RECONSTRUCTING THE GIANT SLOTH DINOSAUR, THERIZINOSAURUS

I first learnt about the giant therizinosaur Therizinosaurus cheloniformis from the Walking with Dinosaurs special “The Giant Claw” with Nigel Marven -  and since then … I wasn’t really interested in it. Of course, I loved the episode (sea monsters better) but I was never enthralled by these bizarre long clawed herbivorous theropods – however, this description gives far too little credit to these amazing animals, and I’ve begun to truly appreciate these wonderful animals. At the recent TetZooCon I attended a talk by Jack Ashby about the Platypus – Jack brought up an interesting point, that we tend to “other” these animals by calling them bizarre / weird when in fact, these animals have solved an evolutionary problem by, basically, becoming really good at what they do. From here on out, I won’t refer to therizinosaurs as anything other than amazingly adapted and evolved animals perfectly fit for their ecological role, which is stuffing plants in to their head and digesting it at an enormous rate. (CONT 1.)

EDIT 9/12/22:

A few people have commented on the size of the pelvic girdle compared to the body, in that it looked too small - and I thought I’d touch on that. This is due to the contradictory measurements in the Suzhousaurus You et al. (2008) paper. The detail in which they took the measurements for the pelvic girdle is very well figured, all straight line measurements - but the problem is, they contradict each other. For example, the stated length of the ilium is 79 cm (which is how it is displayed in my twitter skeletal, and this seems to fit pretty well with the sacrum when using the provided sacrum length. However, the other measurements contradict - the ischium length is displayed as 66 cm, which would result in an enormous pelvic girdle, and the preacetabular length is 35 cm (which seems to match pretty well with the given scale bar, and doesn’t produce an outlandish sized pelvis).

Therefore, I decided to present two versions of the skeletal. As I used the pelvic girdle of Suzhousaurus to estimate the size of the pectoral girdle and arms in Therizinosaurus, this affects the size. My larger of the versions is my first published version, version 1, and the smaller of the versions is version 2, with the larger pelvic girdle (shown below). I’ll let you decide which to use based on your preferences.

Version 1 - using Suzhousaurus ilium length and comparison with sacrum

Version 2 - using Suzhosaurus preacetabular + scale bar length

…(CONT. 2)             

Therizinosaurs take the theropod anatomy to an extreme – in some cases resembling sauropods and other ornithischians: the long, highly pneumatic neck with enlarged cervicals compared to the dorsal body, the distinct kink in the tail (at least in Nothronychus - we’ll get to this later), the rectangular, flat tibia lacking a large cnemial crest seen in other theropods… you get the point. But what about Therizinosaurus in particular?

              Therizinosaurus is not known by much – but what we do have of the animal, show that it was very big. The humerus of referred specimen MPC-D 100/15 is 760 mm long, which is massive even for therizinosaur standards. The claws are also equally massive for therizinosaurs, which although massive in other related genera are not nearly as massive compared to the rest of the arm (in Therizinosaurus the 2nd digit claw is nearly as long as the radius). We also have a partial pes and tibia from a specimen, MPC-D 100/45, which also details an astragalus which fits along a large length of the tibia (seen in other therizinosaurs and some maniraptorans like oviraptorosaurs etc.) – perhaps this made it easier for the robust metacarpals to rotate parallel to the ground in a resting posture? Who knows. What we can tell from the pes is that, like other therizinosaurs, the hallux (digit 1) was enlarged, and the toe would have contacted the ground, unlike most other theropods (interestingly, this is a trait shared in Spinosaurus).

Do we have an idea of where it fits phylogenetically? From a recent paper describing the closely related Paralitherizinosaurus, Therizinosaurus clades with this taxa alongside the Nothronychus sized Suzhousaurus – which is thankfully decently complete and well figured. Thus, I used this taxa to fill in much of the gaps alongside Nothronychus. Suzhousaurus has some notable differences from Nothronychus – for example a much stouter, wider scapula/coracoid, and the presence of five sacral vertebrae rather than 6(?) in Nothronychus (statements are contradictory – it’s possible the “6th” sacral is a fused caudal vertebra). But regardless, I’ll move on to each section of the body below:

VERTEBRAL COLUMN

The first thing I needed to figure out was the vertebral count. Most theropods have roughly 9-10 cervicals, and 13-14 dorsals. It’s almost certain therizinosaurs did this differently – Nanshiungosaurus, another therizinosaur, has at least 11 cervicals (excluding the atlas, so 12 in total), a preserved 10 dorsals and 5 sacrals (Dong, 1979). Going by the total number of presacrals in most theropods, it can be assumed that the dorsal count was roughly 12. Suzhousaurus also has 5 sacrals, but this is the general state in most other theropods (excluding other maniraptorans and a few other clades).  Neimongosaurus preserves 13 cervicals however, which increases the possibility that my current model could switch 1 or 2 dorsals to the cervical column, or vice versa. I decided to go for a longer dorsal series to fit more alongside Nanshiungosaurus, a more derived therizinosaur.

As is probably evident, there is a slight upwards bend to the sacral vertebrae – unlike Nothronychus (from conversation with Scott Hartman) which has a much more pronounced sacral bend and thus a dorsal column that projects upward much further. To fit with the COM, I rotated the vertebral column slightly upward.

Now, that Kink in the tail! This was recovered when articulating the vertebrae of Nothronychus (which I used to restore most of the tail). This bend seems to be derived from two separate vertebrae, which unlike the other caudals have centra that angle inwards/outwards. If it is assumed that these two caudals are contacting, then the bend would be much more dramatic than this - but this doesn’t seem to be the case as the more posterior of the two has a more elongated centrum comparatively, suggesting it’s 1 or two places behind in the series. Moving it back any further is not an option, as the other caudals decrease rapidly in size. Thus - a bend, not too dissimilar to some sauropods. Wonder what it could be for?

PECTORAL AND PELVIC GIRDLE

The pectoral girdle in Therizinosaurus is mostly preserved, including much of the arm (as discussed before). TRhe scapula of Suzhousaurus was used to restore the top of the shaft. I wanted to show the curvature of the claws, so in the background forelimb I decided to go for a semi pronated arm look (it looks cool) – I’m gonna link a blog by Albertonykus here which they sent to me detailing the possible level of pronation/supination (spoiler alert, they would not have been able to fully supinate their palms).

              For the pelvic girdle, I used the hip of Suzhousaurus specimen FRDC-GSJB-2004–001 (which I also used for most of the vertebral column) which is essentially complete. As touched upon earlier, the tibia is very rectangular and box like (thanks to Shane Wheatley for pointing this out). The large pedal unguals were based on Erlikosaurus, which is also what I based a decent portion of the skull on.

SKULL

The skull was based mainly around one brilliant specimen of Erlikosaurus, MPC-D 100/111. Kobayashi et al. (2022) found that Therizinosaurus clades alongside the “Bissekty therizinosaur”, which luckily for us preserves a few portions of the skull, such as a frontal and a dentary - we can tell from the dentary that the skull would have been much more downturned in life compared to Erlikosaurus, a feature also present in the dentary of Segnosaurus (which also preserves the rest of the mandible). Thus, I took this all in to account and edited the skull accordingly. Both Segnosaurus and the Bissekty therizinosaur have a toothless end of the dentary, which would have perhaps been the site of a keratinized beak in life (this was reflected in the silhouette).

The final thing to be tackled was the Centre Of Mass (COM) - with such a small tail, this was imperative, otherwise in life the animal would have fallen face first - so getting the angle of the leg and the body was incredibly important in this reconstruction. Once again, Toxic Midget on discord came through and made a wonderful volumetric analysis as well as a COM for the animal (with a makeshift top view I produced using the material of various derived therizinosaurs) shown below:

Model thanks to Toxic_Midget - dot in middle is where the COM is - though it was likely further back due to the completely apneumatic sacral vertebrae (plus I angled the tibia further forward in my final reconstruction so the toes were bang in the middle)

As you can tell, Therizinosaurs were very wide around the hips and the posterior torso - this made them incredible at eating lots of plant matter and digesting it. This also helps keep the COM fairly near the hip, where all the mass of the animal is. I needed to angle the femur forward to accomodate slightly, however, which made me think about the possibility that this would put excess stress on the leg muscles. My talk with Scott Hartman (who has tackled therizinosaurs before) was very helpful, Scott showed me a video of bipedal elephants (yes, poor elephants being forced to walk on two legs, but they can still do this in the wild) whos femur is set out at a 45 degree angle from the hip, with the tibia/fibula being perpendicular to the ground. Scott’s point was that therizinosaurs didn’t need to be pursuit predators like other large theropods who probably did have columnar legs basically perpendicular to the ground. Instead, the slow movements of large therizinosaur, which probably rested a lot, wouldn’t have put much stress on the leg muscles (which are enormous, judging by the length and width of the hip) and thus the femur makes perfect sense to have been angled to the degree it is in the skeletal - ignoring the fact that therizinosaurs are much better adapted to bipedal locomotion than elephants.

Thanks to Toxic’s model, we were also able to estimate the weight of the animal. Smith et al. (2020) describes the incredibly pneumatic vertebra of therizinosaurs, which are apparently comparable to ornithomimosaurs and oviraptorosaurs in this regard. The bone density for large ground birds and other similarly pneumatic theropods is described as ~0.95 kg/dm^3 by Larramendi et al. (2020) - using this estimate, the results are as follows for the volumetric analysis:

Weight (of VERSION 1): 7763 kg (7.8 t) using 0.95 kg/dm^3 bone density

Weight (of VERSION 2):

SPECIAL THANKS TO:

Scott Hartman

Shane Wheatley

Oliver Demuth

Toxic_Midget

For your help with this project! If you have any comments or questions, please send them over below!