Ornithoprion

Ornithoprion is an extinct genus of cartilaginous fish. The only species, O. hertwigi, lived during the Moscovian stage of the Pennsylvanian, which spanned from 315 to 307 million years ago. Its fossils are preserved in black shales from what is now the Midwestern United States. The study of Ornithoprion was performed primarily via x-ray imaging, and at the time of its discovery it represented one of the best known Paleozoic holocephalans. The classification of the genus has been the subject of debate due to its unique anatomy, and it is now placed in the order Eugeneodontiformes and the family Caseodontidae. Ornithoprion's genus name, which may be translated from Greek as 'bird saw', was inspired by the animal's vaguely bird-like skull and the saw-like appearance of the teeth in the lower jaw, while the specific name honors German zoologist Oscar Hertwig.

Ornithoprion is unique among known eugeneodonts for the extremely long mandibular rostrum extending from the lower jaw, which was covered by a beak of fused bony scales and which the function of in life is not known. It inhabited shallow, seasonal marine and brackish water environments, alongside a variety of other cartilaginous fishes. The rounded shape of Ornithoprion's teeth suggests that it hunted hard-shelled invertebrates, and bite marks and damage to its fossils indicate it was fed on by other carnivores. Ornithoprion was small relative to other members of its order, with a cranium length of up to 10 cm (3.9 in) and an estimated body length of up to approximately 91 cm (36 in).

The genus Ornithoprion and its only species, O. hertwigi, were formally described in 1966 by paleontologist Rainer Zangerl.^[1]: 1 This description was based largely on seven fossils collected from the Mecca Quarry Shale of Parke County, Indiana since the 1950s,^[2][3]: 6 in rocks which are part of the Linton Formation.^{[1]: 10–11 [2]} A single specimen was also described from the Logan Quarry Shale of the Staunton Formation (alternatively the Tradewater Formation),^[4][5] also in Parke County,^{[1]: 10–11 [6][note 1]} which along with the Mecca Quarry material is in the collection of the Field Museum of Natural History.^{[1]: 1 [note 2][2]} Another specimen from near Wilmington, Illinois was also described, although the stratigraphic unit it originated from is unknown and it is part of a private collection. The specimens are all preserved in carbonaceous shale, and the preservation mode of the Illinois specimen has been described as pyritic (composed of the mineral pyrite).^{[1]: 1–11} Specimen FMNH PF-2710 from the Mecca Quarry was designated as the holotype (the defining specimen of the species).^[2] Since Ornithoprion was named, additional fossil specimens from the Excello Shale and Carbondale Formation of Indiana,^[8][9] as well as material from Kansas, have been assigned to the species.^[10]: 39 These specimens are housed in the collections of the American Museum of Natural History and the Natural History Museum, London.^{[10]: 39 [note 3]}

Like many other fish fossils from the Mecca and Logan quarries,^[3][6][13] study of the holotype and paratypes of Ornithoprion were primarily performed by radiographic imaging. The specimens, which are extremely delicate, were not extracted from the surrounding rock matrix and were instead scanned via stereoscopic X-rays to study the hard parts of the body from within the shale.^{[1]: 1–6 [7]} The Staunton Formation specimen, FMNH PF-2656, was also cut into multiple cross-sections, which allowed for study of the internal anatomy of the scales and teeth.^[1]: 3 At the time of its discovery Ornithoprion was one of the few members of its order known from postcranial fossils, alongside the genera Fadenia, Erikodus, and what would later be described as Eugeneodus.^{[14]: 155–156 [note 4]} It also represented one of only a small number of holocephalans from the Paleozoic era in which the endoskeleton was known, and alongside the related Fadenia was unique in preserving the gill arches in detail.^{[14]: 156, 159}

The genus name, Ornithoprion, translates literally from Greek as 'bird saw' and was given in reference to the saw-like row of teeth in the lower jaw and the animal's pointed, beaked skull.^[7] The specific name, hertwigi, honors German zoologist Oscar Hertwig.^[1]: 10

Ornithoprion is known from multiple articulated but incomplete skeletons, and none of the described specimens preserve the body past the pectoral girdle. Most of these are preserved in lateral view, and all, including the holotype, are flattened.^{[1]: 10–11, 18 [2]} Additional postcranial remains, including the tail, are known, but these have not been prepared or described in detail.^[10]: 39 The skeleton of Ornithoprion was composed of cartilage reinforced by an outer coating of mineralized tesserae;^[1]: 30 prismatic structures which strengthen the cartilage skeletons of chondrichthyans.^[15][16] Zangerl noted that O. hertwigi was "very small" in his description of the taxon,^[1]: 6 and in a 2003 book, Richard Ellis suggested Ornithoprion had a total length of 90 cm (3 ft) based on an assumed skull length of less than 15 cm (6 in).^[7]

The lower jaw of Ornithoprion was roughly 1.3 times the length of the rest of the skull,^[17] and had a jointed, forward-facing projection termed the mandibular rostrum.^{[1]: 16–18 [8][14]: 158} While similar rostra are known in several related fish, the structure was significantly longer in Ornithoprion,^[18] and both the rostrum and a correlating section of the snout were uniquely armored by rods of bone embedded in the skin.^[9][19][20] In life, the mandibular rostrum was likely to have been cylindrical in cross section, with the exception of the rear portion which had a small bulging projection that supported the teeth.^{[1]: 18-19 [8]} There is no indication that the mandibular rostrum contained sensory organs.^[6] The rostrum articulated with the Meckel's cartilages (equivalent to the mandible) at a flexible joint, and a flattened keel of cartilage protruded from the bottom of the rostrum near this point of contact. The Meckel's cartilages themselves consisted of a pair of broad, flattened, nearly immobile cartilages,^[1]: 18,41 which articulated with the palatoquadrates (equivalent to the upper jaws).^[8][9][19] Because of the shape of the Meckel's cartilages, the mouth was likely incapable of closing.^[1]: 41

The palatoquadrates were greatly reduced,^{[1]: 6,35 [9]} and potentially fused partially with the cranium.^{[14]: 157–158 [19][21]} This differs greatly from related genera such as Helicoprion, in which the palatoquadrates were large and specialized,^[15][22][23] and potentially Fadenia, which may have had them entirely fused to the cranium or completely lost.^[8][9][18] The condition in O. hertwigi most closely resembled some other related genera in the family Caseodontidae such as Caseodus and Eugeneodus, although the degree of reduction is much greater in Ornithoprion.^[8][18] The palatoquadrate had two points of articulation with the cranium; one with a process which may have been the postorbital process at the back of the skull,^[1]: 12 and one beneath the eye socket.^{[8][14]: 157–158 [24]}

The chondrocranium (or neurocranium) of O. hertwigi had a long, pointed snout and large, high-set eye sockets,^[19] which Zangerl compared to a bird's skull. An indentation set far forward on the snout is reported by Zangerl to have likely held the nasal capsule (skeletal support for the nostrils), although this region of the skull is noted to be poorly known. The brain was small and was positioned along the lower surface of the neurocranium, but little else is known about the cranial nervous system.^{[1]: 12 [14]: 159} Processes on the back of the cranium that Zangerl speculated to be a fused hyoid arch are also known,^{[1]: 12, 16}, although the interpretation of these structures as part of a hyoid arch has subsequently been questioned.^[14]: 157 The largest O. hertwigi cranium measures approximately 10 cm (3.9 in) in length.^[19]

The lower dentition of Ornithoprion consisted of multiple tooth crowns extending from a connected base (or root) along the midline of the lower jaw; an arrangement which resembled a saw and is referred to as a tooth whorl. Additional rows of tightly stacked, flattened teeth were also present, which were likely positioned along the lateral surfaces of the jaws.^{[1]: 8–10 [7][9]} The tooth whorl possessed up to seven broad, rounded, bulbous tooth crowns and was positioned near the point of contact between the Meckel's cartilages and the mandibular rostrum. The tooth crowns of the whorl varied in size, with the smallest teeth being situated at the front and the largest at the back.^{[1]: 24 [9]} The tooth whorl was directly compared with that of the related Erikodus in Zangerl's 1966 description.^[1]: 10,33 The lateral teeth were rod-like and possessed deep pits and grooves in their surface.^[7][9] They formed a brick-like arrangement termed a "tooth pavement" in life similar to that of many other Paleozoic cartilaginous fish.^[8][9]

Additional rows of crushing teeth and larger, pointed V-shaped teeth formed the upper dentition of Ornithoprion. Zangerl, in both his initial description of the taxon and in later works, suggested that these teeth were attached directly to the underside of the cranium,^{[1]: 16 [8][9]} although it has alternatively been suggested they instead attached to a previously unrecognized portion of the palatoquadrates that was fused to the cranium.^{[14]: 158 [19]} The V-shaped teeth are thought to have formed a row along the midline of either the cranium or fused palatoquadrates.^{[1]: 22-23 [14]: 162 [19]}

Based on thin sectioning, the tooth crowns of Ornithoprion are thought to have been composed primarily of trabecular dentin (a spongy form of dentin present in holocephalan fishes)^{[25]: 197–200} with an outer coating of orthodentin.^{[1]: 31 [9][19]} There is no indication of enameloid (also called vitrodentin), but a thin layer may have been present in life.^[8]

Either five or six pairs of gill arches were present in Ornithoprion,^{[10]: 34 [19]} and positioned behind the arches were the scapulocoradoids (pectoral girdles). The paired left and right scapulocoracoids were unfused and had forward-angled scapular portions,^{[1]: 10,21 [14]: 161 [19]} as well as an unpaired cartilaginous structure which has been tentatively identified as sternal cartilage running beneath their coracoidal portions.^{[1]: 21–22 [19][26]} A similar intercoracoidal cartilage has also been identified in living broadnose sevengill sharks,^[27] as well as the extinct iniopterygians and chimaeriform Ischyodus.^[26] The function of this structure in Ornithoprion is unknown, and it may have been homologous with similar, albeit paired cartilage structures known in other eugeneodonts.^[8] While the pectoral fins of Ornithoprion are not known, paleontologist Svend Erik Bendix-Almgreen suggested that they were likely anatomically unique based on the shape of the animal's pectoral girdle.^[14]: 161 There is no indication that the fins supported defensive fin spines.^[8][19]

The vertebral centra of Ornithoprion are not preserved and were uncalcified in life, although a series of cartilaginous structures are present along the path of the vertebral column. Zangerl suggested that these elements, which he described as large and leaf-shaped,^{[8][14]: 160} represented the neural arches of the animal's vertebrae, and he further proposed that their shape may have been an adaptation associated with the unique morphology of the animal's skull.^[1]: 19 The spinal cord of Ornithoprion was sheathed by a flexible notochord.^[19]

The known body of Ornithoprion was completely covered in tiny, tooth-like dermal denticles with rounded crowns.^{[1]: 28-30 [11]} These possessed a pulp cavity, were predominantly made up of orthodentin, and grew from proportionally large, flattened bases. Many of these denticles formed fused, compound structures termed polyodontode scales, which shared a single mushroom-shaped base with upwards of seven crowns emerging from it.^{[1]: 28–30 [8]} Similar polyodontode scales occurred in the related Sarcoprion, and potentially also in Helicoprion.^[11] However, the bases of the denticles in Ornithoprion may have uniquely been composed of bone, rather than a form of dentin like in other cartilaginous fish.^{[1]: 33–36 [19][20]} Extremely small denticles were also present in the mouth and throat of Ornithoprion, which also had crowns composed of orthodentin and bases of bone.^[1]: 32

In his 1966 description, Zangerl speculates that the reinforcing bony rods present on the snout and mandibular rostrum were formed by the compounding and fusion of polyodontode scales.^{[1]: 33 [9]} He likens this phenomenon to that proposed by Oscar Hertwig as an explanation for the origin of vertebrate dermal armor, although Zangerl acknowledges that this adaptation evolved independently in Ornithoprion.^{[1]: 35–36}

Though sometimes referred to as a shark,^{[1]: 1, 6, 16 [7]} Ornithoprion is only a distant relative of living sharks (Selachii). When first described, the genus was classified in the family Edestidae, which itself was traditionally considered a member of the order (sometimes class) Bradyodonti or the equivalent subclass Holocephali.^{[14]: 156 [11][25]: 200–203} In his description of O. hertwigi, however, Zangerl classified it and other edestids as elasmobranchs.^{[1]: 6,33–35} In the 1971 edition of Paleozoic Fishes, researcher Roger S. Miles considered the genus to be of uncertain position within Chondrichthyes, and only tentatively placed it within Holocephali. He suggested that Ornithoprion's similarities with edestids may be the result of convergent evolution (meaning they developed independently) because of differences in the anatomy of the gills, the tooth histology, and the palatoquadrates.^[19] Bendix-Almgreen similarly expressed belief that the features used to unite the edestids may be convergent, and that the family was likely polyphyletic (not a natural group).^{[8][28][29]: 108–109} His conclusions were supported primarily by the apparent presence or absence of enameloid or vitrodentin between different edestid taxa,^[8][30] and differences in the features of their skulls. He considered Ornithoprion to be possibly related only to several genera traditionally classified as edestids from the Late Permian of Greenland, although only very distantly.^[28] In two 1968 publications, both Bendix-Almgreen and paleontologist Colin Patterson also considered the features of Ornithoprion inconsistent with classification as a holocephalan.^{[14]: 163–168 [25]: 192–194}

In a 1981 book, Zangerl considered O. hertwigi a member of the new family Caseodontidae, as part of the larger superfamily Caseodontoidea and the newly established order Eugeneodontiformes (then Eugeneodontida), in light of the numerous new taxa and characteristics that had been observed since Ornithoprion's original description. Ornithoprion's classification within the Caseodontidae is based on the bulbous, rounded nature of its tooth crowns and the reduction of its palatoquadrates, features which are also found in genera such as Caseodus and Erikodus.^[8][9][31] In this publication he again classified the eugeneodonts as members of Elasmobranchii rather than the traditionally assumed Holocephali or Bradyodonti.^[8] While Zangerl's classification of eugeneodonts as elasmobranchs has been refuted by later publications,^[15][22][23] his erected suborders and families within the group remain in use.^[9][10][31] Eugeneodontiformes is regarded as a monophyletic group in the subclass Holocephali (sometimes defined as the more broadly-encompassing Euchondrocephali),^{[9][31][32][note 5]} and discrepancies in tooth histology previously used to argue against their close relation has been alternatively explained by different members' rates of tooth replacement or wear.^[8] The only living members of Holocephali are three families in the order Chimaeriformes, which together are commonly called the chimaeras.^{[29]: 40 [31]} Despite the group's relation, the chimaeras are highly specialized fish that do not closely resemble the eugeneodonts, which instead developed a lifestyle and appearance much closer to sharks.^{[29]: 143–145}

The Mecca Quarry, Logan Quarry, and Excello shales have been dated to the Moscovian stage of the Pennsylvanian subperiod,^{[9][17][34][note 6]} which spanned from 315 to 307 million years ago and which is part of the longer Carboniferous period.^[35][36] The localities were deposited in the North American Midcontinent Sea; a large inland sea which covered much of what is now North America. The localities that preserve Ornithoprion fossils represented marine or brackish water environments along the margin of the sea,^[34] which were inhabited by a diverse assemblage of species.^[37][38][39] In a 1963 paper, Rainer Zangerl and fellow researcher Eugene S. Richardson Jr. proposed that the Mecca and Logan Quarry sites were extremely shallow water habitats, likely less than a meter (3.3 ft) deep, with isolated deeper areas.^{[3]: 30–31, 191 [13][40]} The presence of peat and coal indicates that the deposits overlaid drowned forests, and are the result of a transgression of a marine environment over a terrestrial one (known as a cyclothem). The rich, black shale which encases the fossils indicates large amounts of decaying organic matter, likely from floating algal mats, was present, which led to anoxic (without oxygen) conditions on the seabed and formed organic mud.^{[3]: 120–122 [13]} Zangerl and Richardson suggested that there is evidence of the water levels lowering significantly during periodic dry seasons, often isolating fishes into small ponds termed "fish traps".^{[3]: 221–224 [41][42]} They also hypothesized that these ponds were extremely overcrowded and had an overabundance of predators, and that hunting and killing by predatory fish resulted in the mass death assemblages preserved at the sites.^{[3]: 222–224} The Logan and Mecca shale environments likely only existed for a brief period, with overlying invertebrate communities and limestone deposits indicating that deeper water eventually flooded the region and created a more stable habitat.^[40] Some subsequent authors have suggested that these shales were instead formed in deep-water environments with anoxic mud bottoms, similar to the conditions seen in some other fossiliferous midwestern shales,^[43][44] although other authors have treated the conditions that formed the Mecca and Logan sites as distinct from those that formed deep-sea shales and continued to accept a shallow water environment.^[40][45][46] The Excello Shale had an equivalent depositional environment to the Mecca and Logan Quarry shales.^[13][46][47]

Slabs of shale containing Ornithoprion fossils sometimes also preserve the remains of other animals, although not directly associated with the Ornithoprion remains. These include isolated spines and denticles from acanthodians, Listracanthus, and Petrodus.^[1]: 7,17 The Mecca fauna, which consists of animals from the Mecca and Logan Quarry sites, also preserves an assemblage of conodonts, palaeoniscoids, brachiopods, orthocones,^[3][40] larger predatory chondrichthyans such as Edestus, Orodus, Cobelodus, Symmorium, and Stethacanthus,^[8][37][38] and several members of the order Iniopterygiformes.^[47][48] The Logan Quarry was inhabited by, in addition to chondrichthyans and palaeoniscoids, an unnamed chondrost-like actinopterygian (ray-finned fish) with an elongated rostrum somewhat similar to that of Ornithoprion.^[6][34][49] Invertebrates such as brachiopods and ammonoids are known from the Excello Shale, as are many fish genera that are also part of the Mecca fauna.^[39] Later work by Rainer Zangerl has suggested that many chondrichthyans of the Mecca fauna, namely the iniopterygians, eugeneodonts and cladodonts, were actually pelagic fishes. He suggested they were vagrants that had migrated into shallower waters and became trapped, rather than being native to these habitats.^[47]

Numerous specimens of O. hertwigi show damage which Zangerl interprets as feeding traces left by predators or scavengers. Portions of the skeleton are often broken, maimed or missing, and it is suggested that the unpreserved rear halves of the animals may have been severed by predators. The skulls of several Ornithoprion specimens also display small crushed or missing chunks, which Zangerl said likely resulted from other fishes biting them and fracturing the cartilage.^{[1]: 3–6,11}Almost all other fish skeletons from the Logan Quarry also show similar feeding traces, or are preserved within fossilized regurgitates or coprolites (feces).^{[3]: 136–142,224}

Similar to many of its close relatives,^[18][24][50] Ornithoprion is believed to have been durophagous, with a diet that consisted of shelled invertebrates. The rounded, bulbous crowns of the tooth whorl were likely adaptations for feeding on shelled prey, which would have been crushed between the upper and lower midline tooth rows.^{[1]: 41 [19][21]} The mandibular rostrum is believed to have been utilized in feeding, although the exact mechanism is uncertain. In 1966, Zangerl hypothesized that the structure may have been used to disturb or probe sediment while hunting for prey living on or in the seabed, and potentially to dislodge prey living on the bottom.^[7][19] He notes that this possible feeding mode is entirely speculative,^{[1]: 41–42} although later works agree with the conclusion that the rostra of caseodonts could have been used to probe sediment or dislodge benthic invertebrates such as brachiopods.^[19][50] Some features of Ornithoprion's skull, such as the armor and articulation of the jaws, were suggested by Zangerl to be shock-absorbing adaptations, although he considered it unlikely that the rostrum was a weapon. The mandibular rostrum of Ornithoprion was compared with those of the unrelated extinct bony fish Saurodon and Saurocephalus, in which the function is also unknown.^[1]: 42

• Alienacanthus, an unrelated arthrodire with a similarly elongated mandibular rostrum^[17][51]
• List of prehistoric cartilaginous fish genera

• Original description published in Fieldiana Geology, provided by the Biodiversity Heritage Library