Therizinosaurus (; meaning 'scythe lizard') is a genus of very large therizinosaurid dinosaurs that lived during the Late Cretaceous period in what is now Asia. It contains a single species, Therizinosaurus cheloniformis, known from the fossiliferous Nemegt Formation. The first remains of Therizinosaurus were found in 1948 by a Mongolian field expedition in the Gobi Desert and later described by Evgeny Maleev in 1954. The genus is only known from a few bones, including gigantic manual unguals (claw bones), from which it gets its name, and additional findings comprising fore and hindlimb elements that were discovered from the 1960s through the 1980s.
Therizinosaurus was a colossal therizinosaurid that could grow up to long and tall, and weigh possibly over . Like other therizinosaurids, it would have been a slow-moving, long-necked, high browser equipped with a rhamphotheca (horny beak) and a wide torso for food processing. Its forelimbs were particularly robust and had three fingers that bore unguals which, unlike other relatives, were very stiffened, elongated, and only had significant curvatures at the tips. Therizinosaurus had the longest known manual unguals of any land animal, reaching above in length. Its hindlimbs ended in four functionally weight-bearing toes differing from other theropod groups in which the first toe was reduced to a dewclaw and also resembling the very distantly related sauropodomorphs.
It was one of the last and the largest representative of its unique group, the Therizinosauria (formerly known as Segnosauria; the segnosaurs). During and after its original description in 1954, Therizinosaurus had rather complex relationships due to the lack of complete specimens and relatives at the time. Maleev thought the remains of Therizinosaurus to belong to a large turtle-like reptile, and also named a separate family for the genus: Therizinosauridae. Later on, with the discovery of more complete relatives, Therizinosaurus and kin were thought to represent some kind of Late Cretaceous sauropodomorphs or transitional ornithischians, even though at some point it was suggested that it may have been a theropod. After years of taxonomic debate, nevertheless, they are now placed in one of the major dinosaur clades, Theropoda, specifically as maniraptorans. Therizinosaurus is widely recovered within Therizinosauridae by most analyses.
The unusual arms and body anatomy (extrapolated after relatives) of Therizinosaurus have been cited as an example of convergent evolution with chalicotheriines and other primarily herbivorous mammals, suggesting similar feeding habits. The elongated hand claws of Therizinosaurus were more useful when pulling vegetation within reach rather than being used for active attack or defense because of their fragility, however, they may have had some role for intimidation. Its arms also were particularly resistant to stress, which suggests a robust use of these limbs. Therizinosaurus was a very tall animal, likely having a reduced competition over the foliage in its habitat and outmatching predators like tyrannosaurid Tarbosaurus.
History of discovery
thumb|left|upright=1.2|[[Cretaceous fossil localities of Mongolia; Therizinosaurus fossils have been collected from the Altan Uul, Hermiin Tsav, and Nemegt localities at the area A (Nemegt Formation)]]
In 1948, several Mongolian Paleontological expeditions organized by the USSR Academy of Sciences were conducted in the Nemegt Formation of the Gobi Desert, Southwestern Mongolia, with the main objective of new fossils findings. The expeditions unearthed numerous dinosaur and turtle fossil remains from the stratotype locality Nemegt (also known as Nemegt Valley), but the most notable elements collected were three partial manual unguals (claw bones) of considerable size. This set of unguals was found on a subdivision of the Nemegt locality designated as Quarry V near the skeleton of a large theropod, but also in association with other elements including a metacarpal fragment and several ribs fragments. It was labelled under the specimen number PIN 551-483 and later on, these fossils were described by the Russian paleontologist Evgeny Maleev in 1954 who used them to scientifically name the new genus and type species Therizinosaurus cheloniformis, becoming the holotype specimen. The generic name, Therizinosaurus, is derived from the Greek (, meaning scythe, reap or cut) and (, meaning lizard) in reference to the enormous manual unguals, and the specific name, cheloniformis, is taken from the Greek (, meaning turtle) and Latin formis as the remains were thought to belong to a turtle-like reptile. Maleev also coined a separate family for this new and enigmatic taxon: Therizinosauridae. Since little was known of Therizinosaurus at the time of the original description, Maleev thought PIN 551-483 belonged to a large, long turtle-like reptile that relied on its giant hand claws to harvest seaweed.
thumb|Holotype unguals PIN 551–483 at the [[Moscow Paleontological Museum; note left ungual cast]]
Though it was not fully understood to what general kind of animal these fossils belonged, in 1970, the Russian paleontologist Anatoly K. Rozhdestvensky was one of the first authors to suggest that Therizinosaurus was a theropod and not a turtle. He made comparisons between Chilantaisaurus and the holotype unguals of Therizinosaurus to propose that the appendages actually came from a carnosaurian dinosaur, thereby interpreting Therizinosaurus as a theropod. Rozhdestvensky also illustrated the three holotypic manual unguals and re-identified the metacarpal fragment as a metatarsal bone, and based on the unusual shape of both metatarsal and ribs fragments he listed them as sauropod remains. These theropodan affinities were also followed by the Polish paleontologist Halszka Osmólska and co-author Ewa Roniewicz in 1970 during their naming and description of Deinocheirus—another large and enigmatic theropod from the formation that was initially known from partial arms. Similar to Rozhdestvensky, they suggested that the holotype unguals were more likely to have belonged to a carnosaurian theropod, rather than a large marine turtle.
Additional specimens
thumb|left|Arms of specimen MPC-D 100/15 with unguals 100/16 and 100/17 at [[CosmoCaixa Barcelona]]
Further expeditions in the Nemegt Formation unearthed more fossils of Therizinosaurus. In 1968 prior to Rozhdestvensky, Osmólska and Roniewicz statements, the upper portion of a manual ungual was found in the Altan Uul locality and labeled as MPC-D 100/17 (formerly IGM or GIN). In 1972, another fragmented ungual (specimen MPC-D 100/16) was discovered at the Upper White Beds of the Hermiin Tsav locality, only preserving its lower portion. During the year 1973, a much more complete, larger, and articulated specimen was collected also from Hermiin Tsav. This specimen was labelled as MPC-D 100/15 and consists of both left and right arms including the scapulocoracoids, both humeri (upper arm bones), right ulna with radius and left ulna, two right carpals, the right metacarpus including a complete digit Il, and some ribs with gastralia (belly ribs). As common with fossils, some elements were not entirely preserved such as the scapulocoracoids with broken ends, and the left arm is less complete than the right one. All of these specimens were first described and referred to Therizinosaurus by the Mongolian paleontologist Rinchen Barsbold in 1976. In this new monograph, he pointed out that the rib fragments in MPC-D 100/15 were more slender than the ones from the holotype, and identified MPC-D 100/16 and 100/17 as pertaining to digits I and III, respectively. It was clear to Barsbold that MPC-D 100/15 represented Therizinosaurus as the ungual in this specimen shared the elongation and flattened morphology of all previous specimens. He concluded that Therizinosaurus was a theropod taxon since MPC-D 100/15 matched multiple theropodan characters. In 1990, Barsbold and Teresa Maryanska agreed with Perle in that the hindlimb material from Hermiin Tsav he described in 1982 was therizinosaurian (then called segnosaurians) given that the metatarsus was stocky and the astragalus had a laterally arched ascending process (bony extension), but cast doubt with his referral of it to Therizinosaurus and the segnosaurian identity for this taxon since it was only known from the pectoral girdle and other forelimb elements, making direct comparisons between specimens impossible. They considered this specimen to represent a Late Cretaceous representative of the Segnosauria, but not Therizinosaurus.
In 2010 however, the North American paleontologist Lindsay E. Zanno in her large taxonomic reevaluation of Therizinosauria considered the referral of MPC-D 100/45 to Therizinosaurus to be likely based on the rationale that it was collected in the same stratigraphic context (Nemegt Formation) as the holotype, and shared the robust and four-toed morphology of other therizinosaurids such as Segnosaurus. She also excluded the rib material from the holotype as it was re-identified by Rozhdestvensky to likely have come from a sauropod dinosaur, and not Therizinosaurus itself.
Description
thumb|left|Size of two specimens ([[holotype in yellow, right) compared to a tall human]]
By maniraptoran standards, Therizinosaurus attained enormous sizes, estimated to have reached in length. In 2010, Gregory S. Paul estimated in his book The Princeton Field Guide to Dinosaurs that the animal weighed around 3 tons. However, the estimate was revised to in a subsequent edition. In 2013, Lindsay E. Zanno and Peter J. Makovicky estimated that Therizinosaurus weighed around . These dimensions make Therizinosaurus the largest known therizinosaur and the largest known maniraptoran. Along with the contemporaneous ornithomimosaur Deinocheirus, it was the largest maniraptoriform. Though body remains of Therizinosaurus are relatively incomplete, inferences can be made about its physical characteristics based on more complete relatives. Like other members of its family, Therizinosaurus had a proportionally small skull, bearing a rhamphotheca (horny beak), atop a long neck; a semi-erect, bipedal gait; a large gut for foliage processing; and sparse feathering. Other traits that were likely present in Therizinosaurus include a heavily pneumatized (air-filled) vertebral column and a robustly-built, (backwards oriented) .
thumb|[[Paleoart|Life restoration]]
In 2010, Senter and James used hindlimb length equations to predict the total length of the hindlimbs in Therizinosaurus and Deinocheirus. They concluded that an average Therizinosaurus may have had approximately long legs. More recently, Mike Taylor and Matt Wedel suggested that the whole neck would be 2.9 times the size of the humerus, which was , resulting in a long neck based on comparisons with the series of Nanshiungosaurus. The most distinctive feature of Therizinosaurus was the presence of gigantic unguals on each of the three digits of its hands. These were common among therizinosaurs but particularly large and stiffened in Therizinosaurus, and they are considered as the longest known from any terrestrial animal. A year later, the new genus Erlikosaurus was named by Barsbold and Perle in 1980. They named a new infraorder called the Segnosauria, composed of Erlikosaurus and Segnosaurus. They also noted that while aberrant and having ornithischian-like pelves, segnosaurs featured similar traits to other theropods. With the discovery of the referred hindlimb to Therizinosaurus in 1982 by Perle, he concluded that Segnosaurus was very similar to the latter based on the morphology and they possibly belonged to a single group.
thumb|Mounted skeleton of [[Alxasaurus at the Royal Tyrrell Museum. Alxasaurus helped to resolve important aspects about the affinities of the long-enigmatic therizinosaurs]]
Consequently, the assignment of segnosaurs started to shift towards sauropodomorphs. In 1984, Gregory S. Paul claimed that segnosaurs, rather than being theropods, were instead morphological and evolutionary intermediates between sauropodomorphs and ornithischians, late survivors of an early herbivorous dinosaur radiation which had persisted into the Cretaceous. He maintained his position in 1988 by placing the Segnosauria into the newly-erected and now-obsolete Phytodinosauria, and was among the first to suggest a segnosaur assignment for the enigmatic Therizinosaurus. Other prominent paleontologists like Jacques Gauthier and Paul Sereno supported this vision. In 1990, Rinchen Barsbold and Teresa Maryańska agreed in that the hindlimb material from Hermiin Tsav referred to Therizinosaurus in 1982 was segnosaurian since it matched several traits, but considered it unlikely to belong to the genus and species as there was no overlapping material. Barsbold and Maryańska also disagreed with previous researchers who classified Deinocheirus as a segnosaur.
thumb|left|Feather impressions from the holotype of [[Beipiaosaurus]]
With the unexpected discovery of Alxasaurus in 1988, and its description in 1993, the widely accepted sauropodomorph affinities of segnosaurs were questioned by paleontologists Dale Russell and Dong Zhiming. This new genus was far more complete than any other segnosaur and multiple anatomical features indicated that it was related to Therizinosaurus. Its description allowed for the recognition of Segnosauridae and Therizinosauridae as a single clade; as the second name predates the first, it holds taxonomic priority. Due to some primitive characters present in Alxasaurus, they coined a new taxonomic rank, the superfamily Therizinosauroidea, containing it and Therizinosauridae. All of the new information provided data on the affinities of the new-named therizinosauroids. Russell and Dong concluded that they were theropods with unusual features. In 1994, Clark and colleagues redescribed the very complete skull of Erlikosaurus and even more theropod traits were found this time. They also validated the synonymy of the Segnosauridae with Therizinosauridae and considered therizinosauroids as maniraptoran dinosaurs. In 1997, Rusell coined the infraorder Therizinosauria in order to contain all segnosaurs. This new infraorder was composed of Therizinosauroidea and the smaller, more derived clade Therizinosauridae. Consequently, Segnosauria became a synonym of Therizinosauria. A small, feathered therizinosauroid from China, Beipiaosaurus, was described in 1999 by Xu Xing and colleagues. It confirmed the placement of therizinosaurs, among theropods and also their taxonomic place on the Coelurosauria. The discovery also indicated that feathers were highly distributed among theropod dinosaurs.
In 2010, Lindsay Zanno revised the taxonomy of therizinosaurs in extensive detail. She found that many parts on therizinosaur holotype and referred specimens were lost or damaged, and sparse specimens with no overlapping elements were disadvantages when concluding the relationships of the members. Zanno accepted the referral of the specimen IGM 100/45 to Therizinosaurus since it matches multiple therizinosaurid traits, but decided not to include the specimen in her taxonomic analysis due to the lack of comparative forelimb remains. She also excluded the supposed ribs that were present on the holotype since they likely came from a different animal and not Therizinosaurus.
thumb|upright=1.5|Skeletal reconstructions of various therizinosaurian genera (not to scale); Therizinosaurus in top left
Paleobiology
Feeding
thumb|Life restoration of [[Erliansaurus in a sitting stance. Therizinosaurus may have sat on its pelvis during feeding]]
In 1993 Dale A. Russell and Donald E. Russell analyzed Therizinosaurus and Chalicotherium, and noted similarities in their respective body plan, even though they form part of different groups. Both genera had large, well-developed, and relatively strong arms; the pelvic girdle was robust and suited for a sitting behavior; and the hindlimb (particularly the foot) structure was robust and shortened. They considered these adaptations to represent an example of convergent evolution—a condition where organisms evolve similar traits without necessarily being related—between extinct mammal and dinosaur genera. Moreover, the body plan is somewhat exhibited by the modern-day gorillas. Because the animals with this type of body plan are known to represent herbivores, the authors suggested this lifestyle for Therizinosaurus. Russell and Russell reconstructed the feeding behavior of Therizinosaurus as being able to sit while consuming foliage from large shrubs and trees. The plant material would have been harvested with its hands and this action was likely favored by its elongated neck which prevented the use of large amounts of force and effort. As its arms were long enough to have touched the ground during certain stances, they could have helped the dinosaur to rise from a prone position. If browsing in a bipedal stance, Therizinosaurus may have been able to reach even higher vegetation supported by its short and robust feet. Whereas Chalicotherium was more suited to hook branches, Therizinosaurus was better at pushing large clumps of foliage because of its long claws. It is also possible that Therizinosaurus was less capable of great precision in its movements than was Chalicotherium, due to the latter having more developed brain, dental and muscular capacities.
Anthony R. Fiorillo and colleagues in 2018 suggested that Therizinosaurus had a reduced bite force that may have been useful for cropping vegetation or foraging, based on relative therizinosaurids such as Erlikosaurus and Segnosaurus. As the bite force started to decrease from primitive to derived therizinosaurians, Therizinosaurus, being a derived member, would have been subject to the evolutionary relationship.
Arms and claws function
thumb|left|Lateral view of the unguals in specimens MPC-D 100/15, 100/16, and 100/17. Therizinosaurus may have used its claws for grasping [[foliage]]
When the genus was first described by Maleev in 1954, he considered that the unusually large claws were used to harvest seaweed. This was however, based on the assumption of a giant marine turtle.
In 2014, Lautenschlager tested the function of various therizinosaur hand claws—including Therizinosaurus—through digital simulations. Three different functional scenarios were simulated for each claw morphology with a force of 400 N applied in each scenario: scratch/digging; hook-and-pull; and piercing. Though the stocky claws of Alxasaurus resulted in low-stress magnitudes, the stress was greater with the curvature and elongation of the claws in Falcarius, Nothronychus and Therizinosaurus. Some of the highest stress, deformation, and strain magnitudes were obtained in the scratch/digging scenario; the hook-and-pull scenario, in contrast, resulted in lower magnitudes, and lesser ones were found in the piercing scenario. Particularly, the overall stress was most pronounced in the unusual claws of Therizinosaurus, which may represent an exceptional case of elongation specialization. Lautenschlager noted the more strongly curved and elongate claws of some therizinosaurian taxa were poorly functional in a scratch/digging fashion, indicating this as the most unlikely function. Though fossorial (digging) behavior has been reported in several dinosaur species, the large body size largely rules out the possibility of burrow digging in therizinosaurs. Nevertheless, an overall digging action would have been done with the foot claws because, since as in other maniraptorans, feathers on the arms would have interfered with this function. Instead of being used for fossorial behavior, it is more likely that Therizinosaurus make use of its hands in a hook-and-pull fashion to pull or grasp vegetation within reach. This herbivorous behavior would make therizinosaurs mostly similar to the extant anteaters and the extinct ground sloths. Lautenschlager could neither confirm nor disregard that the hand claws could have been used for defense, intraspecific competition, stabilization by grasping tree trunks during high browsing, sexual dimorphism, or gripping mates during mating given the lack of more specimens. He clarified that there is no evidence that the long claws of Therizinosaurus would have been used in active defense or attack; however, it is possible that these appendages could have had some role when facing a threat, such as intimidation.
thumb|Manus of Therizinosaurus tested in a hook and pull scenario
Scott A. Lee and Zachary Richards in 2018 based on bending resistance measurements of several dinosaur humeri, found the humeri of carnosaur, therizinosaur, and tyrannosaur dinosaurs to be relatively resilient to stress. This increased ability to withstand stress supports the idea that Therizinosaurus and other therizinosaurians used their arms in a robust fashion that generated significant forces. They also suggested that the prominent claws of some members could have been used as a defense against predators and other various functions. Unlike the generally light and agile ornithomimosaurs who avoided predation with speed, Therizinosaurus and relatives relied on arms and claws to face threats (and were generally slow-runners to begin with).
A 2023 study by Qin, Rayfield, Benton, et al., regarding the claw function in therizinosaurids and alvarezsaurids, which represent the extremes of theropod claw morphology, suggest that there was no mechanical function identifiable for Therizinosaurus, suggesting the claws on its forelimbs were merely decorative rather than functional and a result of peramorphic growth resulting from increased body size.
Paleoenvironment
thumb|left|The Nemegt Formation may have had similar environments to those of the swampy [[Okavango Delta]]
The remains of Therizinosaurus have been found in the well-known Nemegt Formation of the Gobi Desert. Although this formation has never been dated radiometrically because of the discontinuity of exposures and absence of datable volcanic rock facies, the vertebrate fossil assemblage suggests a Maastrichtian stage (and possibly Late Campanian). The Nemegt Formation is separated into three informal members. The lower member is mainly composed by fluvial sediments, while middle and upper members consist of alluvial plain, paludal, lacustrine, and fluvial sedimentation. Therizinosaurus is known from the Altan Uul, Hermiin Tsav, and Nemegt localities, belonging to the lower and middle members of the formation. The wet environments of the Nemegt Formation may have acted as an oasis-like area that attracted oviraptorids from arid neighbour localities such as the Barun Goyot Formation, as evidenced on the presence of Nemegtomaia in both regions. It has been previously suggested that the Nemegt Formation may have been similar to the modern-day Okavango Delta, which is also composed of mesic (well-watered) surroundings.
The area would have been semi-arid during certain times of the year.
thumb|Life restoration of a Therizinosaurus pair along with a small group of [[Prenocephale, and Adasaurus]]
The paleofauna of the Nemegt Formation was diverse and rich, composed of other dinosaurs such as the alvarezsaurs Mononykus and Nemegtonykus; deinonychosaurs Adasaurus, and Zanabazar; ornithomimosaurs Anserimimus and Gallimimus; oviraptorosaurs Avimimus, Conchoraptor, Rinchenia, and Elmisaurus; tyrannosaurids Alectrosaurus, Alioramus, and possibly Bagaraatan; ankylosaurids Saichania and Tarchia; and pachycephalosaurids Homalocephale and Prenocephale. The Nemegt megafauna included the ornithomimosaur Deinocheirus; hadrosaurids Barsboldia and Saurolophus; titanosaurs Nemegtosaurus and Opisthocoelicaudia; and the apex predator Tarbosaurus. Additional paleofauna includes birds like Judinornis or Teviornis; abundant freshwater ostracods at numerous localities; fish; terrestrial and aquatic turtles such as Mongolochelys and Nemegtemys; and the crocodylomorph Paralligator.
As the sediments in which Therizinosaurus remains have been found are fluvial-based, it is suggested that it may have preferred to forage on riparian areas.