Wednesday, 1 May 2013

Tyrannosaur Forelimbs ......... Again

Let me set the scene. It had been a long day and as I walked through the front door of my place I thought I would head straight for the sofa and have a five minute chill before getting some dinner ready. As I lay on the sofa and relaxed I looked upward and there was something that caught my eye that made me realise something about tyrannosaurid forelimbs.
Allow me to explain. Above the sofa there are two wall mounted shelves which have a few fossils on display. These include a mosasaur dentary, a woolly rhinoceros cervical vertebra, a piece of wonderfully preserved tree bark from the Oxford Clay and one of those polished belemnites from Morocco. But central to the display is a cast of a right forelimb from Albertosaurus libratus and, as I looked up, I saw the wickedly hooked manus unguals peering over the edge of the shelf and it occurred to me that tyrannosaur forelimbs were actually very functional. In fact they were not vestigial at all.
Let me explain. I have discussed theropod forelimbs before – particularly those of tyrannosaurs – and one of the more common explanations for the reduction in the size of  tyrannosaur forelimbs was that they actually served no use anymore and that they were, indeed, becoming vestigial. There is a school of thought that had the dinosaurs escaped the end of Cretaceous extinction then tyrannosaur forelimbs would have disappeared altogether.
There is a wonderful depiction by palaeoartist Greg Paul that features in the popular Scientific American Book of Dinosaurs (2000) in which he depicts how dinosaurs may have evolved post Cretaceous and, central to this typical Paul-ian scene, are derived tyrannosaurs with only little stubs remaining where there were once forelimbs. Interesting but I suspect probably incorrect (I would share this image but suspect this may cause a problem).
From Lambe 1917.
The forelimbs of theropod dinosaurs have undergone significant evolutionary change throughout the whole of dinosaurian history. They have displayed various morphologies and evolutionary trends that, although at first they may appear to follow a convergent path ie reduced size due to bipedalism, actually demonstrate significant differences.
For instance, dromaeosaurids and other similar small theropods have well developed, highly functional forelimbs whilst tyrannosaurids, as we are all aware, appear to have highly reduced and less functional forelimbs but, as we have only recently been examining, large theropods such as the therizinosaurs and Deinocheirus, despite them being Late Cretaceous theropods that were contemporaries of tyrannosaurids, evolved spectacular and large specialist forelimbs.
So there is obviously an element of specialisation displayed here. This can evolve in a number of ways that includes loss and/or modification of digits and carpals, the various morphologies in the manus claws and the variable ways that the muscle framework attached.  As far as tyrannosaurids are concerned (and most other large theropods for that matter) the reduction in the size of the forelimbs was largely due, in part, to their large size but this is only part of the story.
The Late Jurassic Ceratosaurus displays a robust, albeit short, forearm adorned with four digits – a primitive trait retained not only by ceratosaurs and abelisaurs but even primitive tyrannosauroids. It is strange that the derived Late Cretaceous abelisaurids, such as Carnotaurus and Majungasaurus also retained the condition since, if these forelimbs were becoming vestigial, one would expect that, since there is around 85 million years of evolution between Ceratosaurus and Majungasaurus, that the forelimbs would have long disappeared – or am I being over simplistic?
We understand, especially from the well sampled Late Cretaceous faunas of North America that the duration of a species, on average, was between 0.9 to 1 million years – in some cases less. Of course, there are exceptions and, by way of example, the hadrosaur Prosaurolophus maximus appears to have survived over a period of 1.6 million years (McGarrity et al 2013) but this, for now, is the exception rather than the rule. It would seem likely that rapid species turnover and evolution would have accelerated the loss of the forelimbs if they were indeed, vestigial.
The caveat here is that, if we use the Dinosaur Park Formation as an example (Mallon et al 2012), theropods actually remain quite stable taxa-wise whilst all around them the mega herbivores are in an almost constant state of evolutionary flux. And yet time and evolution wait for no dinosaur and theropods were no exception. Environmental, ecological and biological pressure would still have driven evolutionary adaption and change.
So was theropod forelimb evolution destined for vestigiality? Hard to quantify but it would appear there is something else going on here. Perhaps one thing that does support vestigiality is the uniform reduction in digit size in these more primitive forms – that is they all appear to be shrinking at the same rate. But this is where tyrannosaurids differ.
Despite having been discussed at length by palaeontologists over the years, the function of the tyrannosaurid forelimb is still no closer to being solved although there are a few good possibilities. However, the one thing that everyone agrees on is that, despite their size, the tyrannosaurid forelimb was actually robust and quiet strong, capable of sustaining high stress loadings and (if it was feasible) able to lift a considerable weight.
Not exactly "vestigial".
So, being simplistic again, if the forelimb was becoming vestigial then why reduce the size of the forelimb only to retain significant strength and power? Of course the dichotomy here is that despite being reasonably functional the reduced size of the forelimb greatly reduces their use! This can drive you to distraction but there are factors worth considering.
First of all, and as we have already mentioned, theropods evolved smaller forelimbs because bipedalism demanded it – that is an undeniable fact. Evolution then had the careful balancing act of ensuring that the forelimb complemented the animals’ lifestyle without compromising balance and agility. This was remarkably successful, not only in terms of dinosaurian evolution, but as one of the greatest terrestrial body plan designs of all time. Theropods were amongst the first and amongst the last dinosaurs ever to exist and this basic blueprint lasted for the duration of their tenure on the planet.
Interestingly, the well-publicised paper looking at the evolution of posture, morphological change and the driving forces behind them in bird-line archosaurs (Allen et al 2013) again highlights the exceptional adaptability of the theropod body plan. One of the key findings in the paper is that as the forelimbs of these bird-line theropods became larger, their centre of gravity edged forward and the gradual evolution of the extant crouching bird position came into effect.
It was originally thought that it was the lightening and shortening of the tail that was the locomotor behind the development of the crouch but this excellent research using sophisticated three dimensional modelling techniques clearly demonstrates that the forelimbs were the major contributing factor.
It came as a surprise to find out just how much influence the forelimbs had on balance and posture especially when compared to the tail but the trend was clear. This new research also has implications for tyrannosaurids in as much as it confirms what we already knew – that is that tyrannosaurs, despite being closely related to the maniraptoran line, retained the large powerful tail to counterbalance the enlarged head and that the reduced forelimbs were clearly not as important to them as it was to those theropods that were heading down the avian road. Obvious I know but nice to have it confirmed by the science.
And yet, just to ensure that there are still more questions to answer, John Hutchinson highlights the fact that since the forelimbs were influential in the development of body mass distribution, then they equally affect how the hind limbs developed and functioned. In other words we have to assess the whole locomotor structure rather than just part. This is obvious in so many ways and yet it is something that we all miss when discussing these kinds of topics. One simple evolutionary trend affects the whole of the animal – everything is affected, everything changes – even if we have no fossil evidence. It is not always easy to remember this but it is always true.
Is it possible then that the tyrannosaurid forelimb affected the balance of the animal? Could they have conceivably had any influence on the agility and stability of a tyrannosaur in motion? It hardly seems possible that these small forelimbs could have any realistic effect and yet I am reminded by the halteres in two winged flying insects. These tiny club-shaped structures are situated behind the wings and enable the insect to maintain a high degree of equilibrium. Insects are not dinosaurs and yet perhaps we should not completely dismiss the possibility of tyrannosaurid forelimbs acting as a balancing aid. 
So if some theropods were clearly evolving their forelimbs for, grasping, flapping, flying and, in the case of Deinocheirus and therizinosaurs, for some as yet undetermined functional use, then tyrannosaurid forelimbs  do appear underwhelming  and yet we have other odd closely related theropods with strange arms – the ornithomimids.
These appear, on the face of it, to be long and highly adaptable – especially for grasping and yet these forelimbs are surprisingly weak when compared to tyrannosaurs. They demonstrate a surprising lack of manoeuvrability due to an under developed antebrachium (that is the area between the elbow of the wrist) and this limited their ability to pronate or supinate. So, despite their impressive appearance, it appears the arms of ornithomimids were best suited for simple reach-and-grasp use.
The thing that should grab your attention with tyrannosaurid forelimbs is not their relative small size but actually that they are quite robust and that the unguals (claws) are potent barbs although in more derived tyrannosaurines such as Tyrannosaurus they do become blunter. Despite the fact that there only the two digits represented, there is a remnant third metacarpal present and indeed does become more vestigial as tyrannosaurids became more derived - so much so that in taxa such as Tarbosaurus it is nearly completely fused with metacarpal 2.
And this is where we came in. Ceratosauria appears to demonstrate a uniform reduction in digit size throughout their long history although I confess to being mystified as to why it would take so long to lose their forelimbs IF they were becoming vestigial. Tyrannosaurids on the other hand clearly lost their digits in sequence – so much so that we find two digits appearing to gradually disappear throughout the tenure of Tyrannosauroidea although there are massive gaps in the fossil record so we must be cautious. The gradual fusion of metacarpal 3 onto metacarpal 2 in derived tyrannosaurids clearly demonstrates this process as digit 3 disappeared.
Metacarpal 3 (bottom) slowly disappearing in Albertosaurus
But the remaining two digits are clearly of importance to tyrannosaurids and, as always, it is difficult to ascertain what and why they were used. In my previous discussion about tyrannosaur forelimbs I speculated that if tyrannosaurids were feathered to some degree then perhaps they were used in courtship rituals – a kind of “don’t eat me – I’m here to mate” signal. You would think that tyrannosaurids must have had some form of intraspecific communication – even if it was just for the mating act.
Indeed, it was gratifying to hear Thomas Holtz Jr during his recent talk at the Burke Museum in Seattle a short while ago suggest exactly this and the more you study these structures the more intriguing they become. However I have been leaning toward a similar but subtly different use for them.
Pathologies in the forelimbs we have are not uncommon and they demonstrate various fractures, remodelling and scarring right through the entire forelimb right up to, and including, the furcula. There are various reasons for these pathologies and they could be as simple as the animal falling down and breaking an arm. Tyrannosaurids are huge animals and a fall onto a forelimb could have caused considerable damage and pain.
However, it occurred to some of us that perhaps the forelimbs were used to assist in copulation and gave the males some purchase during the act – as well as helping to fend off an over aggressive female. This made some sense since trying to hang onto a large multi-tonned female during copulation would have caused considerable stress to the forelimbs and injury would appear inevitable.
Gorgosaurus sp. (BHI#126846) displaying
 acute pathology to the scapulacoracoid
However, as pointed out by others (eg Rothschild & Molnar 2008), this would seem unlikely since the T. rex Sue (FMNH PR2081) shows considerable pathologies to the humerus and scapulacoracoid and – of course – this specimen is widely accepted to be a female. It is still worth pointing out that sexual dimorphism in tyrannosaurids is unproven although it does appear that there are specimens that display both gracile and robust features so dimorphism seems likely.
But I do not believe that this should disqualify the mating hypothesis since tyrannosaurids led such a hazardous life and multiple pathologies are found throughout their entire skeleton. It is entirely conceivable that the forelimb was used in the act of mating and, at the same time, was equally susceptible to injuries that occurred during hunting, falling and the other day to day hazards that befell any carnivorous dinosaur during life.
Of course (and how many times have I said this?) this is all behavioural inference and likely to be unproven and yet the more you look at these enigmatic structures the more you realise that there is so much more to them than meets the eye. For me, I just cannot accept that what is actually a robust and derived specialist bit of equipment was actually vestigial. But, as always, I remain open minded and look forward to the next hypothesis regarding these extraordinary forelimbs.
And a word of thanks
The viewing figures for this blog recently crashed through the 100,000 mark and I feel really humbled by that. Bloggers, such as myself, who want to make a contribution to our science will never have the aura and pulling power of the uber bloggers such as Dave Hone, Brian Switek and Darren Naish but we all contribute in our own way and long may that continue.
So a very big thank you to all of you who have taken the time to read these pages - it really does make all the difference.

Allen, V., Bates, K.T., Li, Z., & Hutchinson, J.R. (2013) Linking the evolution of body shape and locomotor biomechanics in bird-line archosaurs Nature DOI: 10.1038/nature12059
Lambe, L. 1917. The Cretaceous theropodus dinosaur Gorgosaurus, Memoirs of the Geological Survey of Canada. 100, pp. 1–84.
Lipkin, C., & Carpenter, K. (2008) Looking again at the forelimb of Tyrannosaurus rex. In: Tyrannosaurus rex, the tyrant king. Larson, P. & Carpenter, K. (eds.). Indiana University Press. Bloomington & Indianapolis, Larson, P. & Carpenter, K. pp.167-190.
Mallon, Jordan C., Evans, David C., Ryan, Michael J., Anderson, & Jason S. (2012). Megaherbivorous dinosaur turnover in the Dinosaur Park Formation (upper Campanian) of Alberta, Canada Palaeogeography, Palaeoclimatology, Palaeoecology Volumes 350-352 pp. 124 – 138.
McGarrity, C.T, Campione, N.E. & Evans, D.C. (2013) Cranial anatomy and variation in Prosaurolophus maximus (Dinosauria: Hadrosauridae).Zoological Journal of the Linnean Society 167(4): 531–568
Paul, G.S. (Editor) 2000. The Scientific American Book of Dinosaurs. St. Martins Press. pp. 384-385. ISBN-10: 0312262264
Rothschild, B. M. & Molnar, R. E. 2008. Tyrannosaurid pathologies as clues to nature and nurture in the Cretaceous. In: Tyrannosaurus rex, the tyrant king. Larson, P. & Carpenter, K. (eds.). Indiana University Press. Bloomington & Indianapolis, Larson, P. & Carpenter, K. pp.287-304.

Obviously not a true comparison but interesting none the less.


John Conway said...

Interesting post. I've come to the opinion that the forelimbs were either vestigial. or signalling devices. I can't imagine any mechanical use that a six-tonne Tyrannosaurus could put them to that wouldn't rip them of at the slightest shift in weight. I'd like to see the numbers, but I strongly suspect they couldn't take anything like the stress of a shifting tyrannosaur body if they were lodged in something.

As for vestigially, we don't know what sort of genetic locks these theropods were subject to. They probably had to keep relatively robust shoulder girdles for neck musculature - perhaps they were genetically stuck with arms. This is boring from a morphology perspective, and is also difficult to falsify.

Mark Wildman said...

Thanks for the comments John. I have always found the shoulder girdle of intense interest as well since the girdle is independent of the rest of skeletal structure. The muscles and tendons that held it in place in these large predators must have been immense.

Hadiaz said...

John Conway: "I can't imagine any mechanical use that a six-tonne Tyrannosaurus could put them to that wouldn't rip them of at the slightest shift in weight."

Bakker brought back Osborn's tickling idea (See "Theory No. 3": ). Reminds me of how boids use their pelvic spurs.

"It is still worth pointing out that sexual dimorphism in tyrannosaurids is unproven"

I wouldn't say that, given the discovery of medullary tissue in a robust specimen (See the Galton quote). Otherwise, good post (as usual).

BTW, congrats on breaking the 100,000 ceiling.

Quoting Galton ( ): "In the theropod dinosaur Tyrannosaurus rex, there is also a femoral sexual dimorphism (Larson, 2008), with the robust morph identified as female based on pelvic dimensions and fused caudal vertebrae. Analyses for the presence of medullary bone in this same series of femora showed a significant clustering with the robust forms (Schweitzer et al., 2005). This indicates that the robust morph was probably female because female birds form medullary bone as a calcium reservoir to aid in egg shell formation (Dacke et al., 1993)."

Mark Wildman said...

Hi Herman and thanks for the kind words. I do acept that sexual dimorphism in Tyrannosaurus is likely but Holtz recently stated:

"Here's the important issue: before you can assert sexual dimorphism, you
have to demonstrate di-morphism. That is, one has to show that the
population shows two discrete, statistically-distinguishable
subpopulations. This has not yet been done for Tyrannosaurus. Instead, it
has been seen is that there are some specimens which are more robust and
some which are more gracile, but it is not at all clear that they form
something other than a continuum of robusticity.

That said, in one of the few cases where we can reasonable assert the sex
of a dinosaur specimen, the B-rex individual with medullary bone (and
hence mostly likely an ovulating female) is on the robust end of the

Anonymous said...

Pretty good article, Wildman; a good addendum to your Sep/2011 work.

Metacarpal III is an interesting bone in tyrannosauridae; it is also possible that it actually was a useful structure for attaching tendons, not necessary doomed to disappear.[1]

It is true mIII is smaller in at least some Tarbosaurus specimens--but this might be a matter of forelimb variability[2] (also remember the SVP 2012 work by Sara Burch).

Forelimb robustness is also variable; Russell pointed out that albertosaurines seem to have "lightly constructed" forelimbs, while those of tyrannosaurines are more robust.[3]

1. See for example Carpenter K, "RE: T-rex finger count?", Dinosaur Mailing List 6/6 2005.
2. Currie P.J.2003. "Allometric growth in tyrannosaurids (Dinosauria: Theropoda) from the Upper Cretaceous of North America and Asia", Canadian Journal of Earth Sciences 40(4):p.651-665.
3. Russell D.A.1970. "Tyrannosaurs from the Late Cretaceous of western Canada", National Museum of Natural Science Publications in Palaeontology 1:p.1-34.

Mark Wildman said...

Thanks for the comment. Whilst not doubting the variable robustness of tyrannosaurid forelimbs, I was simply alluding to the point that the two remaining digits were still quite formidable as digit 3 virtually disappeared.They were clearly functional but to what degree is uncertain - not vestigial. And you can call me Mark.......

Anonymous said...

Thank you Mark for these admissions (and clarifications).

By the way, are your e-mail address availabile on the blog?

Mark Wildman said...

I avoid being over public with my main email address (for obvious reasons) but I am on Twitter if you want to connect. I'm @saurian65mya. Thanks for your interest.

Anonymous said...

Thank you Mark. Let see if we discuss on Twitter.

The condition of metacarpal III (and some other forelimb elements) in Tarbosaurus seems more like an exception of the general rule among tyrannosauridae (see Currie (2003) and "Tyrannosauroidea" in The Dinosauria (2nd edition). By the way, The attachment of mIII to mII in your albertosaur arm cast might be a flaw in the manufacturing (the other bones are also "fused")?

Russell (1970) argued that tyrannosaurine manual claws might was less curved than albertosaurines. But more blunt? What is your references to this notion?

I think arm condition might partially be a matter of balancing; allosauroids had "large" arms--and lightweight skull/long tail--tyrannosaurids had heavy skull/short tail.

Why not discuss all of this with your colleague Michael Habib? I have heard that he have some interesting points on the subject of forelimbs in tyrannosaurs and other theropods. Have you read the recent work by Williams[1]

1. Williams S "A new subadult tyrannosaurus rex and a reassessment of ontogenetic and phylogenetic changes in tyrannosauroid forelimb proportions", Geological Society of America Abstracts with Programs 43(1):p.120.

Mark Wildman said...

The fusing of metacarpal 3 to metacarpal 2 is likely a preservational artefact but this should not detract from the fact that it was disappearing. And the cast was certainly created with appearance in mind so you are likely correct.

Holtz stated in The Dinosauria that the manus unguals are thicker at the distal end and not so sharp in Tyrannosaurus when compared to those of Gorgosaurus and Albertosaurus. I was inferring that all tyrannosaurines in general likely followed suit.

Anonymous said...

Thank you Mark.

I do not know if we can know for a fact if mIII was disappearing completely (though one might can infer this from Tarbosaurus). I still consider it might had a good function as muscle attachment, at least i T. rex...

Holtz (2004) indeed give many interesting notes about tyrannosaurid forelimbs, but he do not address which specimens he examined--sadly.

With ceratosaurs, you wonder "why it would take so long to lose their forelimbs IF they were becoming vestigial?". They might need more degenerating mutations focusing on the arms? This becmes even more interesting when we consider Eoabelisaurus.

Keep up the good work! :-D

Anonymous said...

Oh, sorry Wildman;

I forgot to admit that hypotheses I favor cannot be known for a fact either...

Mark Wildman said...


Anonymous said...

Sorry Mark (I use to call people with higher qualifications by their surname...). :-(

Regarding metacarpals, I came to remember that Holtz (2004) also says that T. rex and Tarbosaurus have a mII/mI-ratio of 170% or less (other tyrannosaurids have 200%)--and suggested they might be more derived in theropod digit reduction. This might be a problem for those of us who support functional theories?

But I have seen some pictures of the Tarbosaurus skeleton MPC-D 107/2 at Flickr; it seem to have a ratio of 200% (the same with a Tarbosaur arm picture at

Anonymous said...

Alternatively, these differences in metacarpal lenght might be a result of "good old fashioned individual" variation (as Holtz once said).

Even so, this variation in relative forelimb bone length and robustness can be a problem for us who argue for functionality; Michael Habib says (personal communication, 2013) that forelimb variability is also prominent in ratite birds--as their wings are not used for much, there is a low degree of selective pressure, he believe--which might could be applied to tyrannosaurids also.

What do you think, Mark? :-)

Mark Wildman said...

There is no doubt that individual variation is a factor which is why there is never really a maximum size for any animal whether extant or extinct.
Mike have a point but it is still highly subjective. For example, the furcula in ratites is almost absent but not so in tyrannosaurids and, apparently, the furcula is more functional in tyrannosaurids than in dromaeosaurids which is kind of weird. I will dig out the reference for that statement.

Anonymous said...

Interesting point about the tyrannosaurid furcula. But it might had another main function than assisting forelimb use (I just want to speculate)? What about assisting breathing?

As comparison, Holtz suggest[1] the massive shoulders in abelisaurs mainly served as attachment for big muscles in the neck, rather than the arms. But this idea might not fit tyrannosaurids and their thin "strap" scapulae... but they also had quite short neck, unlike abelisaurs.

1. Holtz T jr, 15/8 2012 comment to Switek B, "Carnotautus Had a Hefty Neck", Dinosaur Tracking 15/8 2012.

Mark Wildman said...

The point I was making about the furcula was to reinforce the subjectiveness of comparing ratites with tyrannosaurs. It isn't that any one person is right or wrong - just that we all need to be objective about things. The furcula as a breathing aid? Cannot see that myself but then we wil have to wait and see as more evidence becomes available.

I believe you have actually provided the answer when you made your comment about abelisaurs and tyrannosaurids.

Anonymous said...

Yes Mark--and it was an interesting point. Neither do I think anyone of us have the hole truth. But Lipkin & Carpenter (2008) believed the scapulae of tyrannosauridae attached certain arm muscles.

Regarding proposed use of furcula in breathing; remember I am a layman--I do not know enough about respiratory systems in archosaurs... But I thought of the hollow furcula in Aerosteon (a cavity for air sacs?) and an article by Andrea Cau at his Theropoda blogspot (6/10 2008) which proposed that some theropods (especially abelisaurids) had reduced arms in favor of advanced respiratory systems.

Mark Wildman said...

It is hard to comprehend that the furcula would contain air sacs - it is such a small bone in relative terms. But as part of some sort of framework to aid breating in theropods? Could be - perhaps this an area of research that needs expanding.

Anonymous said...

You might be right, Mark; in second thought, it feels unlikely that the furcula assisisted respiration (if tyrannosaurids had air sacs like those of birds, we might should look at them?).

Still, small structures/features can serve interesting functions; for instance, that metacarpal III of T. rex might attached strong tendons in the wrist,[1] or that the great range of retraction in the uppper arm/shoulder-joint[2] might enabled the arms to be "folded back" to avoid injury during intermutual fights (this is just a personal idea--and it still seems like the arms could receive nasty bites).[3]

1. Quinlan E.D., “Anatomy and function of digit III of the Tyrannosaurus rex manus”, Geological Society of America Abstracts with Programs 39(6):p.77.
2. Senckenbergiana lethaea, 2002, vol.82(1):p.59-75.
3. Ep. 4 of "Dinosaur Revolution" (Discovery Channel, 2011) showed a tyrannosaur humerus with pathologies indicating it almost had been bitten of.

Anonymous said...

Hi again Mark!

Returning to implications of forelimb size/proportional variability; what about ornithomimosaurs? You argued they did not have much use of their forelimbs--which are highly variable in size and proportions, both in upper and lower arm as well as the hands.[1] Also, what is your reference to that metacarpal II and III in Tarbosaurus are fused? Personal observation? Pictures/drawings of fossils/casts?

Also interesting that tyrannosaurid forelimbs (based on the shoulder joint) had limited range of motion--while those of in abelisaurids allowed great range of motion[2]--yet both groups show signs of a forelimb nervous system making their arms limited in usefulness.[3] Why not discuss this sometime?

1. The Dinosauria, 2nd edition (2004), chap. 6:“Ornithomimosauria”.
2. PaleoBios 26(1) (2006).
3. Journal of Zoology 236(3): p.389-410 (1995).

Anonymous said...

Hi again Mark.
I have been thinking about your claim that Metacarpal II and III are fused in Tarbosaurus--do you base it on a text you have read, or have you seen some skeletal cast?

Post a Comment