Ornithocheiridae

Holotype specimens of Ornithocheirus simus (A to D) and junior synonym O. platyrhinus (E to H)

The family Ornithocheiridae is without a doubt, one of the most well-known pterosaur groups, mostly due to their very controversial and convoluted taxonomic history. Most of the ornithocheirid fossil record consists of isolated teeth, as well as fragmentary bones, reaching hundreds or even thousands of remains in some localities.[1] The first uncovered ornithocheirid remains were described in 1861 by British paleontologist Sir Richard Owen, who assigned the fossil remains to a new species of Pterodactylus: P. simus.[2] In 1869, British paleontologist Harry Govier Seeley erected the new generic name Ornithocheirus (from Ancient Greek meaning "bird hand"), and assigned P. simus as its type species, therefore creating Ornithocheirus simus. Later, in 1870, Seeley created the name Ornithocheirae to only contain Ornithocheirus.[3] However, in the same year, this was emended to Ornithocheiridae by Seeley himself following the article 11.7.1.3 of the ICZN.[4] In 1874, Owen had proposed two new genera for the Cretaceous British pterosaurs: Coloborhynchus (meaning "maimed beak") and Criorhynchus (meaning "ram beak") based on highly distinctive jaw fragments.[5] Owen reassigned P. simus as the type species of Criorhynchus, creating Criorhynchus simus. He referred three species to Coloborhynchus, including the addition of a new species called C. clavirostris; no type species was designated, however.[6] In 1876, however, Seeley pointed out that Criorhynchus was a junior synonym of Ornithocheirus, a concept that was followed by paleontologist Richard Lydekker in 1888.[5] In the latter year, Lydekker acknowledged that Ornithocheirus simus was the type species of Ornithocheirus, and also distinguished O. simus by its tall rostrum, while other species referred to Ornithocheirus had lanceolate jaw tips. Therefore, to avoid confusion, Lydekker preferred to use the name Criorhynchus for O. simus, and Ornithocheirus for the species with lanceolate jaw tips, a concept later favored by paleontologist Reginald Walter Hooley in 1914. In his review of Ornithocheirus, he divided the family Ornithocheiridae into two subfamilies: Ornithocheirinae and Criorhynchinae; the former consisted of Ornithocheirus and Lonchodectes, while the latter consisted of Amblydectes and Criorhynchus.[7] In his review, Hooley also considered the species Coloborhynchus clavirostris as a synonym of Criorhynchus simus.[4][8] In 1967, paleontologist Oskar Kuhn placed Criorhynchus within the family Criorhynchidae (which is now considered synonymous to Ornithocheiridae), and recognized Ornithocheirus within the family Ornithocheiridae and subfamily Ornithocheirinae. He also designated the species Coloborhynchus clavirostris as the type species of Coloborhynchus, but agreed with Hooley that it was synonymous with Criorhynchus simus.[4][6] In 1994, however, Yuong-Nam Lee revalidated the genus Coloborhynchus (with C. clavirostris as its type species), and regarded it as distinct from Criorhynchus simus.[6] Later, in 2001, paleontologist David Unwin revised the taxonomic history of the Cambridge Greensand pterosaurs, and divided Ornithocheiridae into three genera: Ornithocheirus, Coloborhynchus and Anhanguera. Unwin also designated Ornithocheirus simus as the type and only species of Ornithocheirus.[4]

In 2003, Unwin defined the family Ornithocheiridae as Haopterus gracilis, Ornithocheirus simus, their most recent common ancestor, and all its descendants. He included the genera Anhanguera, Brasileodactylus, Coloborhynchus, Haopterus, Ludodactylus and Ornithocheirus within the family, and also concluded that Araripesaurus, Arthurdactylus and Santanadactylus may belong to this family as well. However, their taxonomic status and precise relationships with other ornithocheirids are uncertain.[9] In 2019, upon the description of the pterosaur Mimodactylus, Haopterus, which was assigned to this family by Unwin,[9] and recovered as a basal eupterodactyloid by Brian Andres and colleagues,[10] was reassigned by Alexander Kellner and colleagues as the sister taxon of the former.[11]

In 2014, Andres and colleagues defined the Ornithocheiridae with a different definition: the most inclusive clade containing Ornithocheirus simus but not Anhanguera blittersdorffi. They placed the genera Coloborhynchus, Ornithocheirus and Tropeognathus within the Ornithocheiridae, while placing Anhanguera within the separate family Anhangueridae.[10] However, back in 2001, Unwin considered the name Anhangueridae a junior synonym of Ornithocheiridae,[5] a concept that was later followed by several paleontologists such as Mark Witton in 2013.[1] Some phylogenetic analyses, however, contradict this name synonymy, with Ornithocheiridae and Anhangueridae classified as different families, therefore following the 2014 analysis by Andres and colleagues instead.[12][13][14] Other studies such as the ones by Kellner and colleagues in 2019 have used a different concept, classifying Coloborhynchus, Tropeognathus, as well as several other close relatives such as Ludodactylus and Caulkicephalus within the Anhangueridae, which, along with the family Hamipteridae, forms the larger group Anhangueria. They assigned Ornithocheirus outside the Anhangueria due to being undiagnosable.[4][11][15] Most recent studies have since followed this concept.[16][17][18]

Skull comparisons between different anhanguerians, notice the ornithocheirids Caulkicephalus (D), Tropeognathus (E and F) and Uktenadactylus (I and J)

Ornithocheirids had elongated jaws with rounded sagittal crests on both tips, as well as robust fang-like teeth.[13][22] The sagittal crest of the species Ornithocheirus simus and Tropeognathus mesembrinus extended to the anterior end of the rostrum, a feature that is also seen in the ornithocheirid (or alternately anhanguerid)[18] species Siroccopteryx moroccensis; further synapomorphies between these three species were also found, including the premaxilla having a tall and narrow shape in anterior aspect, the anterolateral margins of the premaxilla being convex in both anterior and lateral view, a feature that resulted in a bluntly rounded outline of the tip of the rostrum. The rostrum in S. moroccensis lacks a constriction that is posterior to the anterior rosette, a feature also shared by O. simus and T. mesembrinus, therefore another synapomorphy supported by these three species. Yet another feature shared by these three species is that the teeth are short, straight, and relatively uniform in size, something that is not present in other ornithocheirans such as Coloborhynchus and Anhanguera.[13] In Coloborhynchus, the teeth were found to have been heterodont, elongated, recurved and caniniform, which is similar to those seen in another ornithocheirid (though also recovered as an anhanguerid)[16][18] called Caulkicephalus.[23] The genus Caulkicephalus, though having similarities with other ornithocheirids, including the anterior end of the rostrum being transversely expanded, or having a low, bony sagittal crest that includes a smooth dorsal margin on the rostrum, still possesses some unique features. The most distinct characteristic of Caulkicephalus is that it bores a frontoparietal crest, a feature that is only seen in pteranodontians such as Pteranodon, and in Ludodactylus,[24] a pterosaur once assigned to the Ornithocheiridae, but some recent analysis have placed it within the more inclusive group Anhangueria, outside the Ornithocheiridae.[12][14] Other studies, however, have recovered it within the Anhangueridae instead.[11][16][18]

Restoration of three Tropeognathus in flight, notice their high aspect ratio

The forelimbs of ornithocheirids were proportionally enormous, around five times longer than their legs. Substantial anchorage on the body is required given the mighty arms, and accordingly, ornithocheirids have robust scapulocoracoids, and stout, deeply keeled sterna, which served the purpose of housing their substantial forelimb muscles. The shoulder or pectoral girdle in ornithocheirids is set at a perpendicular angle to the spine, with the coracoids being much longer than the scapulae. The shoulder girdle is also of typical construction for ornithocheiroids. Over 60 percent of the wing length is occupied by the wing fingers, making them among the longest possessed by any pterodactyloids.[25]

In adult ornithocheirids, the sacrum develops a supraneural plate above its neural spines. The tails of ornithocheirids are poorly known, though they appear to be composed of at least eleven short vertebrae, and become relatively circular in cross section toward the end of the series. Like the related istiodactylids, the slender femora of ornithocheirids have femoral heads that project almost in line with the femoral shaft, but seem to lack prominent processes that anchor their hindlimb muscles. Ornithocheirid shinbones (or tibiae) are similarly developed and of equal length to the femora. Although the feet in ornithocheirids are poorly known, they seem to be relatively small and gracile, with undeveloped claws and a hook-like fifth metatarsal.[26]

In the past, many pterosaur genera were assigned to the Ornithocheiridae; however, following recent studies, these supposed ornithocheirids had been reclassified to other groups or families. The pterosaur Boreopterus for example, was initially classified within the Ornithocheiridae; however, later analysis had found it in a different family called Boreopteridae, with Boreopterus being the sister taxon of Zhenyuanopterus; both pterosaurs were recovered as basal members of the Lanceodontia in several recent studies such as the ones by Rodrigo Pêgas and colleagues,[17] and Adele Pentland and colleagues, both studies of which are from 2019.[14]

Aetodactylus is another pterosaur that was initially classified within the Ornithocheiridae; later analysis have found it outside the Ornithocheiridae: Timothy Myers in 2015 and Nicholas Longrich and colleagues in 2018 for example, are two studies that found Aetodactylus as sister taxon to the two species of Cimoliopterus (C. cuvieri and C. dunni).[12][27] In 2019, Pêgas and colleagues have found Aetodactylus, along with two other pterosaurs (Camposipterus and Cimoliopterus), within the clade Targaryendraconia, more specifically placed within the family Cimoliopteridae in a polytomy.[17]

The genus Haopterus was used to define the Ornithocheiridae in Unwin's 2003 study; however, Andres and Myers in 2013 argued that Haopterus had not been previously referred to the Ornithocheiridae besides a note added in proof to Unwin in 2001 that stated that Haopterus appeared to be a small ornithocheirid. Phylogenetic analyses since then have found Haopterus as a rogue taxon either within the Pterodactyloidea, the Ornithocheiroidea, the Pteranodontoidea, or the Istiodactylidae. In the phylogenetic analysis by Andres and Myers, Haopterus was recovered as a stable sister taxon to the group Ornithocheiroidea.[28] Some later analyses have also recovered this concept, with both Haopterus and the Ornithocheiroidea placed within the larger group Eupterodactyloidea.[10][12][29] Another study in 2019 recovered Haopterus within a different group called Mimodactylidae.[11]

Different phylogenetic analysis have found Ornithocheiridae to comprise different genera, the most typical ones being Tropeognathus, Coloborhynchus and Ornithocheirus.[27][28] In 2014, Andres and colleagues created the subfamily Ornithocheirinae to contain Coloborhynchus and Ornithocheirus, as a sister taxon to Tropeognathus, and altogether formed the family Ornithocheiridae.[10] In 2018, Longrich and colleagues had included the genus Siroccopteryx in their phylogenetic analysis, specifically as a member of the ornithocheirine subfamily, sister taxon to Coloborhynchus.[12] In 2019, a study performed by Adele Pentland and colleagues had found the Ornithocheiridae to comprise more genera; while the typical Tropeognathus, Coloborhynchus and Ornithocheirus clade was included in their analysis, Pentland and colleagues found the genera Ferrodraco and Mythunga to belong the Ornithocheiridae as well, specifically sister taxa within the Ornithocheirinae, closely related to Ornithocheirus.[14] In the same year, Megan Jacobs and colleagues have recovered a different set of relationships for ornithocheirids in their analysis. Within the family, three clades emerge: the first one consists of Ornithocheirus, Tropeognathus and Siroccopteryx, the second one comprises Uktenadactylus with several Coloborhynchus species, and the third one comprises Cimoliopterus and Camposipterus. The close relationship between Siroccopteryx, Ornithocheirus and Tropeognathus is supported by several synapomorphies, such the teeth being short, straight, and relatively uniform in size.[13] Several other recent studies such as the ones by Alexander Kellner and colleagues,[11][15] or the one by Rodrigo Pêgas and colleagues, both in 2019, have recovered Coloborhynchus, Siroccopteryx, Tropeognathus and Uktenadactylus within the family Anhangueridae instead of the Ornithocheiridae.[17] In 2020, a study by Borja Holgado and Pêgas had also recovered both Ferrodraco and Mythunga within the Anhangueridae instead of this family.[18]

Cimoliopterus has generally been recovered outside the Ornithocheiridae; subsequent analyses have found it as sister taxon to Aetodactylus, as mentioned earlier.[12][27] Another possible position for Cimoliopterus is within the clade Targaryendraconia, again, closely related to Aetodactylus, and together with Camposipterus, the three formed the family Cimoliopteridae.[17] In the analysis by Jacobs and colleagues, the two Cimoliopterus species had been found as sister taxa to the three Camposipterus species (C. nasutus, C. colorhinus and C. segwickii), altogether formed an unnamed clade within the Ornithocheiridae. However, as noted by Jacobs and colleagues, support for some of these arrangements is relatively weak due to the limited number of characters that can be scored, and the levels of homoplasy are very high.[13]

Below are two cladograms showing the possible genera that are included within the ornithocheirid family. The cladogram to the left is a topology recovered by Jacobs and colleagues,[13] and the one to the right is a topology recovered by Pentland and colleagues.[14]

Restoration of Cimoliopterus stealing prey from a Lonchodectes

Ornithocheirids are generally considered piscivorous animals, mainly because they seem to have been suited for flight over marine settings;[30] in fact, most ornithocheirids are known from lagoonal, coastal and marine deposits.[31] Although the manner in which ornithocheirids gathered their food has not been researched in detail, it is generally thought that members of this family either fed like modern-day skimmers, pushing their lower jaw through the water to snap up food upon impact, or fed by gleaning food from the water surface like some modern-day terns and frigatebirds. The skim-feeding hypothesis on ornithocheirids has been discounted in recent assessments of pterosaur skim-feeding, while dip-feeding is supported by a number of anatomical features. The elongated rostra of ornithocheirids are considered ideal for reaching into the water to grab swimming creatures; the rostral crests of ornithocheirids would have worked well as stabilizers for the jaws tips while being plunged into the water. Large, forward-facing eyes and well-developed flocculi are ideal for dip-feeding as well, which permits effective spotting of prey as well as judgement of distances when striking at them; as such, it seems likely that at least several ornithocheirids were efficient dip feeders. Sedate foraging methods might have also been used when hunting -- examples of these methods are: reaching food while being alighted on the water surface, and shallow surface dives.[30]

Similar to modern-day albatrosses, most ornithocheirids used a flight technique called "dynamic soaring", which consists of travelling long distances without flapping using the vertical gradient of wind speed near the ocean surface as an advantage, at moderate flight speed. Several studies showed that most ornithocheirids sprawled their limbs to a large degree, similar to crocodiles, while other studies conclude that ornithocheirids were generally quadrupedal. Yet other studies concluded that ornithocheirids held their limbs more or less vertically extended, similar to an avian or mammalian configuration.[32] Some studies in later genera show that ornithocheirids spend much of their time at sea, and as a result, they perhaps influenced the later pteranodontids with the same piscivorous diet, as well as their well-developed flight techniques.[33][34] Analyses of limb proportions in the genus Anhanguera, however, show that some ornithocheirids were consistent with hopping, but the later genera were suggested that they most likely walked on four limbs, which consists on their wing-fingers as the front limbs, and using their hind limbs to balance.[35]