So a mere nineteen years after its initial description in a tabloid, Gigantoraptor was finally described in useful detail. Afrovenator, Suchomimus, Deltadromeus and Jobaria are jealous. This post involves two aspects of the new description compared to its original skeletal reconstruction- completeness and accuracy.
Back in 2007 we got the initial description of Gigantoraptor in a tabloid (Nature). Fairly useless as most tabloid descriptions are when there aren't tens of supplementary info pages to back them up. The mandible was described in 2017, but now in 2026 we have a description of the postcranium in JVP. It took a while, but sounds great. Here's the skeletal reconstruction from the initial 2007 paper, explicitly described in the caption as "Skeletal reconstruction showing preserved elements..."
 |
| Gigantoraptor erlianensis holotype LH V 0011 skeletal reconstruction supposedly showing preserved elements in white (after Xu et al., 2007). |
Upon reading the new osteology I noticed...
- posterior cervical neural arch- DOESN'T EXIST.
- eight dorsal vertebrae missing neural spines- DON'T EXIST.
- nine complete to partial dorsal ribs- DON'T EXIST.
- sacrum- DOESN'T EXIST.
- full gastral basket- NO GASTRALIA PRESERVED AT ALL.
- complete caudal series of 27 vertebrae- ONLY NINE CAUDALS PRESERVED (five with significant missing parts).
- fourteen chevrons- ONLY THREE PRESERVED (two incomplete).
- furcula- DOESN'T EXIST.
- sternum- DOESN'T EXIST.
- sternal ribs- DON'T EXIST.
- scapholunare/radiale and semilunate carpal- DON'T EXIST.
- distal metacarpal III- DOESN'T EXIST.
- manual phalanges I-1, II-1, proximal II-2, III-1 and III-2- DON'T EXIST.
- manuals unguals I-III- ONLY ONE PRESERVED.
- ventral ilium- DOESN'T EXIST.
- metatarsals I and V- DON'T EXIST.
- pedal phalanx I-1 and pedal ungual I- DON'T EXIST.
And for completion's sake, the proximal pubis and mid femoral shaft aren't preserved, but I would forgive that if the rest wasn't so blatantly wrong.
Obviously we're all used to schematic skeletal reconstructions (Exhibit A- any Sereno paper), where maybe we have a lot of ribs but don't know exactly where they go, so you fill in all the ribs. Messy but not bloggable. But here the artists (I guess "R.-S. Li and W.-D. Zhang of the Institute of Vertebrate Paleontology and Paleoanthropology") chose to color in with white only dorsal ribs 1, the distal parts of 2-6 and all of 9-11. Why? Ditto only coloring in the one cervical neural arch, and only the ends of manual phalanx II-2 and metacarpal III. It REALLY gives the impression some precise work is being done to illustrate only the preserved materials. But apparently not?
 |
| Gigantoraptor erlianensis holotype LH V 0011 skeletal reconstruction with only elements stated to exist by Hao et al. (2026 online) in white (modified from Xu et al., 2007). Note the two distal chevrons aren't necessarily the two incomplete elements that are preserved as these aren't described, and the position of the caudals is approximate. Counterclockwise from upper left are the actual preserved mandible, scapula, humerus, metacarpus, proximal femur, tibia plus proximal tarsals and fibula (most after Hao et al., 2026 online except mandible modified from Ma et al. 2017 and metacarpus after Xu et al., 2007). |
For my second complaint, sure maybe the reconstruction doesn't show what's known, but at least it presents the preserved parts correctly? Nope, not even close. Among the most obvious differences, in the actual material-
- the scapula is sigmoid and decreases in height distally.
- the humerus lacks a significant deltopectoral crest apex.
- metacarpals I and II are very narrow (obvious on the original figure, so that might have been a red flag...).
- the trochanteric crest strongly projected anterodorsally.
- the cnemial crest is low and proximally limited, with fibular condyle strongly and abruptly projected posteriorly.
- the fibular shaft is very slender distally with posteriorly angled proximal edge.
- none of the metatarsals are shaped correctly in the recon, the pedal phalanges are not constricted ventrally nearly enough there and phalanx IV-4 is weirdly short.
Of course, taphonomy is a thing and it's clear e.g. the distal femur and tibia are anteroposteriorly compressed. Maybe most of the humerus is too. But that doesn't describe the differences here. It's as if the artist was just told to draw a generic oviraptorosaur leg, and there ya go! Again, the differences in shape between the mandible illustration and the reconstruction should have been a red flag, as the latter lacks the sigmoidal dorsal and ventral edges of the much lower external mandibular fenestra, and is longer posterior to the fenestra than anterior to it, unlike the actual fossil.
While not part of my initial post idea, I wondered if despite the differences in shape and preserved elements, maybe the major ones were at least proportioned correctly. I used the humerus as a standard length (679 mm) since its proximal and distal ends are visible in the reconstruction and it was complete unlike the femur, tibia or pubis. The results are-
element real length (mm) recon length scaled to humerus (mm) recon/real
Mandible 460 551 1.20
Radius 560 538 0.96
Ulna 563 551 0.98
Metacarpal I 93 99 1.06
Metacarpal II 324 317 0.98
Tibia 1150 1200 1.04
Fibula 1100 1150 1.05
Metatarsal II 501 554 1.11
Metatarsal IV 538 624 1.16
The skull and pes in particular are too large in the reconstruction. The rest is okay, but really getting elements within 10% of their actual size should be the bare minimum expectation.
If you go back to its initial description (Xu et al., 2007), the holotype is listed as- "an incomplete skeleton preserving a nearly complete mandible, several partial presacral vertebrae, most caudal vertebrae, nearly complete right scapula, much of the forelimbs, partial ilium, and nearly complete pubes and hind limbs."
The mandibular description (Ma et al., 2017) lists the holotype as- "a nearly complete mandible and several postcranial bones (right scapula, forelimbs, presacral vertebrae, ilium, pubes, hindlimbs and caudal vertebrae)"
Note it's also a left scapula, and only the left forelimb and hindlimb are preserved except for a right humerus and distal femur. But those sources do suggest at least presacral vertebrae and an ilium were initially referred to the holotype. That it wasn't just an artistic error. Xing Xu was an author on all of these, so it's not likely to be a telephone type error. Similarly, Qingwei Tan and Ling Tan were authors on the initial and mandibular papers, and the former is the second author of the postcranium paper.
So what are we to make of this? I would guess some presacral vertebrae and an ilium were initially preserved. Were they not collected (maybe)? Were they reidentified (unlikely)? Were they so poorly preserved that the postcranium description ignored them? Again maybe, but it would be a bad precedent. Another possibility is they were lost/stolen, as seems to happen especially frequently in Chinese museums. But if any of these reasons are even partially true, it's the authors' job to report it in the official public document.
On the other hand, the artist of Gigantoraptor's initial Figure 1a was either not qualified to draw a skeletal reconstruction or given only vague information about what they were supposed to do without photos of real material or official measurements. Either way, the current paper should have explained it, saying the previous reconstruction was approximate and/or designed on elements they thought were preserved. And no matter what combination of explanations turns out to be real, I'm betting at the very least the sternum, sternal ribs and furcula were lazy errors to keep colored white.
References- Xu, Tan, Wang, Zhao and Tan, 2007. A gigantic
bird-like dinosaur from the Late Cretaceous of China. Nature. 844-847.
Ma, Wang, Pittman, Tan, Tan, Guo and Xu, 2017. Functional anatomy of a
giant toothless mandible from a bird-like dinosaur: Gigantoraptor and the evolution of
the oviraptorosaurian jaw. Scientific Reports. 7: 16247.
Hao, Tan and Xu, 2026 online. Postcranial osteology of the giant oviraptorosaur Gigantoraptor
erlianensis. Journal of Vertebrate Paleontology. e2716748.
