Citipati

Caudal vertebra from the holotype

Though the first specimen of Citipati (IGM 100/971) was briefly reported and discussed, Norell and colleagues in 2001 provided an extensive description of this specimen. As the description was published prior to the formal naming of Citipati, Norell and team tentatively referred this small embryo to a "new large species from Ukhaa Tolgod"—in fact, later known as Citipati osmolskae—based on the shared tall premaxilla morphology among specimens.[9] The more famous IGM 100/979 was extensively described by Clark and team in 1999, also prior to the naming of Citipati. They considered this specimen to be most similar and closely related to Oviraptor than to the other oviraptorids known at that time.[10] Despite being discovered in 1995, the specimen IGM 100/1004 remained partially figured and largely undescribed for years until its formal referral to the taxon Citipati osmolskae in 2018 by Norell and team.[7]

The largest and most complete specimen of Citipati is represented by the holotype IGM 100/978, however, it was preliminarily described and figured in 2001 during the naming of the taxon and during that time, the specimen had not been completely prepared.[8] The skull anatomy of the specimen was later described by Clark and colleagues in 2002,[11] the furcula morphology in 2009 by Sterling J. Nesbitt with team,[12] and the caudal vertebrae by W. Scott Persons and colleagues in 2014 who noted the presence of a pygostyle.[13] Subsequent descriptions have been published in 2018 by Norell and team describing and illustrating some cervical vertebrae and uncinate processes,[7] and Amy M. Balanoff and colleagues describing the endocranium anatomy.[14]

Mounted IGM 100/42; this specimen has been largely used as a reference for Oviraptor an could represent a second species of Citipati

The Zamyn Khondt oviraptorid is a well-known oviraptorid represented by a single and rather complete specimen (IGM 100/42) collected from the Zamyn Khondt (also spelled as Dzamin Khond) locality of the Djadokhta Formation. Since the type skull and body remains of Oviraptor are crushed and partially preserved, the Zamyn Khondt oviraptorid had become the quintessential depiction of the former, even appearing in scientific literature with the label Oviraptor philoceratops.[15]

Clark with team have pointed out that this distinctive-looking, tall-crested oviraptorid has more features of the skull in common with Citipati than it does with Oviraptor. Though being different in the crest shape of the skull, the Zamyn Khondt oviraptorid is similar to Citipati in the shape of the narial region and premaxilla morphology. They considered this oviraptorid to belong to the genus, however, they could neither confirm nor disregard that this specimen represents a second species of Citipati.[8][11] Lü Junchang and colleagues in 2004 found this specimen to be closely related to Oviraptor,[16] Phil Senter with team in 2007 placed it close to neither genus,[17] and in 2020 Gregory F. Funston and colleagues found it to be the sister taxon of Citipati.[18]

Comparison of C. osmolskae and IGM 100/42 skulls

Its skull was unusually short and highly pneumatized (riddled with air-spaced openings), ending in a stout, toothless beak. Perhaps the most distinctive feature of Citipati was its tall crest, superficially similar to that of a modern cassowary. The crest was relatively low in C. osmolskae formed by the premaxilla and nasal bones of the skull, with a nearly vertical front margin grading into the beak. In contrast, the crest of IGM 100/42 was taller with a prominent notch in the front margin, creating a squared appearance.[11]

Citipati embryo IGM 100/971

Life restoration of IGM 100/971

The embryo-bearing egg was otherwise identical to other oviraptorid eggs in shell structure and was found in an isolated nest, again arranged in a circular pattern. Two skulls belonging to very young or embryonic Byronosaurus were found associated with the same nest as the first Citipati embryo. It is possible that these tiny troodontids were preyed upon by the Citipati to feed its nest. Alternately, Mark Norell suggested that the juvenile troodonts were raiding the Citipati nest, or even that an adult Byronosaurus had laid eggs in a Citipati nest as an act of nest parasitism.[1][2]

Although fossilized dinosaur eggs are rare, Citipati eggs and oviraptorid eggs in general, are relatively well known. Along with the two known nesting specimens, dozens of isolated oviraptorid nests have been uncovered in the Gobi Desert. Citipati eggs are elongatoolithid, which are shaped like elongated ovals and resemble the eggs of ratites in texture and shell structure. In the nest, Citipati eggs are typically arranged in concentric circles of up to three layers, and a complete clutch may have consisted of as many of 22 eggs.[22] The eggs of Citipati are the largest known definitive oviraptorid eggs, at 18 cm. In contrast, eggs associated with Oviraptor are only up to 14 cm long.[10]

The two nesting specimens of Citipati are situated on top of egg clutches, with their limbs spread symmetrically on each side of the nest, front limbs covering the nest perimeter. This brooding posture is found today only in birds and supports a behavioral link between birds and theropod dinosaurs.[10] The nesting position of Citipati also supports the hypothesis that it and other oviraptorids had feathered forelimbs. Thomas P. Hopp and Mark J. Orsen in 2004 analyzed the brooding behavior of extinct and extant dinosaur species, including oviraptorids, in order to evaluate the reason for the elongation and development of wing and tail feathers. Given that IGM 100/979 was found in a very avian-like posture, with the forelimbs in a near-folded posture and the pectoral region, belly, and feet in contact with the eggs, Hopp and Orsen indicated that long pennaceous feathers and a feather covering were most likely present in life. The "wings" and tail of oviraptorids would have granted protection for the eggs and hatchlings against climate factors like the sunlight, wind, and rainfalls. However, the arms of this specimen were not extremely folded as in some modern birds, instead, the are more extended resembling the style of large flightless birds like the ostrich. The extended position of the arm is also similar to the brooding behavior of this bird, which is known to nest in large clutches like oviraptorids. Based on the forelimb position of nesting oviraptorids, Hopp and Orsen proposed brooding as the ancestral reason behind wing and tail feather elongation, as there was a greater need to provide optimal protection for eggs and juveniles.[23]

In 2014, W. Scott Persons and colleagues suggested that oviraptorosaurs were secondarily flightless and several of the traits in their tails may indicate a propensity for display behaviour, such as courtship display. The tail of several oviraptorosaurs and oviraptorids ended in pygostyles, a bony structure at the end of the tail that, at least in modern birds, is used to support a feather fan. Furthermore, the tail was notably muscular and had a pronounced flexibility, which may have aided in courtship movements.[13]