Dipteronotus ornatus, copyright Ghedoghedo.

Belongs within: Actinopteri.
Contains: Holostei, Teleostei.

Who left all this fish lying around?
Published 19 May 2009
Two species of the swordfish-like Cretaceous pachycormid Protosphyraena. This genus was not even closely related to the modern swordfish (contra Wikipedia), and represents a case of convergence. Reconstruction by Dmitry Bogdanov.

The Neopterygii, or “new fins” (not, as it is often translated, “new wings”) are one of the most successful clades of fishes today. One particular subgroup of the Neopterygii, the teleosts, includes almost all the living ray-finned fishes. However, just to be difficult, I decided that the most appropriate tack for a post on Neopterygii was to leave the teleosts in all their diversity for another time, and focus on the non-teleost neopterygians. This, as it turns out, was a mistake. The non-teleost neopterygians seem, to a fish, to be almost universally ignored, and most of what there is out there was covered by Toby White almost seven years ago. Nevertheless, I’ll see what I can do.

The origins of the Neopterygii date back to sometime in the Permian (Hurley et al. 2007). Compared to earlier actinopterygians, the ancestors of Neopterygii lost their clavicle, beginning a trend of lightening and strengthening their skeletons, while at the same time reducing the weight of their scales. Early fish had been heavily armoured arrangements, but like the origins of the modern military, neopterygians were to trade in their clunky plate armour for something a bit more like a bullet-proof jacket*.

*Something that has almost nothing to do with the main post, but which struck me when I was thinking about it yesterday evening: When one looks at the living vertebrates only, it is easy to imagine that there was a progressive development of the bony skeleton—at the base of the tree, we have the living cartilaginous fishes and jawless fishes with little or no ossification, followed by the bony fishes and the tetrapods mostly with full skeletons. The fossil record, however, indicates that things were a little more complicated—early fishes such as placoderms had extensive skeletons, and the modern unossified fishes are actually the descendants of vertebrates that lost most of their skeletons. However, the original vertebrate bony skeleton did differ from the modern bony skeleton in one major regard—it was on the outside. Early fish had great coverings of bony armour, but little ossified interior skeleton. So over the course of evolution, vertebrates have gone from having their skeletons on the outside and meaty parts in the middle, to have the meaty parts on the outside and the skeletons in the middle. In other words, vertebrates have effectively been turned inside out.

Longnose gar Lepisosteus osseus, one of the few living non-teleost neopterygians. Photo from here.

There are few living groups of non-teleost neopterygians—in fact, there’s only two, both restricted to fresh waters of North America. One group, the Halecostomi, is represented in the modern fauna by only a single species, the bowfin, Amia calva. As Toby has noted before me, perhaps the single most remarkable feature of the bowfin is that it has absolutely nothing remarkable about it whatsoever. Amiid fishes go all the way back to the Jurassic, and don’t look too much different from each other in all that time. The other living group, the American gars of the family Lepisosteidae, are entirely a different matter—gigantic carnivorous fish, with long beaks and sharp teeth. The largest gars can be over two metres long, and according to this site Rafinesque referred to gars up to twelve feet long. They also lay eggs that are toxic to humans. Unfortunately, it looks like American gars don’t have green bones, despite common rumour—the green-boned “garfish” is a quite different, marine fish (Belone) nestled well within the teleosts.

Bowfin, Amia calva, the other survivor. Photo from here.

Relationships between the neopterygian clades are almost completely obscure—while features of the jaw musculature support a relationship between Amia and teleosts to the exclusion of gars, other authors have supported an Amia-Lepisosteidae clade that excludes teleosts. Hurley et al. (2007) found the latter result in a morphological analysis, but the former in a molecular analysis. While a number of fossil groups of non-teleost neopterygians are known, few authors seem to have plugged them into a phylogenetic analysis except for Hurley et al. (2007) and Arratia (2001) (the latter of which I don’t have access to). A number of authors have supported a relationship between the gars and the extinct Semionotiformes (Olsen & McCune 1991), while the Pachycormiformes and Aspidorhynchiformes seem likely to be stem-teleosts. Finally, the Dapediidae and Pycnodontiformes were found by Hurley et al. (2007) to form a third clade in a polytomy with the Amia-Lepisosteidae clade and the teleosts.

The pycnodontiform Coelodus costai. Photo by Giovanni Dall’Orto.

Some of these were decidedly odd fishes. The Pycnodontiformes were deep-bodied fish, about as tall as they were long. They had strong teeth, and would have fed on shellfish. The Pachycormiformes, mostly pelagic hunters, are best known through the monster Leedsichthys, a gigantic filter feeder growing to lengths over ten metres, which is probably the largest known ray-finned fish.

Figure from McCune (2004), showing a reconstruction of Semionotus, and variation in dorsal spine row morphology and overall body shape in Newark Semionotus.

Perhaps the coolest of all, though, were the Semionotidae. Semionotus wasn’t anything much to look at—not spectacularly large (probably about half a foot) and pretty generalised morphologically. During the Mesozoic it was found in freshwater deposits pretty much around the world, so it would have been dirt common. Where things get interesting is when you get to the Late Triassic and Early Jurassic Newark Supergroup of eastern North America. The Newark Supergroup comprises a series of lake deposits, formed by a process of rifting similar to the modern Great Lakes of Africa. And Semionotus was the Newark deposits’ cichlid. Within a single lake deposit, a whole series of Semionotus species can be found, varying from long and narrow to deep-bodied and humpbacked (McCune 2004). And that is very cool—that the incredible African cichlid radiation is not so incredible after all, but represents patterns and processes that were just as active 100 million years ago.

To drop jaw or not?
Published 11 March 2021

The vast majority of living ray-finned fishes (that is, all of them except for bichirs, sturgeons and paddlefishes) fall under the auspices of the clade Neopterygii. I have commented on this clade in earlier posts and in those posts I have noted that modern neopterygians can theselves be divided between three basal lineages. By far the largest of these is the teleosts with only a handful of species representing the other two: the seven or so species of gar in the Lepisosteidae, and the phylogenetically isolated bowfin Amia calva. However, the exact relationships between these three lineages have been the subject of debate.

Close-up on bowfin Amia calva head, from Big Fishes of the World. Note the membranous attachment of the back of the upper jaw.

Historically, the bowfin and the gars were recognised as a group Holostei in apposition to the Teleostei. When first established, this division was motivated primarily by the nature of their scales: the heavy, solid scales of the holosteans contrasted with the thinner, lighter scales of the teleosts. Hence the name ‘Holostei’ meaning ‘entirely bone’: the holosteans have both a completely bony skeleton on the inside (as opposed to the partially cartilaginous skeletons of more basal fishes) and a complete covering of bony scales on the outside. However, the heavy scales of the Holostei are a primitive feature, indicating that the two lineages diverged before the evolution of the lighter teleost scales but not indicating a direct relationship with each other.

With the increasing emphasis on evolutionary relationships as the primary informer of classifications, a different system was proposed. This saw the gars as the most divergent lineage of the Neopterygii with the bowfin being united with the teleosts as a clade Halecostomi. This time, the primary evidence for this division was in how their jaws worked. The ancestral condition for vertebrate jaws has them working much as our own still do. The upper jaw, the maxilla, is largely fixed in place against the base of the neurocranium (the brain-holding bit) while the movement of opening and closing the mouth is achieved by the lower jaw, the mandible, pivoting around its hinge towards the back of the skull. In the bowfin and teleosts, however, the maxilla is hinged with the skull at its anterior end and with the mandible at the back. When the mouth opens, the maxila pivots downwards from this anterior hinge, dropping the mandible as a whole downwards. The bowfin and teleosts also possess a bone in the cheek, the interopercular bone, that is not found in other fishes; a muscle attached to this bone rotates the gill operculum as the mouth opens (Lauder 1980). Functionally, the expansion of the mouth cavity in this manner of opening the jaws creates a suction that pulls prey or other food into the fish’s mouth.

Though it was by no means universally accepted, it is probably fair to say that the halecostomes vs gars picture of neopterygian evolution became the majority view. But then came the advent of molecular phylogenetic analysis, all ready and willing to cast the proverbial spanner. Rather than confirming halecostome monophyly, molecular analyses pointed the other way, back towards a clade of the bowfin and gars. Following this, a detailed study of gar systematics published by Grande (2010) also supported a gar plus bowfin monophylum on morphological grounds and resurrected the concept of Holostei (albeit redefined on phylogenetic grounds).

Skull of a longnose gar Lepisosteus osseus, from Grande (2010). In the lower diagram, the maxilla is labelled ‘mx’ and the lacrimomaxillaries are labelled ‘lmx’.

Gar jaws, it should be noted at this point, are a bit weird. Rather than being primarily composed of a single maxilla on each side, the upper jaws are made up of a series of tooth-bearing bones, each bone carrying just a few teeth, that have been dubbed the lacrimomaxillaries. When the jaws open, as well as the lower jaw opening in the standard manner, the flexible upper jaw also bends upwards. Rather than using suction to draw in their food like other neopterygians, gars capture prey by sneaking up to it then using a quick sideways jerk of the head to bring the open jaws around the prey (Lauder 1980). Gars were excluded from the Halecostomi on the basis of their lack of a long, mobile maxilla but, as explained by Grande (2010), a mobile maxilla is indeed present in gars but reduced to a remnant splint at the back of the jaw (in mature alligator gars Atractosteus spatula, the maxilla does not ossify). In very young juvenile gars, the mobile maxilla remains a significant part of the upper jaw with the lacrimomaxillaries being added in front of it as the jaw lengthens. As for the interopercular, this is genuinely absent in modern gars but it is present in close fossil relatives of gars such as semionotids. Rather than retaining a primitive jaw structure that was superseded in the bowfin and teleosts, it appears that gars evolved their own derived jaw structure from ‘halecostome’ ancestors.

Given that suction-assisted feeding is generally regarded as a major advance in fish evolution, how did gars end up abandoning it? That I can only speculate about. Is it related to the evolution of their elongate rostra? Long beaks are certainly a thing for a number of teleosts, but I don’t know if any have a beak as long and robust as a gar’s. Could it be that the greater precision of gars’ snapping mode of feeding is an advantage in the low-oxygen, muck-filled waters in which gars thrive? Or could it be a side effect somehow of gars’ more heavily armoured condition than other early-diverging neopterygians?

It’s only fair to note that monophyly of Holostei is still not universally accepted; there are sill researchers who are inclined to think the bowfin closer to teleosts. But even if the ‘Halecostomi’ hypothesis was to rise once more to the surface, it would not be for the same reasons it did before.

Systematics of Neopterygii
<==Neopterygii [Halecostomi]XG11
    |--Discoserra [Guildayichthyidae]GX17
    |    `--D. pectinodonFS10
    `--+--+--Bobasatraniidae [Bobasatraniformes]RE01
       |  |    |--Ebenaqua ritchiei Campbell & Duy Phuoc 1983RE01
       |  |    |--Ecrinesomus dixoni Woodward 1910G93
       |  |    |--Polzbergia brochatus Griffith 1977G93
       |  |    `--Bobasatrania White 1932GX17, G93
       |  |         `--B. groenlandicaGX17
       |  `--Dorypterus [Dorypteridae, Dorypteriformes, Dorypteroidei]C93
       |       |--D. althausi (Münster 1842)G93
       |       `--D. hoffmanni Germar 1842G93
       `--+--Venusichthys comptusGX17
          `--+--+--Hulettia americanaGX17
             |  `--+--HolosteiFS10
             |     `--TeleosteiGX17
                |  |    |--ThoracopteridaeG93
                |  |    |    |--Thoracopterus Bronn 1858G93
                |  |    |    `--Gigantopterus Abel 1904G93
                |  |    `--LuganoiidaeG93
                |  |         |--Besania micrognathus Brough 1939G93
                |  |         `--Luganoia Brough 1939GX17, G93
                |  |              `--L. lepidosteoidesGX17
                |  `--PeltopleuriformesG93
                |       |--HabroichthyidaeG93
                |       |    |--Habroichthys minimus Brough 1939G93
                |       |    `--Nannolepis elegans Griffith 1977G93
                |       `--PeltopleuridaeG93
                |            |--Placopleurus Brough 1939G93
                |            |--Peltopleurus Kner 1866GX17, G93
                |            |    `--P. lissocephalusGX17
                |            `--Platysiagum Egerton 1872G93
                |                 `--P. sclerocephalus Egerton 1872G93
                     |--Aetheodontus [Aetheodontidae]G93
                     |    `--A. besanensis Brough 1939G93
                     |    |--CleithrolepisG93
                     |    |    |--C. alta Woodward 1890 [=C. altus]F71
                     |    |    `--C. granulata Egerton 1864G93 [=C. granulatusF71]
                     |    `--DipteronotusGX17
                     |         |--D. cyphus Egerton 1854G93
                     |         `--D. ornatusGX17
                          |--Colobodus Agassiz 1844G93
                          |--Perleidus Deeke 1911G93
                          |--Meidiichthys Borough 1931G93
                          |--Boreichthys Selezneva 1982G93
                          |--Plesioperleidus Dazae & Li 1983G93
                          |--Tripelta Wade 1940G93
                          |    `--T. dubia (Woodward 1890) [=Peltopleurus dubius]F71
                          |--Chrotichthys Wade 1940G93
                          |    `--C. gregarius (Woodward 1890) [=Pholidophorus gregarius]F71
                          |--Zeuchthiscus Wade 1940G93
                          |    `--Z. australis (Woodward 1890) [=Semionotus australis; incl. S. tenuis Woodward 1908]F71
                          `--Pristisomus Woodward 1890G93
                               `--P. gracilis Woodward 1890 [incl. P. crassus Woodward 1890, P. latus Woodward 1890]F71
Neopterygii incertae sedis:
  Cephaloxenus [Cephaloxenidae, Cephaloxeniformes]G93
    `--C. macropterus Brough 1939G93
  Acentrophorus Traquair 1877 [Acentrophoridae]G93
    |--Oligopleurus Tholliére 1850G93
    |--Spathiurus Davis 1887G93
    `--Oshunia Wentz & Kellner 1986G93
    |--Callopterus Tholliére 1858G93
    `--Ionoscopus Tholliére 1858G93
  Uarbrichthys Wade 1941 [Uarbrichthyidae]G93
    |--U. incertus Wade 1953F71
    `--U. latus Wade 1941F71

*Type species of generic name indicated


Arratia, G. 2001. The sister group of Teleostei: consensus and disagreements. Journal of Vertebrate Paleontology 21 (4): 767–773.

[C93] Coates, M. I. 1993. New actinopterygian fish from the Namurian Manse Burn Formation of Bearsden, Scotland. Palaeontology 36: 123–146.

[F71] Fletcher, H. O. 1971. Catalogue of type specimens of fossils in the Australian Museum, Sydney. Australian Museum Memoir 13: 1–167.

[FS10] Friedman, M., K. Shimada, L. D. Martin, M. J. Everhart, J. Liston, A. Maltese & M. Triebold. 2010. 100-million-year dynasty of giant planktivorous bony fishes in the Mesozoic seas. Science 327: 990–993.

[G93] Gardiner, B. G. 1993. Osteichthyes: basal actinopterygians. In: Benton, M. J. (ed.) The Fossil Record 2 pp. 611–619. Chapman & Hall: London.

[GX17] Giles, S., G.-H. Xu, T. J. Near & M. Friedman. 2017. Early members of ‘living fossil’ lineage imply later origin of modern ray-finned fishes. Nature 549: 265–268.

Grande, L. 2010. An empirical synthetic pattern study of gars (Lepisosteiformes) and closely related species, based mostly on skeletal anatomy. The resurrection of Holostei. Copeia 2010 (2A): iii–x, 1–871.

Hurley, I. A., R. Lockridge Mueller, K. A. Dunn, E. J. Schmidt, M. Friedman, R. K. Ho, V. E. Prince, Z. Yang, M. G. Thomas & M. I. Coates. 2007. A new time-scale for ray-finned fish evolution. Proceedings of the Royal Society of London Series B 274: 489–498.

Lauder, G. V., Jr. 1980. Evolution of the feeding mechanism in primitive actinopterygian fishes: a functional anatomical analysis of Polypterus, Lepisosteus, and Amia. Journal of Morphology 163: 283–317.

McCune, A. R. 2004. Diversity and speciation of semionotid fishes in Mesozoic rift lakes. In: Dieckmann, U., M. Doebeli, J. A. J. Metz & D. Tautz (eds) Adaptive Speciation pp. 362–379. Cambridge University Press.

Olsen, P. E., & A. R. McCune. 1991. Morphology of the Semionotus elegans species group from the Early Jurassic part of the Newark Supergroup of eastern North America with comments on the family Semionotidae (Neopterygii). Journal of Vertebrate Paleontology 11 (3): 269–292.

[RE01] Ritchie, A., & G. D. Edgecombe. 2001. An odontogriphid from the Upper Permian of Australia. Palaeontology 44 (5): 861–874.

[XG11] Xu, G.-H. & K.-Q. Gao. 2011. A new scanilepiform from the Lower Triassic of northern Gansu Province, China, and phylogenetic relationships of non-teleostean Actinopterygii. Zoological Journal of the Linnean Society 161: 595–612.

Leave a comment

Your email address will not be published. Required fields are marked *