 2005). Occasional piscivory may therefore be more widespread across Syngnathiformes than the emphasis on crustacean prey in the review literature indicates (Manning et al., 2019), and this could be tested more broadly by micro‐CT screening of voucher material. Non‐destructive imaging recovered diagnosable prey osteology from the holotype of S. snuffleupagus more than three decades after fixation, so preserved specimens retain dietary information long after collection. Broader screening of Solenostomus specimens in museum collections would establish how widespread piscivory is across the genus.

AUTHOR CONTRIBUTIONS

Graham Short conceived and designed the study, performed micro‐CT imaging of S. cyanopterus (SAIAB 14841), interpreted the osteological data and drafted the manuscript. Kerryn Parkinson produced the micro‐CT images of the S. snuffleupagus type specimens and of S. cyanopterus (AMS I.51717‐001) and contributed to the writing of the manuscript. Thomas J Peachey performed the digital segmentation of the gut contents of S. snuffleupagus. David Harasti collected the specimen AMS I.51717‐001 and contributed the field observation of S. cyanopterus feeding at Nelson Bay. Richard Smith contributed the field observation of S. halimeda at Wakatobi and the photograph reproduced as Figure S1. All authors contributed to the interpretation of the results, critically revised the manuscript and approved the final version.

CONFLICT OF INTEREST STATEMENT

Graham Short, Richard Smith and David Harasti are members of the IUCN SSC Seahorse, Pipefish and Seadragon Specialist Group. The authors declare no other conflicts of interest.

Supporting information

Figure S1. Solenostomus halimeda Orr, Fritzsche & Randall, 2002 (upper left) and the pygmy seahorse Hippocampus pontohi Lourie & Kuiter, 2008 (lower right) photographed together among Halimeda at Wakatobi, Indonesia, during the feeding attempt reported in the text. Photograph by R.S.

JFB-109-2512-s001.png (34.8MB, png)

ACKNOWLEDGEMENTS

We thank Adam Summers (Friday Harbour Laboratories, University of Washington) for access to CT scanning facilities and Gen Paras (Aqua‐Venture Dive Resort, Philippines) for sharing field observations of Solenostomus. This research was supported in part by funding from the National Science Foundation (grants DBI‐1759637 and DEB‐170166) to Adam Summers and by the Australian Museum Research Institute. Open access publishing facilitated by University of Technology Sydney, as part of the Wiley ‐ University of Technology Sydney agreement via the Council of Australasian University Librarians

DATA AVAILABILITY STATEMENT

The micro‐CT datasets generated and analysed during this study are available from the corresponding author on reasonable request.

REFERENCES

  1. Agara, C. (1967). The third specimen of the ghost pipefish Solenostomus paegnius Jordan & Thompson, from Tanabe Bay. Publications of the Seto Marine Biological Laboratory, 15(1), 1–4. [Google Scholar]
  2. Allen, G. R. , & Erdmann, M. V. (2012). Reef fishes of the east indies. Tropical Reef Research. [Google Scholar]
  3. Ape, F. , Corriero, G. , Mirto, S. , Pierri, C. , Lazic, T. , & Gristina, M. (2019). Trophic flexibility and prey selection of the wild long‐snouted seahorse Hippocampus guttulatus Cuvier, 1829 in three coastal habitats. Estuarine, Coastal and Shelf Science, 224, 1–10. 10.1016/j.ecss.2019.04.034 [DOI] [Google Scholar]
  4. Ascione, S. J. , Camins, E. , Correia, M. , & Vincent, A. C. J. (2026). The state of knowledge on four families of Syngnathoidei fishes (Teleostei: Syngnathiformes): Aulostomidae, Centriscidae, Fistulariidae and Solenostomidae. Journal of Fish Biology, 109, 400–449. 10.1111/jfb.70393 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Fedorov, A. , Beichel, R. , Kalpathy‐Cramer, J. , Finet, J. , Fillion‐Robin, J.‐C. , Pujol, S. , Bauer, C. , Jennings, D. , Fennessy, F. , Sonka, M. , Buatti, J. , Aylward, S. , Miller, J. V. , Pieper, S. , & Kikinis, R. (2012). 3D slicer as an image computing platform for the quantitative imaging network. Magnetic Resonance Imaging, 30, 1323–1341. 10.1016/j.mri.2012.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Fishelson, L. (1966). Solenostomus cyanopterus Bleeker (Teleostei, Solenostomidae) in Elat (gulf of Akaba). Israel Journal of Zoology, 15, 95–103. [Google Scholar]
  7. Fritzsche, R. A. , & Thiesfeld, K. G. (1999). Family Solenostomidae. In Carpenter K. E. & Niem V. H. (Eds.), The living marine resources of the Western Central Pacific. FAO species identification guide for fisheries purposes (Vol. 4, p. 2263). FAO. [Google Scholar]
  8. Hamilton, H. , Saarman, N. , Short, G. , Sellas, A. , Moore, B. , Hoang, T. , Grace, C. , Pardo, B. , Jones, A. , Simison, W. B. , & Longo, S. (2017). Molecular phylogeny and patterns of diversification in Syngnathiformes. Molecular Phylogenetics and Evolution, 107, 388–403. 10.1016/j.ympev.2016.10.003 [DOI] [PubMed] [Google Scholar]
  9. Kendrick, A. J. , & Hyndes, G. A. (2005). Variations in the dietary compositions of morphologically diverse syngnathid fishes. Environmental Biology of Fishes, 72(4), 415–427. 10.1007/s10641-004-2597-y [DOI] [Google Scholar]
  10. Kuiter, R. H. (2009). Seahorses and their relatives. Aquapress. [Google Scholar]
  11. Manning, C. G. , Foster, S. J. , & Vincent, A. C. J. (2019). A review of the diets and feeding behaviors of a family of biologically diverse marine fishes (family Syngnathidae). Reviews in Fish Biology and Fisheries, 29(2), 197–221. 10.1007/s11160-019-09549-z [DOI] [Google Scholar]
  12. Mihalitsis, M. , & Bellwood, D. R. (2017). A morphological and functional basis for maximum prey size in piscivorous fishes. PLoS One, 12(9), e0184679. 10.1371/journal.pone.0184679 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Mihalitsis, M. , & Bellwood, D. R. (2019). Morphological and functional diversity of piscivorous fishes on coral reefs. Coral Reefs, 38, 945–954. 10.1007/s00338-019-01820-w [DOI] [Google Scholar]
  14. Mihalitsis, M. , & Bellwood, D. R. (2021). Functional groups in piscivorous fishes. Ecology and Evolution, 11, 12765–12778. 10.1002/ece3.8020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Orr, J. W. , & Fritzsche, R. A. (1993). Revision of the ghost pipefishes, family Solenostomidae (Teleostei: Syngnathoidei). Copeia, 1993(1), 168–182. 10.2307/1446307 [DOI] [Google Scholar]
  16. Ross, R. , & Wandell, M. (2012). Revealing the mysterious: Collection, husbandry, and spawning of these enigmatic relatives of the seahorses and true pipefishes. CORAL Magazine, 9(4), 128–143. [Google Scholar]
  17. Sabaj, M. H. (2020). Codes for natural history collections in ichthyology and herpetology. Copeia, 108(3), 593–669. 10.1643/asihcodons2020 [DOI] [Google Scholar]
  18. Short, G. , & Harasti, D. (2026). Solenostomus snuffleupagus sp. nov., a hairy ghost pipefish (Teleostei: Solenostomidae) from the Southwest Pacific, with an integrative comparison to S. paegnius . Journal of Fish Biology, 109, 1018–1033. 10.1111/jfb.70497 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Stiller, J. , Short, G. , Hamilton, H. , Saarman, N. , Longo, S. J. , Wainwright, P. C. , Rouse, G. W. , & Simison, W. B. (2022). Phylogenomic analysis of Syngnathidae reveals novel relationships, origins of endemic diversity and variable diversification rates. BMC Biology, 20(1), 75. 10.1186/s12915-022-01271-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Tornabene, L. , Deis, B. , & Erdmann, M. V. (2018). Evaluating the phylogenetic position of the goby genus Kelloggella (Teleostei: Gobiidae), with notes on osteology of the genus and description of a new species from Niue in the south central Pacific Ocean. Zoological Journal of the Linnean Society, 183, 143–162. 10.1093/zoolinnean/zlx060 [DOI] [Google Scholar]
  21. van Wassenbergh, S. , Dries, B. , & Herrel, A. (2014). New insights into muscle function during pivot feeding in seahorses. PLoS One, 9(10), e109068. 10.1371/journal.pone.0109068 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Vaz, D. F. , Goatley, C. H. , & Tornabene, L. (2025). Osteology of dwarfgobies Eviota and Sueviota (Gobiidae: Gobiomorpharia), with phylogenetic inferences within coral gobies. Journal of Morphology, 286(3), e70039. 10.1002/jmor.70039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Wetzel, J. , & Wourms, J. P. (1995). Adaptations for reproduction and development in the skin‐brooding ghost pipefishes, Solenostomus . Environmental Biology of Fishes, 44, 363–384. 10.1007/BF00008252 [DOI] [Google Scholar]
  24. Winterbottom, R. (2003). A new species of the gobiid fish Trimma from the western Pacific and northern Indian Ocean coral reefs, with a description of its osteology. Zootaxa, 218, 1–24. 10.11646/zootaxa.218.1.1 [DOI] [Google Scholar]
  25. Woods, C. M. C. (2002). Natural diet of the seahorse Hippocampus abdominalis . New Zealand Journal of Marine and Freshwater Research, 36(3), 655–660. 10.1080/00288330.2002.9517121 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1. Solenostomus halimeda Orr, Fritzsche & Randall, 2002 (upper left) and the pygmy seahorse Hippocampus pontohi Lourie & Kuiter, 2008 (lower right) photographed together among Halimeda at Wakatobi, Indonesia, during the feeding attempt reported in the text. Photograph by R.S.

JFB-109-2512-s001.png (34.8MB, png)

Data Availability Statement

The micro‐CT datasets generated and analysed during this study are available from the corresponding author on reasonable request.

Articles from Journal of Fish Biology are provided here courtesy of Wiley

RESOURCES