Dilophosaurus

Dilophosaurus Temporal range: Early Jurassic,
Reconstructed cast of the holotype specimen (UCMP 37302) in burial position, Royal Ontario Museum
Scientific classification
Domain:	Eukaryota
Kingdom:	Animalia
Phylum:	Chordata
Clade:	Dinosauria
Clade:	Saurischia
Clade:	Theropoda
Genus:	†Dilophosaurus Welles, 1970
Species:	†D. wetherilli
Binomial name
†Dilophosaurus wetherilli (Welles, 1954)
Synonyms
Megalosaurus wetherilli Welles, 1954

Dilophosaurus (/daɪˌloʊfəˈsɔːrəs, -foʊ-/ dy-LOHF-o-SOR-əs) is a genus of theropod dinosaur that lived in what is now North America during the Early Jurassic Period, about 193 million years ago. In 1942, three skeletons were found in Northern Arizona, the two most complete of which were collected. The most complete of these was made the holotype specimen of a new species of the genus Megalosaurus, named M. wetherilli by Samuel Welles in 1954. Welles found a larger skeleton belonging to the same species in 1964, and realized it bore crests on its skull, and therefore moved the species to the new genus Dilophosaurus in 1970. The genus name means "two-crested lizard", and the species name honors John Wetherill, a Navajo councilor. More specimens have since been found, including an infant, and footprints have also bee attributed to the animal. Another species from China, D. sinensis, was named in 1993, but was later found to belong to the genus Sinosaurus.

At about 7 meters (23 ft) in length, with a weight of about 400 kilograms (880 lb), Dilophosaurus was one of the first big predatory dinosaurs, though it was smaller than later theropods. It was lightly built and slender, and though the skull was proportionally large, it was delicate. The snout was narrow, and the upper jaw had a gap below the nostril. It had a pair of longitudinal crests on its skull, similar to a cassowary with two crests. The mandible was slender and delicate at the front, but deep at the back. The teeth were long, curved, thin, and compressed sideways. The teeth of the lower jaw were much smaller than those of the upper jaw. The teeth had serrations at their front and back edges, and were covered in enamel. The neck was long, and its vertebrae were hollow, and very light. The arms were powerful, with a long and slender upper arm bone. The hands had four fingers; the first was short but strong and bore a large claw, the two following fingers were longer and more slender with smaller claws, and the fourth was vestigial. The feet were stout and the toes bore large claws.

Description

Size of two specimens compared to a human, with holotype in green

Dilophosaurus has been described as a medium-sized theropod. It was one of the first big predatory dinosaurs, though small compared to some of the later theropods. It was lightly built, slender, and has been compared to a brown bear in size. The largest known specimen measured about 7 meters (23 ft) in length, its skull was 590 millimeters (1.94 ft) long, and it weighed about 400 kilograms (880 lb). The smaller holotype specimen was about 6.03 meters (19.8 ft) long, with a hip height of about 1.36 meters (4.5 ft), its skull was 523 millimeters (1.716 ft) long, and it weighed about 283 kilograms (624 lb). A resting trace of a large theropod that may be Dilophosaurus or a similar theropod has markings around the belly and feet that have been interpreted as feather impressions by some researchers. Others have concluded that the markings are sedimentological artifacts, while noting that this interpretation does not rule out that the track maker bore feathers.

Dilophosaurus had 9 cervical (neck) vertebrae, 5 pectoral vertebrae (the vertebrae that transition between those of the neck and the back), 10 dorsal (back) vertebrae, and 45 caudal (tail) vertebrae. It had a long neck, which was probably flexed nearly 90° by the skull and by the shoulder, thereby holding the skull in a horizontal posture. The cervical vertebrae were unusually light; their centra (the "bodies" of the vertebrae) were hollowed out by pleurocoels (depressions on the sides) and centrocoels (cavities on the inside). The arches of the cervical vertebrae also had chonoses, conical recesses so large that the bones separating them were sometimes paper-thin. The centra were plano-concave, flat at the font and cupped (or concave) at the back. This indicates that the neck was flexible, though it had long, overlapping cervical ribs, which were fused to the centra. The cervical ribs were slender, and may therefore have bent easily. The atlas bone (the first cervical vertebra which attaches to the skull) had a small, cubic centrum, and had a concavity at the front where it formed a cup for the occipital condyle (protuberance which connects with the atlas vertebra) at the back of the skull. The axis bone (the second cervical vertebra) had a heavy spine, and its postzygapophyses (articular processes) were met by long prezygapophyses that curved upwards from the third cervical vertebra. The centra and spines of the cervical vertebrae were long and low, and the spines had caps which gave the appearance of a Maltese cross when seen from above. The neural spines of the dorsal vertebrae were also low and expanded front and back, which formed strong attachments for ligaments. There were four sacral vertebrae (dorsal vertebrae between the hips), which occupied the length of the ilium blade, and these did not appear to be fused. The centra of the caudal vertebrae were very consistent in length, but their diameter became smaller towards the back, and they went from elliptical to circular in cross-section.

Diagram showing the forelimb of Dilophosaurus in hypothesized resting posture

The scapulae (shoulder blades) were moderate in length, wide (particularly the upper part), and were concave on their inner sides to follow the body's curvature. The coracoids were elliptical, and not fused to the scapulae. The arms were powerful, and had deep pits and stout processes for attachment of muscles and ligaments. The humerus (upper arm bone) was large and slender, with stout epipodials, and the ulna (lower arm bone) was stout and straight, with a stout olecranon. The hands had four fingers; the first was shorter but stronger than the following two fingers, with a large claw, and the two following fingers were longer and more slender, with smaller claws. The third finger was reduced, and the fourth was vestigial (was retained, but without function). The crest of the ilium was highest over the peduncle, and its outer side was concave. The foot of the pubic bone was only slightly expanded, whereas the lower end was much more expanded on the ischium, which also had a very thin shaft. The hind-legs were large, with a slighter longer femur (thigh bone) than tibia (lower leg bone), the opposite of for example Coelophysis. The femur was massive, its shaft was sigmoid-shaped, and its greater trochanter was centered on the shaft. The tibia had a developed tuberosity and was expanded at the lower end. The astragalus bone (ankle bone) was separated from the tibia and the calcaneum, and formed half of the socket for the fibula. It had long, stout feet with three well-developed toes that bore large claws. The third toe was the stoutest, and the smaller first toe (the hallux) was kept off the ground.

Skull

Reconstructed skull and neck, Royal Tyrrell Museum; the crests and the gap in the upper jaw are distinguishing features of this dinosaur

The skull of Dilophosaurus was large in proportion to the skeleton, yet delicate. The snout was narrow in front view, and narrowing towards the rounded top. The premaxilla (front bone of the upper jaw) was long and low when seen from the side, and its outer surface became less convex from snout to naris (bony nostril). The nostrils were placed further back than in most other theropods. The premaxilla was weakly attached to the maxilla (the following bone of the upper jaw), only connecting at the middle of the palate, with no connection art the side. Hindwards and below, the premaxilla formed a wall for a gap between itself and the maxilla, called the subnarial gap (also called a "kink" or "notch"). Within the gap was a deep excavation behind the toothrow of the premaxilla, called the subnarial pit, which was walled by a downwards keel of the premaxilla. The outer surface of the premaxilla was covered in foramina (openings) of varying sizes. The nasal process (backwards projection from the premaxilla) was long and low, and formed most of the upper border of the elongated naris. It had a dip towards the font, which made the area by its base concave in profile. Seen from below, the premaxilla had an oval area which contained alveoli (tooth sockets). The maxilla was shallow, and was depressed around the antorbital fenestra (a large opening in front of the eye), forming a recess that was rounded towards the front, and smoother than the rest of the maxilla. A foramen called the preantorbital fenestra opened into this recess at the front bend. Large foramina ran on the side of the maxilla, above the alveoli. A deep nutrient groove ran backwards from the subnarial pit along the base of the interdental plates (or rugosae) of the maxilla.

Dilophosaurus bore two high and thin crests longitudinally on the skull roof, formed by the nasal and lacrimal bones. The paired crests extended upwards and gave the appearance of a cassowary with two crests. Each crest also had a finger-like backwards projection. The upper surface of the nasal bone between the crests was concave, and the nasal part of the crest overlapped the lacrimal part. As only one specimen preserves the shape of the crests, it is unknown if they differed in other individuals. The lacrimal bone had a thickened upper rim, where it formed the upper border at the back of the antorbital fenestra. The prefrontal bone formed the roof of the orbit (eye socket), and had an L-shaped bar that made part of the upper surface of the orbit concave. The orbit was oval, and narrow towards the bottom. The jugal bone had two upwards pointing processes, the first of which formed part of the lower margin of the antorbital fenestra, and part of the lower margin of the orbit. A projection from the quadrate bone into the lateral temporal fenestra (opening behind the eye) gave this a reniform (kidney-shaped) outline. The foramen magnum (the large opening at the back of the braincase) was about half the breadth of the occipital condyle, which was itself cordiform (heart-shaped), and had a short neck and a groove on the side.

Restoration showing hypothetical feathers and crest-shape

The mandible was slender and delicate at the front, but the articular region (where it connected with the skull) was massive, and the mandible was deep around the mandibular fenestra (an opening on its side). The retroarticular process of the mandible (a backwards projection) was long, the surangular shelf was strongly horizontal. The dentary bone (the front part of the mandible where most of the teeth there attached) had an up-curved rather than pointed chin. The chin had a large foramen at the tip, and a row of small foramina ran in rough parallel with the upper edge of the dentary. On the inner side, the mandibular symphysis (where the two halves of the lower jaw connected) was flat and smooth, and showed no sign of being fused with its opposite half. A Meckelian foramen ran along the outer side of the dentary.

Dilophosaurus had 4 teeth in each premaxilla, 12 in each maxilla, and 17 in each dentary. The teeth were generally long, thin, and recurved, with relatively small bases. They were compressed sideways, oval in cross-section at the base, lenticular (lens-shaped) above, and slightly concave on their outer and inner sides. The teeth of the dentary were much smaller than those of the maxilla. The teeth had serrations on the front and back edges, which were offset by vertical grooves, and were smaller at the front. There were 31 to 41 serrations on the front edges, and 29 to 33 on the back. The teeth were covered in a thin layer of enamel, 0.1 to 0.15 millimeters (0.0039 to 0.0059 in), which extended far towards their bases. The alveoli were elliptical to almost circular, and all were larger than the bases of the teeth they contained, which may therefore have been loosely held in the jaws. Though the number of alveoli in the dentary indicates the teeth were very crowded, they were rather far apart, due to the larger size of their alveoli. The jaws contained replacement teeth at various stages of eruption. The interdental plates between the teeth were very low.

History of discovery

Front part of holotype cast, ROM; the skull, crests, and pelvis are reconstructed, the neck straightened, the left leg moved up, and the tail has been curved upwards. The rest of the skeleton is in the position of burial

In the summer of 1942, the American paleontologist Charles L. Camp led a field party from the University of California Museum of Paleontology in search of fossil vertebrates in the Navajo County of Northern Arizona. Word of this was spread among the native Americans there, and the Navajo Jesse Williams brought three members of the expedition to some fossil bones he had discovered in 1940. The area was part of the Kayenta Formation, about 32 kilometers (20 mi) north of Cameron near Tuba City in the Navajo Indian Reservation. Three dinosaur skeletons were found arranged in a triangle, about 9.1 meters (30 ft) long at one side, in purplish shale. The first was nearly complete, and lacked only the front of the skull, parts of the pelvis, and some vertebrae, the second was very eroded, and included the front of the skull, lower jaws, some vertebrae, limb bones, and an articulated hand, and the third was so eroded that it consisted only of vertebra fragments. The first, good skeleton was encased in a block of plaster after ten days of work and loaded onto a truck, the second skeleton was easily collected as it was almost entirely weathered out of the ground, but the third skeleton was almost gone.

The nearly complete first specimen was cleaned and mounted at the UCMP under supervision of the American paleontologist Wann Langston Jr., a process that took three men two years. The skeleton was wall mounted in bas relief, with the tail curved upwards, the neck straightened, and the left leg moved up for visibility, but the rest of the skeleton was kept in the position of burial. As the skull was crushed, it was reconstructed based on the back of the skull of the first specimen and the front of the second. The pelvis was reconstructed after that of Allosaurus, and the feet were also reconstructed. At the time, it was one of the best preserved skeletons of a theropod dinosaur, though incomplete. In 1954, the American paleontologist Samuel P. Welles, who was part of the group that excavated the skeletons, preliminarily described and named this dinosaur as a new species in the existing genus Megalosaurus; M. wetherilli. The nearly complete specimen (catalogued as UCMP 37302) became the holotype, and the second specimen (UCMP 37303) was included in the hypodigm (the sample that defines a taxon) of the species. The specific name honored John Wetherill, a Navajo councilor who Welles described as an "explorer, friend of scientists, and trusted trader". Wetherill's nephew, Milton, had first informed the expedition of the fossils. Welles placed the new species in Megalosaurus due to the similar limb proportions of it and M. bucklandii, and because he did not find great differences between them. At the time, Megalosaurus was used as a "wastebasket taxon", wherein many species of theropod were placed, regardless of their age or locality.

Reconstructed skull, showing the two crests that were the basis of the name Dilophosaurus, American Museum of Natural History

Welles returned to Tuba City in 1964 to determine the age of the Kayenta Formation (it had been suggested to be Late Triassic in age, whereas Welles thought it was Early to Middle Jurassic), and discovered another skeleton about 402.3 kilometers (250.0 mi) south of where the 1942 specimens had been found. The nearly complete specimen (catalogued as UCMP 77270) was collected with the help of William Breed of the Museum of Northern Arizona and others. During preparation of this specimen, it became clear that it was a larger individual of M. wetherilli, and that it would have had two crests on the top of its skull. Being a thin plate of bone, one crest was originally thought to be part of the missing left side of the skull which had been pulled out of its position by a scavenger. When it became apparent it was a crest, it was also realized that there would have been a corresponding crest on the left side, since the right crest was right of the midline, and was concave along its middle length. This discovery led to reexamination of the holotype specimen, which was found to have bases of two thin, upwards extended bones, which were crushed together. These also represented crests, but it had formerly been assumed they were part of a misplaced cheek bone. It was also concluded that the two 1942 specimens were juveniles, while the 1964 specimen was an adult, about one third larger than the others.

Reconstructed skeleton at RTM (based on the holotype), with outdated (pronated) hand pose

Welles and an assistant subsequently corrected the wall mount of the holotype specimen based on the new skeleton, by restoring the crests, redoing the pelvis, and making the neck ribs longer, and placed closer together. After studying the skeletons of North American theropods, and seeing most of the material in western Europe in 1969, Welles realized that the dinosaur did not belong to Megalosaurus, and needed a new genus name. At this time, no other theropods with large longitudal crests on their heads were known, and the dinosaur had therefore gained interest by paleontologists. A mold of the holotype specimen was made, and fiberglass casts of it were distributed to various exhibits; to make labeling these casts easier, Welles decided to name the new genus in a brief note, rather than wait until the publication of a detailed description. In 1970, Welles coined the new genus name Dilophosaurus, from the Greek words di (δι) meaning "two", lophos (λόφος) meaning "crest", and sauros (σαυρος) meaning "lizard"; "two-crested lizard". Welles published a detailed osteological description of Dilophosaurus in 1984, but the 1964 specimen has not yet been described. Dilophosaurus was the first well-known theropod from the Early Jurassic, and remains one of the best preserved examples of that age.

In 2001, the American paleontologist Robert J. Gay identified the remains of at least three new Dilophosaurus specimens (this number is based on the presence of three pubic bone fragments and two differentially sized femora) in the collections of the Museum of Northern Arizona. The specimens were found in 1978 in the Rock Head Quadrangle, 190 kilometers (120 mi) away from where the original specimens were found, and had been labeled as a "large theropod". Though most of the material is damaged, it is significant in including elements not preserved in the earlier known specimens, including part of the pelvis, and some ribs. Some elements in the collection belonged to an infant specimen (MNA P1.3181), the youngest known example of this genus, and one of the earliest known infant theropods from North America, only preceded by some Coelophysis specimens. The juvenile specimen includes a partial humerus, a partial fibula, and a tooth fragment. In 2005, the American paleontologist Ronald S. Tykoski assigned a specimen (TMM 43646-140) from Gold Spring, Arizona, to Dilophosaurus, but in 2012 the American paleontologist Matthew T. Carrano and colleagues found it to differ in some details.

Attributed track in Red Fleet Dinosaur Tracks Park, Utah

In 1971 Welles reported dinosaur footprints from the Kayenta Formation, 14 meters (45 ft) and 112 meters (367 ft) below the level where the original Dilophosaurus specimens were found. The lower footprints were tridactyl (three-toed), and could have been made by Dilophosaurus; Welles created the new ichnogenus Dilophosauripus williamsi based on them, in honor of Jesse Williams, the discoverer of the first Dilophosaurus skeletons. The type specimen is a cast of a large footprint catalogued as UCMP 79690-4, with casts of three other prints included in the hypodigm. In 1984 Welles conceded that there was no way to prove or disprove that the footprints belonged to Dilophosaurus. Tracks of Eubrontes and Gigandipus of the Connecticut River Valley, which have been found in both Connecticut and in Massachusetts, have often been attributed to Dilophosaurus, however no fossil remains of Dilophosaurus have been directly attributed to either one of the footprint types. The size and shape suggested they were made by a theropod around 20 feet long similar to that of Dilophosaurus, suggesting they were either made by Dilophosaurus or a very close relative. Two similar footprints, Anchisauripus and Grallator, also found in the valley, have often been attributed to Dilophosaur's smaller relatives, Podokesaurus and Coelophysis, and occasionally attributed to Dilophosaurus itself.

In 1991, the Polish paleontologist Gerard Gierlinski examined tridactyl footprints from the Holy Cross Mountains in Poland and concluded they belonged to a theropod like Dilophosaurus. He coined the new ichnospecies Grallator (Eubrontes) soltykovensis based on them, with a cast of footprint MGIW 1560.11.12 as the holotype. In 2009 the American paleontologists Andrew R. C. Milner and colleagues used the combination Kayentapus soltykovensis instead. They also suggested that Dilophosauripus may not be distinct from Eubrontes and Kayentapus, as the long claw marks that distinguish Dilophosauripus may be an artifact of dragging. They found Dilophosaurus a suitable match for a trackway and resting trace (SGDS.18.T1) from the St. George Dinosaur Discovery Site in the Moenave Formation of Utah, though the dinosaur is not itself known from the formation, which is slightly older than the Kayenta Formation.

Formerly assigned species

Reconstructed skeleton of Sinosaurus sinensis, which was originally described as a species of Dilophosaurus, in MUSE, Trento

In 1984, Welles suggested that the 1964 specimen (UCMP 77270) did not belong to Dilophosaurus after all, but to a new genus, based on differences in the skull, vertebrae, and femora. He maintained that both genera bore crests, but that the exact shape of these was unknown in Dilophosaurus. Welles died in 1997, before he could name this supposed new dinosaur, but the idea that the two were separate genera has generally been ignored or forgotten since. In 1999, the American amateur paleontologist Stephan Pickering privately published the new name Dilophosaurus "breedorum" based on this specimen, named in honor of Breed, who assisted in collecting it. This name is considered a nomen nudum, an invalidly published name, and Gay pointed out in 2005 that there are no significant differences between D. "breedorum" and other D. wetherilli specimens. Carrano and colleagues found differences between UCMP 77270 and the holotype specimen, but attributed them to variation between individuals rather than species.

In 1987, a nearly complete theropod skeleton (KMV 8701) was discovered in the Lufeng Formation, in Yunnan Province, China. It was similar to Dilophosaurus in having a pair of crests and a gap that separated the premaxilla from the maxilla, but differed in some details. The Chinese paleontologist Shaojin Hu therefore named it as a new species of Dilophosaurus in 1993, D. sinensis (from Greek sino, pertaining to China). In 1998, the American paleontologist Matthew C. Lamanna and colleagues found D. sinensis to be identical to Sinosaurus triassicus, a theropod from the same formation, named in 1940. This conclusion was confirmed by the Chinese paleontologist Lida Xing and colleagues in 2013, and though the Chinese paleontologist Guo-Fu Wang and colleagues agreed the species belonged in Sinosaurus in 2017, they suggested it may be as a separate species, S. sinensis.

Classification

Skeleton of Coelophysis, which Dilophosaurus was often grouped with, Cleveland Museum of Natural History

Though Welles originally thought Dilophosaurus to be a megalosaur in 1954, he realized this to be incorrect in 1970, after discovering it had crests, at the time the only theropod with such a feature. In 1984, Welles found that Dilophosaurus exhibited features of both coelurosauria and carnosauria, the two main groups theropods had hitherto been divided into, based on body size, and he suggested this division were inaccurate. He found Dilophosaurus to be closest to those theropods that were usually placed in the family Halticosauridae, particularly Liliensternus.

In 1988, the American paleontologist Gregory S. Paul classified the halticosaurs as a subfamily of the family Coelophysidae, and suggested that Dilophosaurus could have been a direct descendant of Copelophysis. Paul also considered the possibility that spinosaurs were late surviving dilophosaurs, based on similarity of the kinked snout, nostril position, and slender teeth of Baryonyx. In 1994, the American paleontologist Thomas R. Holtz placed Dilophosaurus in the group Coelophysoidea, along with but separate from Coelophysidae. He placed Coelophysoidea itself in the group Ceratosauria. In 2000, the American paleontologist James H. Madsen and Welles divided Ceratosauria into the families Ceratosauridae and Dilophosauridae, with Dilophosaurus as the sole member of the latter family.

Reconstructed skeleton of the crested theropod Cryolophosaurus, once considered a close relative of Dilophosaurus, Science World

In 1998, Lamanna and colleagues pointed out that since Dilophosaurus was found to have crests on its skull, other similarly crested theropods have been discovered (including Sinosaurus), and that this feature is therefore not unique to the genus, and of limited use for determining interrelationships within their group. In 2005, the Australian paleontologist Adam M. Yates described the new genus Dracovenator from South Africa, and found it closely related to Dilophosaurus and Zupaysaurus. His cladistic analysis suggested they did not belong in Coelophysoidea, but rather in Neotheropoda, a more derived (or "advanced") group. He proposed that if Dilophosaurus was more derived than Coelophysoidea, the features it shared with this group may have been inherited from basal (or "primitive") theropods, indicating that theropods may have passed through a "coelophysoid stage" in their early evolution.

In 2007, the American paleontologist Nathan D. Smith and colleagues found the crested theropod Cryolophosaurus to be the sister species of Dilophosaurus, and grouped them with Dracovenator and Sinosaurus. This clade was more derived than Coelophysoidea, but more basal than Ceratosauria, thereby placing basal theropods in a ladder-like arrangement. In 2012, Carrano and colleagues found that the group of crested theropods proposed by Smith and colleagues was based on features that relate to the presence of such crests, but that the features of the rest of the skeleton were less consistent. They instead found that Dilophosaurus was a Coelophysoid, with Cryolophosaurus and Sinosaurus being more derived, as basal members of the group Tetanurae.

In 2015, the French paleontologist Christophe Hendrickx and colleagues defined Dilophosauridae to include Dilophosaurus and Dracovenator, and noted that while there is general uncertainty about the placement of this group, it appears to be slightly more derived than Coelophysoidea, and the sister group to Averostra. Dilophosauridae shares features with Coelophysoidea such as the subnarial gap and the front teeth of the maxilla pointing forwards, while features shared with Averostra include a fenestra at the front of the maxilla and a reduced number of teeth in the maxilla. They suggested that the cranial crests of Cryolophosaurus and Sinosaurus had either evolved convergently, or was a feature inherited by a common ancestor. The following following cladogram is based on that published by Hendrickx and colleagues, itself based on earlier studies:

Skull bones of Dracovenator, which may be the closest relative of Dilophosaurus (on whose skull the bones are plotted), ROM

Neotheropoda	Coelophysidae Liliensternus Zupaysaurus Dilophosauridae Dilophosaurus Dracovenator Averostra Ceratosauria Tetanurae Cryolophosaurus Sinosaurus Monolophosaurus Orionides

Paleobiology

Restoration of Dilophosaurus in resting pose, based on tracks at the St. George Dinosaur Discovery Site, Utah

Welles envisioned Dilophosaurus as an active, clearly bipedal animal, similar to an enlarged ostrich. He found the forelimbs to have been powerful weapons, strong and flexible, and not used for locomotion. He noted that the hands were capable of grasping and slashing, of meeting each other, and reaching two thirds up the neck. He proposed that during a sitting posture, it would rest on the large "foot" of its ischium, as well as its tail and feet. In 1990, the American paleontologists Stephen and Sylvia Czerkas suggested that the weak pelvis of Dilophosaurus could have been an adaptation for an aquatic life-style, where the water would help support its weight, and that it could have been an efficient swimmer. They found it doubtful that it would have been restricted to a watery environment, though, due to the strength and proportions of it hind-limbs, which would have made it fleet-footed and agile during bipedal locomotion.

In 2005, the American paleontologists Phil Senter and James H. Robins examined the range of motion in the forelimbs of Dilophosaurus and other theropods. They found that Dilophosaurus would have been able to hold its humerus almost parallel with its scapula, but not more horizontal than that. The elbow could approach full extension, but not achieve it completely, and not flex at a right angle. The fingers do not appear to have been voluntarily hyperextensible (able to extend backwards, beyond their normal range), but they may have been passively hyperextensible, to resist dislocation during violent movements by captured prey. A 2015 article by Senter and Robins gave recommendations of how to reconstruct forelimb pose in bipedal dinosaurs, based on examination of various taxa, including Dilophosaurus. The scapulae were held very horizontally, the resting orientation of the elbow would have been close to a right angle, and the orientation of the hand would not have deviated much from that of the lower arm.

Welles interpreted the fact that three individuals were found closely together, and the presence of criss-crossed trackways nearby as indications that the animal traveled in groups. Gay agreed that they may have traveled in small groups, but noted that there was no direct evidence for this, and that flash floods could have picked up scattered material from different individuals and deposited them together.

Feeding and diet

Skull diagrams showing closed and open jaws, as well as hypothetically reconstructed crest-shape

Welles found that Dilophosaurus did not have a poweful bite, due to weakness caused by the subnarial gap. He thought that it used its front premaxillary teeth for plucking and tearing rather than biting, and the maxillary teeth further back for piercing and slicing. He thought that it was probably a scavenger rather than a predator, and that if it did kill large animals, it would have done so with its hands and feet rather than its jaws. Welles did not find evidence of cranial kinesis in the skull of Dilophosaurus, which allows individual skull-bones of the skull to move in relation to each other.

In 1986, the American paleontologist Robert T. Bakker instead found Dilophosaurus, with its massive neck and skull and large upper teeth, to have been adapted for killing large prey, and strong enough to attack any Early Jurassic herbivores. In 1988, Paul dismissed the idea that Dilophosaurus was a scavenger, and claimed that scavenging terrestrial animals are a myth. He stated that the snout of Dilophosaurus was better braced than what had been thought previously, and that the very large, slender maxillary teeth were more lethal than the claws. Paul suggested that it hunted large animals such as prosauropods, and that it was more capable of snapping small animals than other theropods of a similar size. Paul also depicted Dilophosaurus bouncing on its tail while lashing out at an enemy, similar to a kangaroo.

In 2007, Milner and James I. Kirkland suggested that Dilophosaurus had features that indicates it may have eaten fish. They pointed out that the ends of the jaws were expanded to the sides, forming a "rosette" of interlocking teeth, similar to those of spinosaurids, which are known to have eaten fish, and the gharials, which is the modern crocodile that eats most fish. The nasal openings were also retracted back on the jaws, similar to spinosaurids which have even more retracted nasal openings, and this may have limited water splashing into the nostrils during fishing. Both groups also had long arms with well-developed claws, which could help when catching fish. Lake Dixie, a large lake which extended from Utah to Arizona and Nevada, would have provided abundant fish in the "post-cataclysmic", biologically more impoverished world that followed the Triassic–Jurassic extinction event.

Crest function

Head of a model nicknamed "Dyzio", Geological Museum of the State Geological Institute, Warsaw

Welles conceded that suggestions as to the function of the crests were conjectural, but that though they had no grooves to indicate vascularization, they could have been used for thermoregulation. He also suggested they could have been used for species recognition or ornamentation. The Czerkas pointed out that the crests could not have been used during battle, as they would have been easily damaged due to their delicate structure. They suggested them to be a visual display for attracting a mate, or thermoregulation as well.

In 2011, American palaeontologists Kevin Padian and John R. Horner proposed that "bizarre structures" in dinosaurs in general (including crests, frills, horns, and domes) were primarily used for species recognition, and dismissed other explanations as unsupported by evidence. They noted that foo few specimens of cranially ornamented theropods, including Dilophosaurus, were known to test their evolutionary function statistically, and whether they represented sexual dimorphism or sexual maturity.

In a response to Padian and Horner the same year, Rob J. Knell and Scott D. Sampson argued that species recognition was not unlikely as a secondary function for "bizarre structures" in dinosaurs, but that sexual selection (used in display or combat to compete for mates) was a more likely explanation, due to the high cost of developing them, and because such structures appear to be highly variable within species. In 2013, the British palaeontologists David E. Hone and Darren Naish criticized the "species recognition hypothesis", and argued that no extant animals use such structures primarily for species recognition, and that Padian and Horner had ignored the possibility of mutual sexual selection (where both sexes are ornamented).

Development

Restoration showing adult size

Welles originally interpreted the smaller Dilophosaurus, specimens as juveniles, and the larger specimen as an adult. Paul suggested that the differences between the specimens was perhaps due to sexual dimorphism, as was seemingly also apparent in Copelophysis, which has "robust" and "gracile" forms of the same size, that might otherwise have been regarded as separate species. The smaller Dilophosaurus specimen would represent a "gracile" example following this scheme.

In 2005, Tykoski found that most Dilophosaurus specimens known were juvenile individuals, with only the largest specimen (UCMP 77270) being an adult, based on the level of coossification (fusion during bone tissue formation) of the bones. In 2005, Gay found no evidence of the sexual dimorphism suggested by Paul (but supposedly present in Copelophysis), and attributed the variation seen between Dilophosaurus specimens to individual variation and ontogeny (changes during growth). Though there was no dimorphism in the skeletons, he did not rule that there could have been in the crests, though more data was needed to determine this. Based on the tiny nasal crests on a juvenile specimen Yates tentatively assigned to the related genus Dracovenator, he suggested that these would have grown larger in adults.

In 1996, the American paleontologist Joe S. Tkach reported a histological study (microscopical study of internal features) of Dilophosaurus, by taking thin-sections of long bones and ribs of specimen UCMP 37303 (the lesser preserved of the original skeletons). The bone tissue were well-vascularized and had a fibro-lamella structure similar to that found in other theropods and the sauropodomorph Massospondylus. The plexiform (woven) structure of the bones suggested rapid growth, and Dilophosaurus may have attained a growth rate of 30 to 35 kilograms (66 to 77 lb) per year early in life.

Welles found that the replacement teeth of Dilophosaurus and other theropods originated deep inside the bone, decreasing in size the farther they were from the alveolar border. There were usually two or three replacement teeth in the alveoli, with the youngest being a small, hollow crown. The replacement teeth erupted on the outer side of the old teeth. When a tooth neared the gum line, the inner wall between the interdental plates was resorbed and formed a nutrient notch. As the new tooth erupted, it moved outwards to center itself in the alveolus, and the nutrient notch closed over.

Paleopathology

Paleopathologies in bones of the holotype, plotted onto a life restoration

Welles noted various paleopathologies (ancient signs of disease, such as injuries and malformations) in Dilophosaurus. The holotype had a sulcus (groove or furrow) on the neural arch of a cervical vertebra that may have been due to an injury or crushing, and two pits on the right humerus may have been abscesses (collections of pus) or artifacts. Welles also noted that it had a smaller and more delicate left humerus than the right, but with the reverse condition in its foreams. In 2001 the Australian paleontologist Ralph E. Molnar suggested that this was due to a developmental anomaly called fluctuating asymmetry. This anomaly can be caused by stress in animal populations, for example due to disturbances in their environment, and may indicate more intense selective pressure. Assymmetry can also result from traumatic events in early development of an animal, which would be more randomly distributed in time. A 2001 study conducted by the American paleontologist Bruce Rothschild and colleagues examined 60 foot bones assigned to Dilophosaurus for signs of stress fracture (which are caused by strenuous, repetitive actions), but none were found. Such injuries can be the result of very active, predatory lifestyles.

Restoration of the right hand of the holotype in flexion, with the deformed third finger (below) unable to flex

In 2016, Senter and Sara L. Juengst examined the paleopathologies of the holotype specimen, and found that it bore the greatest and most varied number of such maladies on the pectoral girdle and forelimb of any theropod dinosaur so far described, some of which are not known from any other dinosaur. It had eight afflicted bones, whereas no other theropod specimen is known with more than four; only six other theropods are known with more than one paleopathology on the pectoral girdle and forelimbs. On its left side it had a fractured scapula and radius, and fibriscesses (like abscesses) in the ulna and the outer phalanx bone of the thumb. On the right side it had torsion of its humeral shaft, three bony tumors on its radius, a truncated articular surface of its third metacarpal bone, and deformities on the first phalanx bone of the third finger. This finger was permanently deformed and unable to flex. The deformities of the humerus and the third finger may have been due to osteodysplasia, which had not been reported from dinosaurs before, but is known in birds. Affecting malnutritioned juvenile birds, this disease can cause pain in one limb, which makes the birds prefer to use the other limb instead, which therefore develops torsion.

Though the number of traumatic events that lead to these features is not certain, it is possible that they were all caused by a single encounter, for example by crashing into a tree or rock during a fight with another animal, which may have caused puncture wounds with its claws. Since all the injuries had healed, it is certain that the Dilophosaurus survived for a long time after these events; for months, perhaps years. The use of the forelimbs for prey-capture must have been compromised during the healing process. The dinosaur may therefore have endured a long period of fasting or subsisted on prey that was small enough for it to dispatch it with the mouth and feet, or with one forelimb. The high degree of pain it must have experienced in multiple locations for long durations also shows that it was a hardy animal. Senter and Juengst noted that paleopathologies in dinosaurs are under reported, and that even though Welles had thoroughly described the holotype, he had mentioned only one of the pathologies found by them. They suggested that such omissions may sometimes be omitted because descriptions of species are concerned with their characteristics rather than abnormalities, or because such features are difficult to recognize. Senter and Juengst therefore proposed that already described specimens should be reexamined for paleopathologies so that their life histories can be reconstructed.

Paleoichnology

Resting trace attributed to Dilophosaurus, St. George Dinosaur Discovery Site

The Dilophosauripus footprints described by Welles were all on the same level, and were described as a "chicken yard hodge-podge" of footprints, with few forming a trackway. The footprints had been imprinted in mud, which allowed the feet to sink down 5–10 centimeters (2.0–3.9 in). The prints were sloppy, and the varying breadth of the toe prints indicates mud had clung to the feet. The impressions varied to differences in the substrate and the manner they were made; sometimes the foot was planted directly, but there was often a backwards or forwards slip as the foot came down. The positions and angles of the toes also varied considerably, which indicates they must have been quite flexible. The Dilophosauripus footprints had an offset second toe with a thick base, and very long, straight claws which were were in line with the axes of the toe-pads. One of the footprints was missing the claw of the second toe, perhaps due to injury.

Fossilized footprints, discovered in 200-million-year-old sedimentary rock, which were assigned to Dilophosaurus were discovered in the Höganäs Formation in Vallåkra, Sweden, during the 1970s. The footprints appears to show that these dinosaurs lived in herds. Fossilized footprints assigned to Dilophosaurus have also been discovered in Sala, Sweden. Other tracks discovered in the Höganäs Formation have been assigned to the ichnogenus Grallator (Eubrontes) cf. giganteus, which were discovered in Rhaetian strata, and Grallator (Eubrontes) soltykovensis, which were discovered in Hettangian strata. A few of the tracks were taken to museums, but most of them disappeared in natural floodings. In 1994, Gierlinski and Ahlberg assigned these tracks from the Hoganas Formation of Sweden to Dilophosaurus as well.

Gierlinski (1996) observed unusual traces associated with a track specimen in the collection at the Pratt Museum in Amherst, Massachusetts. Specimen AC 1/7 is a "dinosaur sitting imprint", made when a dinosaur is resting its body on the ground, leaving an impression of its belly between a pair of footprints. Traces associated with AC 1/7 were interpreted by Gierlinski as the imprints of feathers, suggesting that Dilophosaurus was a feathered dinosaur. Further analysis proved, however, that the lines that seemed to be feathers were in reality just cracks in the mud where the animal sat. While this does not rule out the possibility of feathery covering on this species, there is no evidence for it and it currently remains as speculation.

Paleoecology

Restoration of Dilophosaurus chasing Scutellosaurus

Dilophosaurus is known from the Kayenta Formation, which dates to the Sinemurian and Pliensbachian stages of the Early Jurassic, approximately 196-183 million years ago. The Kayenta Formation is part of the Glen Canyon Group that includes formations not only in northern Arizona but also parts of southeastern Utah, western Colorado, and northwestern New Mexico. It is composed mostly of two facies, one dominated by silty deposition and the other dominated by sandstone. The siltstone facies is found in much of Arizona, while the sandstone facies is present in areas of northern Arizona, southern Utah, western Colorado, and northwestern New Mexico. The formation was primarily deposited by rivers, with the silty facies as the slower, more sluggish part of the river system. Kayenta Formation deposition was ended by the encroaching dune field that would become the Navajo Sandstone. A definitive radiometric dating of this formation has not yet been made, and the available stratigraphic correlation has been based on a combination of radiometric dates from vertebrate fossils, magnetostratigraphy and pollen evidence.

The Kayenta Formation has yielded a small but growing assemblage of organisms. Most fossils are from the silty facies. Most organisms known so far are vertebrates. Non-vertebrates include microbial or "algal" limestone, petrified wood, plant impressions, freshwater bivalves and snails, ostracods, and invertebrate trace fossils. Vertebrates are known from both body fossils and trace fossils. Vertebrates known from body fossils include hybodont sharks, indeterminate bony fish, lungfish, salamanders, the frog Prosalirus, the caecilian Eocaecilia, the turtle Kayentachelys, a sphenodontian reptile, lizards, several early crocodylomorphs including Calsoyasuchus, Eopneumatosuchus, Kayentasuchus, and Protosuchus, and the pterosaur Rhamphinion. Apart from Dilophosaurus, several dinosaurs are known, including the theropods Megapnosaurus (a species which is in need of a new genus name, and may also have had crests), Kayentavenator, and the "Shake N Bake" theropod, the sauropodomorph Sarahsaurus, a heterodontosaurid, and the armored dinosaurs Scelidosaurus and Scutellosaurus. Synapsids include the tritylodontids Dinnebiton, Kayentatherium, and Oligokyphus, morganucodontids, the possible early true mammal Dinnetherium, and a haramyid mammal. The majority of these finds come from the vicinity of Gold Spring, Arizona. Vertebrate trace fossils include coprolites and the tracks of therapsids, lizard-like animals, and several types of dinosaur.

Taphonomy

Welles outlined the taphonomy of the original specimens, changes that happened during their decay and fossilization. The holotype skeleton was found lying on its right side, and its head and neck were recurved backwards in the "death pose" dinosaur skeletons are often found in. This pose was thought to be opisthotonus (due to death-spasms) at the time, but may instead have been the result of how a carcass was embedded in sediments. Though the back was straight, the hindmost dorsal vertebrae were turned on their left sides. The caudal vertebrae extended irregularly from the pelvis, and the legs were articulated, with little displacement. Welles concluded that the specimens were buried at the place of their deaths, without having been transported much, but that the holotype specimen appears to have been disturbed by scavengers, indicated by the rotated dorsal vertebrae and crushed skull.

Gay noted that the specimens he described in 2001 showed evidence of having been transported by a stream. As none of the specimens were complete, they may have been transported over some distance, or have lied on the surface and weathered for some time before transport. They may have been transported by a flood, as indicated by the variety of animals found as fragments, and by bone breakage.

Cultural significance

The first three dimensional, standing skeleton of Dilophosaurus, Museum of Northern Arizona (the hand pose is outdated)

Dilophosaurus was featured in the 1990 novel Jurassic Park by the American writer Michael Crichton and its 1993 movie adaptation by the American director Steven Spielberg. The fictionalized Dilophosaurus of Jurassic Park was acknowledged as the "only serious departure from scientific veracity" in the movie's making-of book. Crichton had invented the dinosaur's ability to spit venom for the novel, and the movie's art department added an additional feature, a cowl folded against its neck that expanded and vibrated as the animal prepared to attack. The latter feature is similar to that of the frill-necked lizard of Australia. In addition, to avoid confusion with Velociraptor as featured in the movie, it was decided to make Dilophosaurus only 1.2 meters (4 ft) tall, instead of its assumed true height of about 3.0 meters (10 ft). Nicknamed "the spitter", the Dilophosaurus of the movie was realized though puppeteering, and required a full body with three interchangeable heads to produce the actions required in the script. Separate legs were also constructed for a shot where the dinosaur hops by. Unlike most of the other dinosaurs in the movie, no computer-generated imagery was employed when showing the Dilophosaurus.

The American geologist J. Bret Bennington noted that though Dilophosaurus probably did not have a frill or could spit venom, its bite could have been venomous like that of the Komodo dragon. He found that adding venom to the dinosaur was no less allowable than giving its skin color, which is also unknown. If the dinosaur had a frill, there would have been evidence for this in the bones, in the shape of a rigid structure to hold up the frill, or markings for where the muscles used to move it were attached. He also added that if it did have a frill, it would not have used it to intimidate its meal, but rather a competitor. Welles himself was "thrilled" to see Dilophosaurus in Jurassic Park, and though he noted the inaccuracies, he found them minor points and enjoyed the movie, and was happy to find the dinosaur "an internationally known actor".

In 2017, Dilophosaurus was designated as the state dinosaur of the US state of Connecticut, to become official with the new state budget in 2019. Dilophosaurus was chosen because tracks thought to be made by a related dinosaur were discovered in Rocky Hill in 1966, during excavation for the Interstate Highway 91. The six tracks were assigned to the ichnospecies Eubrontes giganteus, which was therefore made the state fossil of Connecticut in 1991. The area they were found in had been a Triassic lake, and when the significance of the area was confirmed, the highway was rerouted, and the are was made a Connecticut state park; Dinosaur State Park. In 1981, a sculpture of Dilophosaurus, the first life-sized reconstruction of this dinosaur, was donated to the state park.

External links

• Dilophosaurus! A Narrated Exhibition - "guided tour" narrated by Samuel Welles