Theropoda

Specimen of the troodontid Jinfengopteryx elegans, with seeds preserved in the stomach region

Theropods exhibit a wide range of diets, from insectivores to herbivores and carnivores. Strict carnivory has always been considered the ancestral diet for theropods as a group, and a wider variety of diets was historically considered a characteristic exclusive to the avian theropods (birds). However, discoveries in the late 20th and early 21st centuries showed that a variety of diets existed even in more basal lineages.[4] All early finds of theropod fossils showed them to be primarily carnivorous. Fossilized specimens of early theropods known to scientists in the 19th and early 20th centuries all possessed sharp teeth with serrated edges for cutting flesh, and some specimens even showed direct evidence of predatory behavior. For example, a Compsognathus longipes fossil was found with a lizard in its stomach, and a Velociraptor mongoliensis specimen was found locked in combat with a Protoceratops andrewsi (a type of ornithischian dinosaur).

The first confirmed non-carnivorous fossil theropods found were the therizinosaurs, originally known as "segnosaurs". First thought to be prosauropods, these enigmatic dinosaurs were later proven to be highly specialized, herbivorous theropods. Therizinosaurs possessed large abdomens for processing plant food, and small heads with beaks and leaf-shaped teeth. Further study of maniraptoran theropods and their relationships showed that therizinosaurs were not the only early members of this group to abandon carnivory. Several other lineages of early maniraptorans show adaptations for an omnivorous diet, including seed-eating (some troodontids) and insect-eating (many avialans and alvarezsaurs). Oviraptorosaurs, ornithomimosaurs and advanced troodontids were likely omnivorous as well, and some early theropods (such as Masiakasaurus knopfleri and the spinosaurids) appear to have specialized in catching fish.[5][6]

Diet is largely deduced by the tooth morphology,[7] tooth marks on bones of the prey, and gut contents. Some theropods, such as Baryonyx, Lourinhanosaurus, ornithomimosaurs, and birds, are known to use gastroliths, or gizzard-stones.

The majority of theropod teeth are blade-like, with serration on the edges,[8] called ziphodont. Others are pachydont or phyllodont depending on the shape of the tooth or denticles. The morphology of the teeth is distinct enough to tell the major families apart,[7] which indicate different diet strategies. An investigation in July 2015 discovered that what appeared to be "cracks" in their teeth were actually folds that helped to prevent tooth breakage by strengthening individual serrations as they attacked their prey.[9] The folds helped the teeth stay in place longer, especially as theropods evolved into larger sizes and had more force in their bite.[10][11]

Fossil of an Anchiornis, showing large preserved feather imprints

Mesozoic theropods were also very diverse in terms of skin texture and covering. Feathers or feather-like structures (filaments) are attested in most lineages of theropods (see feathered dinosaur). However, outside the coelurosaurs, feathers may have been confined to the young, smaller species, or limited parts of the animal. Many larger theropods had skin covered in small, bumpy scales. In some species, these were interspersed with larger scales with bony cores, or osteoderms. This type of skin is best known in the ceratosaur Carnotaurus, which has been preserved with extensive skin impressions.[12]

The coelurosaur lineages most distant from birds had feathers that were relatively short and composed of simple, possibly branching filaments.[13] Simple filaments are also seen in therizinosaurs, which also possessed large, stiffened "quill"-like feathers. More fully feathered theropods, such as dromaeosaurids, usually retain scales only on the feet. Some species may have mixed feathers elsewhere on the body as well. Scansoriopteryx preserved scales near the underside of the tail,[14] and Juravenator may have been predominantly scaly with some simple filaments interspersed.[15] On the other hand, some theropods were completely covered with feathers, such as the troodontid Anchiornis, which even had feathers on the feet and toes.[16]

Tyrannosaurus was for many decades the largest known theropod and best known to the general public. Since its discovery, however, a number of other giant carnivorous dinosaurs have been described, including Spinosaurus, Carcharodontosaurus, and Giganotosaurus.[17] The original Spinosaurus specimens (as well as newer fossils described in 2006) support the idea that Spinosaurus is longer than Tyrannosaurus, showing that Spinosaurus was possibly 3 meters longer than Tyrannosaurus though Tyrannosaurus could still be more massive than Spinosaurus.[18] Specimens such as Sue and Scotty are both estimated to be the heaviest theropods known to science. There is still no clear explanation for why these animals grew so heavy and bulky compared to the land predators that came before and after them.

The largest extant theropod is the common ostrich, up to 2.74 m (9 ft) tall and weighing between 90 and 130 kg (200 - 290 lb).[19]

Size comparison of selected giant theropod dinosaurs – the largest (left) is Spinosaurus aegyptiacus, smallest (right) is Carcharodontosaurus saharicus.

The smallest non-avialan theropod known from adult specimens is the troodontid Anchiornis huxleyi, at 110 grams in weight and 34 centimeters (1 ft) in length.[16] When modern birds are included, the bee hummingbird Mellisuga helenae is smallest at 1.9 g and 5.5 cm (2.2 in) long.[20]

Recent theories propose that theropod body size shrank continuously over a period of 50 million years, from an average of 163 kilograms (359 lb) down to 0.8 kilograms (1.8 lb), eventually evolving into modern birds. This was based on evidence that theropods were the only dinosaurs to get continuously smaller, and that their skeletons changed four times as fast as those of other dinosaur species.[21][22]

In order to estimate the growth rates of theropods, scientists need to calculate both age and body mass of a dinosaur. Both of these measures can only be calculated through fossilized bone and tissue, so regression analysis and extant animal growth rates as proxies are used to make predictions. Fossilized bones exhibit growth rings that appear as a result of growth or seasonal changes, which can be used to approximate age at the time of death.[23] However, the amount of rings in a skeleton can vary from bone to bone, and old rings can also be lost at advanced age, so scientists need to properly control these two possibly confounding variables.

Body mass is harder to determine as bone mass only represents a small proportion of the total body mass of animals. One method is to measure the circumference of the femur, which in non-avian theropod dinosaurs has been shown to be a relatively proportional to quadrupedal mammals,[24] and use this measurement as a function of body weight, as the proportions of long bones like the femur grow proportionately with body mass.[24] The method of using extant animal bone proportion to body mass ratios to make predictions about extinct animals is known as the extant-scaling (ES) approach.[25] A second method, known as the volumetric-density (VD) approach, uses full-scale models of skeletons to make inferences about potential mass.[25] The ES approach is better for wide range studies including many specimens and doesn't require as much of a complete skeleton as the VD approach, but the VD approach allows scientists to better answer more physiological questions about the animal, such as locomotion and center of gravity.[25]

The current consensus is that non-avian theropods didn't exhibit a group wide growth rate, but instead had varied rates depending on their size. However, all non-avian theropods had faster growth rates than extant reptiles, even when modern reptiles are scaled up to the large size of some non-avian theropods. As body mass increases, the relative growth rate also increases. This trend may be due to the need to reach the size required for reproductive maturity.[26] For example, one of the smallest known theropods was Microraptor zhaoianus, which had a body mass of 200 grams, grew at a rate of approximately .33 grams per day.[27] A comparable reptile of the same size grows at half of this rate.[27] The growth rates of medium-sized non-avian theropods (100–1000 kg) approximated those of precocial birds, which are much slower than altricial birds. Large theropods (1500–3500 kg) grew even faster, similar to rates displayed by eutherian mammals.[27] The largest non-avian theropods, like Tyrannosaurus rex had similar growth dynamics to the largest living land animal today, the African elephant, which is characterized by a rapid period of growth until maturity, subsequently followed by slowing growth in adulthood.[28]

An ostrich walking on a road in Etosha National Park, Namibia

As a hugely diverse group of animals, the posture adopted by theropods likely varied considerably between various lineages through time.[29] All known theropods are bipedal, with the forelimbs reduced in length and specialized for a wide variety of tasks (see below). In modern birds, the body is typically held in a somewhat upright position, with the upper leg (femur) held parallel to the spine and with the forward force of locomotion generated at the knee. Scientists are not certain how far back in the theropod family tree this type of posture and locomotion extends.[29]

Non-avian theropods were first recognized as bipedal during the 19th century, before their relationship to birds was widely accepted. During this period, theropods such as carnosaurs and tyrannosaurids were thought to have walked with vertical femurs and spines in an upright, nearly erect posture, using their long, muscular tails as additional support in a kangaroo-like tripodal stance.[29] Beginning in the 1970s, biomechanical studies of extinct giant theropods cast doubt on this interpretation. Studies of limb bone articulation and the relative absence of trackway evidence for tail dragging suggested that, when walking, the giant, long-tailed theropods would have adopted a more horizontal posture with the tail held parallel to the ground.[29][30] However, the orientation of the legs in these species while walking remains controversial. Some studies support a traditional vertically oriented femur, at least in the largest long-tailed theropods,[30] while others suggest that the knee was normally strongly flexed in all theropods while walking, even giants like the tyrannosaurids.[31][32] It is likely that a wide range of body postures, stances, and gaits existed in the many extinct theropod groups.[29][33]

Although rare, complete casts of theropod endocrania are known from fossils. Theropod endocrania can also be reconstructed from preserved brain cases without damaging valuable specimens by using a computed tomography scan and 3D reconstruction software. These finds are of evolutionary significance because they help document the emergence of the neurology of modern birds from that of earlier reptiles. An increase in the proportion of the brain occupied by the cerebrum seems to have occurred with the advent of the Coelurosauria and "continued throughout the evolution of maniraptorans and early birds."[34]

Studies show that theropods had very sensitive snouts. It is suggested they might have been used for temperature detection, feeding behavior, and wave detection.[35][36]

Mummified enantiornithean wing (of an unknown genus) from Cenomanian amber from Myanmar

Shortened forelimbs in relation to hind legs was a common trait among theropods, most notably in the abelisaurids (such as Carnotaurus) and the tyrannosaurids (such as Tyrannosaurus). This trait was, however, not universal: spinosaurids had well developed forelimbs, as did many coelurosaurs. The relatively robust forelimbs of one genus, Xuanhanosaurus, led Dong Zhiming to suggest that the animal might have been quadrupedal.[37] However, this is no longer thought to be likely.[38]

The hands are also very different among the different groups. The most common form among non-avian theropods is an appendage consisting of three fingers; the digits I, II and III (or possibly II, III and IV), with sharp claws. Some basal theropods (e.g. Herrerasaurus, Eoraptor) had four digits, and also a reduced metacarpal V. Ceratosaurians usually had four digits, while most tetanurans had three.[39]

The forelimbs' scope of use is also believed to have also been different among different families. The spinosaurids could have used their powerful forelimbs to hold fish. Some small maniraptorans such as scansoriopterygids are believed to have used their forelimbs to climb in trees.[14] The wings of modern birds are used primarily for flight, though they are adapted for other purposes in certain groups. For example, aquatic birds such as penguins use their wings as flippers.

Diagram of Deinonychus (left) and Archaeopteryx (right) forelimbs illustrating wing-like posture

Contrary to the way theropods have often been reconstructed in art and the popular media, the range of motion of theropod forelimbs was severely limited, especially compared with the forelimb dexterity of humans and other primates.[40] Most notably, theropods and other bipedal saurischian dinosaurs (including the bipedal prosauropods) could not pronate their hands—that is, they could not rotate the forearm so that the palms faced the ground or backwards towards the legs. In humans, pronation is achieved by motion of the radius relative to the ulna (the two bones of the forearm). In saurischian dinosaurs, however, the end of the radius near the elbow was actually locked into a groove of the ulna, preventing any movement. Movement at the wrist was also limited in many species, forcing the entire forearm and hand to move as a single unit with little flexibility.[41] In theropods and prosauropods, the only way for the palm to face the ground would have been by lateral splaying of the entire forelimb, as in a bird raising its wing.[40]

In carnosaurs like Acrocanthosaurus, the hand itself retained a relatively high degree of flexibility, with mobile fingers. This was also true of more basal theropods, such as herrerasaurs. Coelurosaurs showed a shift in the use of the forearm, with greater flexibility at the shoulder allowing the arm to be raised towards the horizontal plane, and to even greater degrees in flying birds. However, in coelurosaurs, such as ornithomimosaurs and especially dromaeosaurids, the hand itself had lost most flexibility, with highly inflexible fingers. Dromaeosaurids and other maniraptorans also showed increased mobility at the wrist not seen in other theropods, thanks to the presence of a specialized half-moon shaped wrist bone (the semi-lunate carpal) that allowed the whole hand to fold backward towards the forearm in the manner of modern birds.[41]

In 2001, Ralph E. Molnar published a survey of pathologies in theropod dinosaur bone. He found pathological features in 21 genera from 10 families. Pathologies were found in theropods of all body size although they were less common in fossils of small theropods, although this may be an artifact of preservation. They are very widely represented throughout the different parts of theropod anatomy. The most common sites of preserved injury and disease in theropod dinosaurs are the ribs and tail vertebrae. Despite being abundant in ribs and vertebrae, injuries seem to be "absent... or very rare" on the bodies' primary weight supporting bones like the sacrum, femur, and tibia. The lack of preserved injuries in these bones suggests that they were selected by evolution for resistance to breakage. The least common sites of preserved injury are the cranium and forelimb, with injuries occurring in about equal frequency at each site. Most pathologies preserved in theropod fossils are the remains of injuries like fractures, pits, and punctures, often likely originating with bites. Some theropod paleopathologies seem to be evidence of infections, which tended to be confined only to small regions of the animal's body. Evidence for congenital malformities have also been found in theropod remains. Such discoveries can provide information useful for understanding the evolutionary history of the processes of biological development. Unusual fusions in cranial elements or asymmetries in the same are probably evidence that one is examining the fossils of an extremely old individual rather than a diseased one.[42]

The trackway of a swimming theropod, the first in China of the ichnogenus named Characichnos, was discovered at the Feitianshan Formation in Sichuan.[43] These new swim tracks support the hypothesis that theropods were adapted to swimming and capable of traversing moderately deep water. Dinosaur swim tracks are considered to be rare trace fossils, and are among a class of vertebrate swim tracks that also include those of pterosaurs and crocodylomorphs. The study described and analyzed four complete natural molds of theropod foot prints that are now stored at the Huaxia Dinosaur Tracks Research and Development Center (HDT). These dinosaur footprints were in fact claw marks, which suggest that this theropod was swimming near the surface of a river and just the tips of its toes and claws could touch the bottom. The tracks indicate a coordinated, left-right, left-right progression, which supports the proposition that theropods were well-coordinated swimmers.[43]

Othniel Charles Marsh, who coined the name Theropoda. Photo c. 1870

O. C. Marsh coined the name Theropoda (meaning "beast feet") in 1881.[53] Marsh initially named Theropoda as a suborder to include the family Allosauridae, but later expanded its scope, re-ranking it as an order to include a wide array of "carnivorous" dinosaur families, including Megalosauridae, Compsognathidae, Ornithomimidae, Plateosauridae and Anchisauridae (now known to be herbivorous sauropodomorphs) and Hallopodidae (subsequently revealed as relatives of crocodilians). Due to the scope of Marsh's Order Theropoda, it came to replace a previous taxonomic group that Marsh's rival E. D. Cope had created in 1866 for the carnivorous dinosaurs: Goniopoda ("angled feet").[38]

By the early 20th century, some palaeontologists, such as Friedrich von Huene, no longer considered carnivorous dinosaurs to have formed a natural group. Huene abandoned the name "Theropoda", instead using Harry Seeley's Order Saurischia, which Huene divided into the suborders Coelurosauria and Pachypodosauria. Huene placed most of the small theropod groups into Coelurosauria, and the large theropods and prosauropods into Pachypodosauria, which he considered ancestral to the Sauropoda (prosauropods were still thought of as carnivorous at that time, owing to the incorrect association of rauisuchian skulls and teeth with prosauropod bodies, in animals such as Teratosaurus).[38] Describing the first known dromaeosaurid (Dromaeosaurus albertensis) in 1922,[54] W. D. Matthew and Barnum Brown became the first paleontologists to exclude prosauropods from the carnivorous dinosaurs, and attempted to revive the name "Goniopoda" for that group, but other scientists did not accept either of these suggestions.[38]

Allosaurus was one of the first dinosaurs classified as a theropod.

In 1956, "Theropoda" came back into use—as a taxon containing the carnivorous dinosaurs and their descendants—when Alfred Romer re-classified the Order Saurischia into two suborders, Theropoda and Sauropoda. This basic division has survived into modern palaeontology, with the exception of, again, the Prosauropoda, which Romer included as an infraorder of theropods. Romer also maintained a division between Coelurosauria and Carnosauria (which he also ranked as infraorders). This dichotomy was upset by the discovery of Deinonychus and Deinocheirus in 1969, neither of which could be classified easily as "carnosaurs" or "coelurosaurs". In light of these and other discoveries, by the late 1970s Rinchen Barsbold had created a new series of theropod infraorders: Coelurosauria, Deinonychosauria, Oviraptorosauria, Carnosauria, Ornithomimosauria, and Deinocheirosauria.[38]

With the advent of cladistics and phylogenetic nomenclature in the 1980s, and their development in the 1990s and 2000s, a clearer picture of theropod relationships began to emerge. Jacques Gauthier named several major theropod groups in 1986, including the clade Tetanurae for one branch of a basic theropod split with another group, the Ceratosauria. As more information about the link between dinosaurs and birds came to light, the more bird-like theropods were grouped in the clade Maniraptora (also named by Gauthier in 1986). These new developments also came with a recognition among most scientists that birds arose directly from maniraptoran theropods and, on the abandonment of ranks in cladistic classification, with the re-evaluation of birds as a subset of theropod dinosaurs that survived the Mesozoic extinctions and lived into the present.[38]

Ceratosaurus, a ceratosaurid

Irritator, a spinosaurid

Mapusaurus, a carcharodontosaurid

Microraptor, a dromaeosaurid

Passer domesticus, an avian, is the world's most widespread extant wild theropod.[55]

The following is a simplified classification of theropod groups based on their evolutionary relationships, and organized based on the list of Mesozoic dinosaur species provided by Holtz.[1] A more detailed version can be found at dinosaur classification. The dagger (†) is used to signify groups with no living members.

• †Coelophysoidea (small, early theropods; includes Coelophysis and close relatives)
• †Ceratosauria (generally elaborately horned, the dominant southern carnivores of the Cretaceous; includes Carnotaurus and close relatives, like Majungasaurus and Chenanisaurus)
• Tetanurae ("stiff tails"; includes most theropods)

• †Megalosauroidea (early group of large carnivores including the semi-aquatic spinosaurids)
• †Allosauroidea (Allosaurus and close relatives, like Carcharodontosaurus)
• †Megaraptora (A group of medium to large Orionides with unknown affinities, quite common in the southern hemisphere)
• Coelurosauria (feathered theropods, with a range of body sizes and niches)

• †Compsognathidae (early coelurosaurs with reduced forelimbs)
• †Tyrannosauridae (Tyrannosaurus and close relatives; had reduced forelimbs)
• †Ornithomimosauria ("ostrich-mimics"; mostly toothless; carnivores to possible herbivores)
• Maniraptora ("hand snatchers"; had long, slender arms and fingers)

• †Alvarezsauroidea (small insectivores with reduced forelimbs each bearing one enlarged claw)
• †Therizinosauria (bipedal herbivores with large hand claws and small heads)
• †Scansoriopterygidae (small, arboreal maniraptors with long third fingers)
• †Oviraptorosauria (mostly toothless; their diet and lifestyle are uncertain)
• †Archaeopterygidae (small, winged protobirds)
• †Dromaeosauridae (small to medium-sized theropods)
• †Troodontidae (small, gracile theropods)
• Avialae (birds and extinct relatives)

• †Omnivoropterygidae (large, early short-tailed avialans)
• †Confuciusornithidae (small toothless birds)
• †Enantiornithes (primitive tree-dwelling, flying birds)
• Euornithes (advanced flying birds)

• †Yanornithiformes (toothed Cretaceous Chinese birds)
• †Hesperornithes (specialized aquatic diving birds)
• Aves (modern, beaked birds and their extinct relatives)

The following family tree illustrates a synthesis of the relationships of the major theropod groups based on various studies conducted in the 2010s.[56]

A large study of early dinosaurs by Dr Matthew G. Baron, David Norman and Paul M. Barrett (2017) published in the journal Nature suggested that Theropoda is actually more closely related to Ornithischia, to which it formed the sister group within the clade Ornithoscelida. This new hypothesis also recovered Herrerasauridae as the sister group to Sauropodomorpha in the redefined Saurischia and suggested that the hypercarnivore morphologies that are observed in specimens of theropods and herrerasaurids were acquired convergently.[57][58] However, this phylogeny remains controversial and additional work is being done to clarify these relationships.