Stegosaurus

S. stenops skull cast, Natural History Museum of Utah

The long and narrow skull was small in proportion to the body. It had a small antorbital fenestra, the hole between the nose and eye common to most archosaurs, including modern birds, though lost in extant crocodylians. The skull's low position suggests that Stegosaurus may have been a browser of low-growing vegetation. This interpretation is supported by the absence of front teeth and their likely replacement by a horny beak or rhamphotheca. The lower jaw had flat downward and upward extensions that would have completely hidden the teeth when viewed from the side, and these probably supported a turtle-like beak in life. The presence of a beak extended along much of the jaws may have precluded the presence of cheeks in these species.[6][7] Such an extensive beak was probably unique to Stegosaurus and some other advanced stegosaurids among ornithischians, which usually had beaks restricted to the jaw tips.[6][8] Other researchers have interpreted these ridges as modified versions of similar structures in other ornithischians which might have supported fleshy cheeks, rather than beaks.[9] Stegosaurian teeth were small, triangular, and flat; wear facets show that they did grind their food.[10]

Despite the animal's overall size, the braincase of Stegosaurus was small, being no larger than that of a dog. A well-preserved Stegosaurus braincase allowed Othniel Charles Marsh to obtain, in the 1880s, a cast of the brain cavity or endocast of the animal, which gave an indication of the brain size. The endocast showed the brain was indeed very small, the smallest proportionally of all dinosaur endocasts then known. The fact that an animal weighing over 4.5 metric tons (5 short tons) could have a brain of no more than 80 g (2.8 oz) contributed to the popular old idea that all dinosaurs were unintelligent, an idea now largely rejected.[11] Actual brain anatomy in Stegosaurus is poorly known, but the brain itself was small even for a dinosaur, fitting well with a slow, herbivorous lifestyle and limited behavioral complexity.[12]

Mounted composite skeleton referred to S. ungulatus, Carnegie Museum of Natural History

In Stegosaurus stenops there are 27 bones in the vertebral column anterior to the sacrum, a varying number of vertebrae in the sacrum, with four in most subadults, and around 46 caudal (tail) vertebrae. The presacrals are divided into cervical (neck) and dorsal (back) vertebrae, with around 10 cervicals and 17 dorsals, the total number being one greater than in Hesperosaurus, two greater than Huayangosaurus, although Miragaia preserves 17 cervicals and an unknown number of dorsals. The first cervical vertebra is the axis bone, which is connected and often fused to the atlas bone. Farther posteriorly, the proportionately larger the cervicals become, although they do not change greatly in anything other than size. Past the first few dorsals, the centrum of the bones become more elongate front-to-back, and the transverse processes become more elevated dorsal. The sacrum of S. stenops includes four sacral vertebrae, but one of the dorsals is also incorporated into the structure. In some specimens of S. stenops, a caudal is also incorporated, as a caudosacral. In Hesperosaurus there are two dorsosacrals, and only four fused sacrals, but in Kentrosaurus there may be as many as seven vertebrae in the sacrum, with both dorsosacrals and caudosacrals. S. stenops preserves 46 caudal vertebrae, and up to 49, and along the series both the centrums and the neural spines become smaller, until the neural spines disappear at caudal 35. Around the middle of the tail, the neural spines become bifurcated, meaning they are divided near the top.[13]

With multiple well-preserved skeletons, S. stenops preserves all regions of the body, including the limbs. The scapula (shoulder blade) is sub-rectangular, with a robust blade. Though it is not always perfectly preserved, the acromion ridge is slightly larger than in Kentrosaurus. The blade is relatively straight, although it curves towards the back. There is a small bump on the back of the blade, that would have served as the base of the triceps muscle. Articulated with the scapula, the coracoid is sub-circular.[13] The hind feet each had three short toes, while each fore foot had five toes; only the inner two toes had a blunt hoof. The phalangeal formula is 2-2-2-2-1, meaning the innermost finger of the fore limb has two bones, the next has two, etc.[14] All four limbs were supported by pads behind the toes.[15] The fore limbs were much shorter than the stocky hind limbs, which resulted in an unusual posture. The tail appears to have been held well clear of the ground, while the head of Stegosaurus was positioned relatively low down, probably no higher than 1 m (3.3 ft) above the ground.[16]

Life restoration of S. stenops based on the "Sophie" specimen

The most recognizable features of Stegosaurus are its dermal plates, which consisted of between 17 and 22 separate plates and flat spines.[17] These were highly modified osteoderms (bony-cored scales), similar to those seen in crocodiles and many lizards today. They were not directly attached to the animal's skeleton, instead arising from the skin. The largest plates were found over the hips and could measure over 60 cm (24 in) wide and 60 cm (24 in) tall.[17]

In a 2010 review of Stegosaurus species, Peter Galton suggested that the arrangement of the plates on the back may have varied between species, and that the pattern of plates as viewed in profile may have been important for species recognition. Galton noted that the plates in S. stenops have been found articulated in two staggered rows, rather than paired. Fewer S. ungulatus plates have been found, and none articulated, making the arrangement in this species more difficult to determine. However, the type specimen of S. ungulatus preserves two flattened spine-like plates from the tail that are nearly identical in shape and size, but are mirror images of each other, suggesting that at least these were arranged in pairs.[18] Many of the plates are manifestly chiral[19][20] and no two plates of the same size and shape have been found for an individual; however plates have been correlated between individuals. Well preserved integumentary impressions of the plates of Hesperosaurus show a smooth surface with long and parallel, shallow grooves. This indicates that the plates were covered in keratinous sheaths.[21]

One of the major subjects of books and articles about Stegosaurus is the plate arrangement.[28] The argument has been a major one in the history of dinosaur reconstruction. Four possible plate arrangements have been proposed over the years:

1901 life restoration of S. ungulatus by Charles R. Knight with paired dorsal plates and eight tail spikes

• The plates lie flat along the back, as a shingle-like armor. This was Marsh's initial interpretation, which led to the name 'roof lizard'. As further and complete plates were found, their form showed they stood on edge, rather than lying flat.
• By 1891, Marsh published a more familiar view of Stegosaurus,[29] with a single row of plates. This was dropped fairly early on (apparently because it was poorly understood how the plates were embedded in the skin and they were thought to overlap too much in this arrangement). It was revived, in somewhat modified form, in the 1980s, by Stephen Czerkas,[30] based on the arrangement of iguana dorsal spines.
• The plates were paired in a double row along the back, such as in Knight's 1901 reconstruction and the 1933 film King Kong.
• Two rows of alternating plates. By the early 1960s, this had become (and remains) the prevalent idea, mainly because some S. stenops fossils in which the plates are still partially articulated show this arrangement. This arrangement is chiral and so demands that a specimen be distinguished from its distinct, hypothetical mirror-image form.[19][20]

Many of the species initially described have since been considered to be invalid or synonymous with earlier named species,[31] leaving two well-known and one poorly known species. Confirmed Stegosaurus remains have been found in the Morrison Formation's stratigraphic zones 2–6, with additional remains possibly referrable to Stegosaurus recovered from stratigraphic zone 1.[32]

• Stegosaurus ungulatus, meaning "hoofed roof lizard", was named by Marsh in 1879,[33] from remains recovered at Como Bluff, Wyoming (Quarry 12, near Robber's Roost).[31] It might be synonymous with S. stenops.[13] At 9 m (29.5 ft), it was the longest species within the genus Stegosaurus. A fragmentary Stegosaurus specimen discovered in Portugal and dating from the upper Kimmeridgian-lower Tithonian stage has been tentatively assigned to this species.[34] Stegosaurus ungulatus can be distinguished from S. stenops by the presence of longer hind limbs, proportionately smaller, more pointed plates with wide bases and narrow tips, and by several small, flat, spine-like plates just before the spikes on the tail. These spine-like plates appear to have been paired, due to the presence of at least one pair that are identical but mirrored. S. ungulatus also appears to have had longer legs (femora) and hip bones than other species. The type specimen of S. ungulatus was discovered with eight spikes, though they were scattered away from their original positions. These have often been interpreted as indicating that the animal had four pairs of tail spikes. No specimens have been found with complete or articulated sets of tail spikes, but no additional specimens have been found that preserve eight spikes together. It is possible the extra pair of spikes came from a different individual, and though no other extra bones were found with the specimen, these may be found if more digging were done at the original site.[18] Specimens from other quarries (such as a tail from Quarry 13, now forming part of the composite skeleton AMNH 650 at the American Museum of Natural History), referred to S. ungulatus on the basis of their notched tail vertebrae, are preserved with only four tail spikes.[31] The type specimen of S. ungulatus (YPM 1853) was incorporated into the first ever mounted skeleton of a stegosaur at the Peabody Museum of Natural History in 1910 by Richard Swann Lull. It was initially mounted with paired plates set wide, above the base of the ribs, but was remounted in 1924 with two staggered rows of plates along the midline of the back.[17] Additional specimens recovered from the same quarry by the United States National Museum of Natural History, including tail vertebrae and an additional large plate (USNM 7414), belong to the same individual as YPM 1853.[18]

Articulated holotype of S. stenops (USNM 4934), known as the "road kill" specimen

• Stegosaurus stenops, meaning "narrow-faced roof lizard", was named by Marsh in 1887,[35] with the holotype having been collected by Marshall Felch at Garden Park, north of Cañon City, Colorado, in 1886. This is the best-known species of Stegosaurus, mainly because its remains include at least one complete articulated skeleton. It had proportionately large, broad plates and rounded tail plates. Articulated specimens show that the plates were arranged alternating in a staggered double row. S. stenops is known from at least 50 partial skeletons of adults and juveniles, one complete skull, and four partial skulls. It was shorter than other species, at 7 m (23 ft). Found in the Morrison Formation, Colorado, Wyoming, and Utah.[9]
• Stegosaurus sulcatus, meaning "furrowed roof lizard", was described by Marsh in 1887 based on a partial skeleton.[35] It has traditionally been considered a synonym of S. armatus,[9] though more recent studies suggest it is not.[18] S. sulcatus is distinguished mainly by its unusually large, furrowed spikes with very large bases. A spike associated with the type specimen, originally thought to be a tail spike, may in fact come from the shoulder or hip, since its base is much larger than the corresponding tail vertebrae. A review published by Maidment and colleagues in 2008 regarded it as an indeterminate species possibly not even belonging to Stegosaurus at all, but to a different genus.[36][37] Peter Galton suggested it should be considered a valid species due to its unique spikes.[18]

Susannah Maidment and colleagues in 2008 proposed extensive alterations to the taxonomy of Stegosaurus. They advocated synonymizing S. stenops and S. ungulatus with S. armatus, and sinking Hesperosaurus and Wuerhosaurus into Stegosaurus, with their type species becoming Stegosaurus mjosi and Stegosaurus homheni, respectively. They regarded S. longispinus as dubious. Thus, their conception of Stegosaurus would include three valid species (S. armatus, S. homheni, and S. mjosi) and would range from the Late Jurassic of North America and Europe to the Early Cretaceous of Asia.[36] However, this classification scheme was not followed by other researchers, and a 2017 cladistic analysis co-authored by Maidment with Thomas Raven rejects the synonymy of Hesperosaurus with Stegosaurus.[18][38] In 2015, Maidment et al. revised their suggestion due to the recognition by Galton of S. armatus as a nomen dubium and its replacement by S. stenops as type species.[13]

• Stegosaurus armatus, meaning "armored roof lizard", was the first species to be found and the original type species named by O.C. Marsh in 1877.[22] It is known from a partial skeletons, and more than 30 fragmentary specimens have been referred to it.[9] However, the type specimen was very fragmentary, consisting only of a partial tail, hips, and leg, parts of some back vertebrae, and a single fragmentary plate (the presence of which was used to give the animal its name). No other plates or spikes were found, and the entire front half of the animal appears not to have been preserved.[31] Because the type specimen is very fragmentary, it is extremely difficult to compare it with other species based on better specimens, and it is now generally considered to be a nomen dubium. Because of this, it was replaced by S. stenops as the type species of Stegosaurus in a ruling of the ICZN in 2013.[39]
• Stegosaurus "affinis", named by Marsh in 1881, is only known from a pubis which has since been lost. Because Marsh did not provide an adequate description of the bone with which to distinguish a new species, this name is considered a nomen nudum.[23]
• Diracodon laticeps was described by Marsh in 1881, from some jawbone fragments.[40] Bakker resurrected D. laticeps in 1986 as a senior synonym of S. stenops,[41] although others note that the material is not diagnostic and is only referable to Stegosaurus sp., making it a nomen dubium.[31]
• Stegosaurus duplex, meaning "two plexus roof lizard" (in allusion to the greatly enlarged neural canal of the sacrum which Marsh characterized as a "posterior brain case"), was named by Marsh in 1887 (including the holotype specimen). The disarticulated bones were actually collected in 1879 by Edward Ashley at Como Bluff. Marsh initially distinguished it from S. ungulatus based on the fact that each sacral (hip) vertebra bore its own rib, which he claimed was unlike the anatomy of S. ungulatus; however, the sacrum of S. ungulatus had not actually been discovered. Marsh also suggested that S. duplex may have lacked armor, since no plates or spikes were found with the specimen, though a single spike may actually have been present nearby, and re-examination of the site maps has shown that the entire specimen was found highly disarticulated and scattered.[31] It is generally considered a synonym of S. ungulatus today, and parts of the specimen were actually incorporated into the Peabody Museum S. ungulatus skeletal mount in 1910.[18]

• Stegosaurus marshi, which was described by Lucas in 1901, was renamed Hoplitosaurus in 1902.[42]
• Stegosaurus priscus, described by Nopcsa in 1911, was reassigned to Lexovisaurus,[9] and is now the type species of Loricatosaurus.[36]
• Stegosaurus longispinus was named by Charles W. Gilmore.[23] It is now the type species of the genus Alcovasaurus.[43]
• Stegosaurus madagascariensis from Madagascar is known solely from teeth and was described by Piveteau in 1926. The teeth were variously attributed to a stegosaur, the theropod Majungasaurus,[44] a hadrosaur or even a crocodylian, but is now considered a possible ankylosaur.[36]

The origin of Stegosaurus is uncertain, as few remains of basal stegosaurs and their ancestors are known. Recently, stegosaurids have been shown to be present in the lower Morrison Formation, existing several million years before the occurrence of Stegosaurus itself, with the discovery of the related Hesperosaurus from the early Kimmeridgian.[46] The earliest stegosaurid (the genus Lexovisaurus) is known from the Oxford Clay Formation of England and France, giving it an age of early to middle Callovian.

The earlier and more basal genus Huayangosaurus from the Middle Jurassic of China (some 165 million years ago – Mya) antedates Stegosaurus by 20 million years and is the only genus in the family Huayangosauridae. Earlier still is Scelidosaurus, from Early Jurassic England, which lived about 190 Mya. It possessed features of both stegosaurs and ankylosaurs. Emausaurus from Germany was another small quadruped, while Scutellosaurus from Arizona was an even earlier genus and was facultatively bipedal. These small, lightly armored dinosaurs were closely related to the direct ancestor of both stegosaurs and ankylosaurs. A trackway of a possible early armored dinosaur, from around 195 Mya, has been found in France.[47]

Back plate cast, Museum of the Rockies

The function of Stegosaurus' plates has been much debated. Marsh suggested that they functioned as some form of armor,[48] though Davitashvili (1961) disputed this, claiming that they were too fragile and ill-placed for defensive purposes, leaving the animal's sides unprotected.[53] Nevertheless, others have continued to support a defensive function. Bakker suggested in 1986 that the plates were covered in horn comparing the surface of the fossilized plates to the bony cores of horns in other animals known or thought to bear horns.[54] Christiansen and Tschopp (2010), having studied a well-preserved specimen of Hesperosaurus with skin impressions, concluded that the plates were covered in a keratin sheath which would have strengthened the plate as a whole and provided it with sharp cutting edges.[21] Bakker stated that Stegosaurus could flip its osteoderms from one side to another to present a predator with an array of spikes and blades that would impede it from closing sufficiently to attack the Stegosaurus effectively. He contends that they had insufficient width for them to stand erect easily in such a manner as to be useful in display without continuous muscular effort.[54] Mobility of the plates, however, has been disputed by other paleontologists.[55]

Another possible function of the plates is they may have helped to control the body temperature of the animal,[55] in a similar way to the sails of the pelycosaurs Dimetrodon and Edaphosaurus (and modern elephant and rabbit ears). The plates had blood vessels running through grooves and air flowing around the plates would have cooled the blood.[56] Buffrénil, et al. (1986) found "extreme vascularization of the outer layer of bone",[57][55] which was seen has evidence that the plates "acted as thermoregulatory devices".[57] Likewise, 2010 structural comparisons of Stegosaurus plates to Alligator osteoderms seem to support the conclusion that the potential for a thermoregulatory role in the plates of Stegosaurus definitely exists.[58]

Plates of Stegosaurus specimen "Sophie"

The thermoregulation hypothesis has been seriously questioned, since other stegosaurs such as Kentrosaurus, had more low surface area spikes than plates, implying that cooling was not important enough to require specialized structural formations such as plates. However, it has also been suggested that the plates could have helped the animal increase heat absorption from the sun. Since a cooling trend occurred towards the end of the Jurassic, a large ectothermic reptile might have used the increased surface area afforded by the plates to absorb radiation from the sun.[59] Christiansen and Tschopp (2010) state that the presence of a smooth, insulating keratin covering would have hampered thermoregulation, but such a function cannot be entirely ruled out as extant cattle and ducks use horns and beaks to dump excess heat despite the keratin covering.[21] Histological surveys of plate microstructure attributed the vascularization to the need to transport nutrients for rapid plate growth.[57][60] The vascular system of the plates have been theorized to have played a role in threat displaying as Stegosaurus could have pumped blood into them, causing them to "blush" and give a colorful, red warning.[61] However, Christiansen and Tschopp (2010) consider this unlikely, as stegosaur plates were covered in horn rather than skin.[21]

The plates' large size suggests that they may have served to increase the apparent height of the animal, either to intimidate enemies[23] or to impress other members of the same species in some form of sexual display.[53] A 2015 study of the shapes and sizes of Hesperosaurus plates suggested that they were sexually dimorphic, with wide plates belonging to males and taller plates belonging to females.[62] Christiansen and Tschopp (2010) proposed that the display function would have been reinforced by the horny sheath which would have increased the visible surface and such horn structures are often brightly colored.[21] Some have suggested that plates in stegosaurs were used to allow individuals to identify members of their species.[60] The use of exaggerated structures in dinosaurs as species identification has been questioned, as no such function exists in modern species.[63]

Thagomizer on mounted tail

Debate has been going on about whether the tail spikes were used for display only, as posited by Gilmore in 1914[23] or used as a weapon. Robert Bakker noted the tail was likely to have been much more flexible than that of other dinosaurs, as it lacked ossified tendons, thus lending credence to the idea of the tail as a weapon. However, as Carpenter[16] has noted, the plates overlap so many tail vertebrae, movement would be limited. Bakker also observed that Stegosaurus could have maneuvered its rear easily, by keeping its large hind limbs stationary and pushing off with its very powerfully muscled but short forelimbs, allowing it to swivel deftly to deal with attack.[41] More recently, a study of the tail spikes by McWhinney et al.,[64] which showed a high incidence of trauma-related damage, lends more weight to the position that the spikes were indeed used in combat. This study showed that 9.8% of Stegosaurus specimens examined had injuries to their tail spikes. Additional support for this idea was a punctured tail vertebra of an Allosaurus into which a tail spike fits perfectly.[65]

S. stenops had four dermal spikes, each about 60–90 cm (2.0–3.0 ft) long. Discoveries of articulated stegosaur armor show, at least in some species, these spikes protruded horizontally from the tail, not vertically as is often depicted.[16] Initially, Marsh described S. ungulatus as having eight spikes in its tail, unlike S. stenops. However, recent research re-examined this and concluded this species also had four.[31]

Brain cavity of S. stenops marked with red

At one time, stegosaurs were described as having a "second brain" in their hips. Soon after describing Stegosaurus, Marsh noted a large canal in the hip region of the spinal cord, which could have accommodated a structure up to 20 times larger than the famously small brain. This has led to the influential idea that dinosaurs like Stegosaurus had a "second brain" in the tail, which may have been responsible for controlling reflexes in the rear portion of the body. This "brain" was proposed to have given a Stegosaurus a temporary boost when it was under threat from predators.[10]

This space, however, is more likely to have served other purposes. The sacro-lumbar expansion is not unique to stegosaurs, nor even sauropods. It is also present in birds. In their case, it contains what is called the glycogen body, a structure whose function is not definitely known, but which is postulated to facilitate the supply of glycogen to the animal's nervous system.[66] It also may function as a balance organ, or reservoir of compounds to support the nervous system.[67]

Juveniles of Stegosaurus have been preserved, probably showing the growth of the genus. The two juveniles are both relatively small, with the smaller individual being 1.5 m (4.9 ft) long, and the larger having a length of 2.6 m (8.5 ft). The specimens can be identified as not mature because they lack the fusion of the scapula and coracoid, and the lower hind limbs. Also, the pelvic region of the specimens are similar to Kentrosaurus juveniles.[68] One 2009 study of Stegosaurus specimens of various sizes found that the plates and spikes had delayed histological growth in comparison to the skeleton and when the dinosaur reached maturity, growth in the osteoderms may have greatly outpaced that of the skeleton.[69] A 2013 study concluded, based on the rapid deposition of highly vascularised fibrolamellar bone, that Kentrosaurus had a quicker growth rate than Stegosaurus, contradicting the general rule that larger dinosaurs grew faster than smaller ones.[70]

Stegosaurus and related genera were herbivores. However, their teeth and jaws are very different from those of other herbivorous ornithischian dinosaurs, suggesting a different feeding strategy that is not yet well understood. The other ornithischians possessed teeth capable of grinding plant material and a jaw structure capable of movements in planes other than simply orthal (i.e. not only the fused up-down motion to which stegosaur jaws were likely limited). Unlike the sturdy jaws and grinding teeth common to its fellow ornithischians, Stegosaurus (and all stegosaurians) had small, peg-shaped teeth that have been observed with horizontal wear facets associated with tooth-food contact[71] and their unusual jaws were probably capable of only orthal (up-down) movements.[9] Their teeth were "not tightly pressed together in a block for efficient grinding",[72] and no evidence in the fossil record of stegosaurians indicates use of gastroliths—the stone(s) some dinosaurs (and some present-day bird species) ingested—to aid the grinding process, so how exactly Stegosaurus obtained and processed the amount of plant material required to sustain its size remains "poorly understood".[72]

Tooth crown illustration

The stegosaurians were widely distributed geographically in the late Jurassic.[9] Palaeontologists believe it would have eaten plants such as mosses, ferns, horsetails, cycads, and conifers or fruits.[73] Grazing on grasses, seen in many modern mammalian herbivores, would not have been possible for Stegosaurus, as grasses did not evolve until late into the Cretaceous Period, long after Stegosaurus had become extinct.

One hypothesized feeding behavior strategy considers them to be low-level browsers, eating low-growing fruit of various nonflowering plants, as well as foliage. This scenario has Stegosaurus foraging at most 1 m above the ground.[74] Conversely, if Stegosaurus could have raised itself on two legs, as suggested by Bakker, then it could have browsed on vegetation and fruits quite high up, with adults being able to forage up to 6 m (20 ft) above the ground.[10]

A detailed computer analysis of the biomechanics of Stegosaurus's feeding behavior was performed in 2010, using two different three-dimensional models of Stegosaurus teeth given realistic physics and properties. Bite force was also calculated using these models and the known skull proportions of the animal, as well as simulated tree branches of different size and hardness. The resultant bite forces calculated for Stegosaurus were 140.1 newtons (N), 183.7 N, and 275 N (for anterior, middle and posterior teeth, respectively), which means its bite force was less than half that of a Labrador retriever. Stegosaurus could have easily bitten through smaller green branches, but would have had difficulty with anything over 12 mm in diameter. Stegosaurus, therefore, probably browsed primarily among smaller twigs and foliage, and would have been unable to handle larger plant parts unless the animal was capable of biting much more efficiently than predicted in this study.[75] However, a study published on May 20, 2016 by Stephen Lautenschlager et al. indicates Stegosaurus's bite strength was stronger than previously believed. Comparisons were made between it (represented by a specimen known as "Sophie" from the United Kingdom's Natural History Museum) and two other herbivorous dinosaurs; Erlikosaurus and Plateosaurus to determine if all three had similar bite forces and similar niches. Based on the results of the study, it was revealed that Stegosaurus had a bite similar in strength to that of modern herbivorous mammals, in particular, cattle and sheep. Based on this data, it is likely Stegosaurus also ate woodier, tougher plants such as cycads, perhaps even acting as a means of spreading cycad seeds. The findings were published in the journal Scientific Reports.[76]