Friday 16 June 2017

Revenge of the scaly Tyrannosaurus

Reworked version of my 2012 Tyrannosaurus painting, now in it's third guise. There's something about this painting which recalls reconstructions from 1906 rather than those of 2016.
The skeletal anatomy of Tyrannosaurus rex is probably better known and studied than the skeletons of many living animals, but its soft-tissues - and thus much about its life appearance - are poorly represented by fossil remains. Thus, virtually all of our ideas about muscle bulk, soft-tissue body shape and integument have to be reconstructed by phylogenetic proxy and functional prediction. As with all dinosaurs, we've historically felt pretty confident that Tyrannosaurus was entirely scaly, but relatively recent discoveries of filamented tyrannosauroids in China (Xu et al. 2004, 2012), as well as a growing mountain of fuzzy coelurosaur fossils, point to a different conclusion: that Tyrannosaurus was adorned in simple filaments - hair-like equivalents of feathers. Skin impressions for more derived tyrant species - the tyrannosaurids - have proven rare in fossil record (Hone 2016) and, though rumours have circulated about some, they have largely escaped formal description and publication. In the absence of better evidence, the most parsimonious modern takes on everyone's favourite tyrant have involved a fuzzy covering.

In the recent months two papers have challenged this idea. The first, by Thomas Carr and colleagues (2017), purports to find osteological correlates of scales on the facial anatomy of the tyrannosaurid Daspleteosaurus, which they argue (along with other lines of evidence), to suggest crocodylian-like facial tissues and sensitivity. The second, by Phil Bell et al. (2017), describes scaly skin impressions from multiple postcranial regions of a Tyrannosaurus skeleton, and argues that the distribution of these impressions implies a uniform (or near uniform) covering of scales across the body, without much in the way of fuzz.

Because this is Tyrannosaurus, media sites and bloggers have spilled great amounts of ink over these stories. The scientific press has often been far from objective or unbiased. Popular articles have suggested Jurassic World fans might have 'won' the debate over scientists, that science fans are 'due' a return to scaly tyrants after 'losing' Pluto, and that the findings mean 'all is well in the dinosaur world'. The implication is a ridiculous one, like evidence of scalier tyrants is a moral victory rather than a test of a scientific hypothesis. But while the popular press has been celebrating the new papers, members of the palaeoblogosphere have been less enamoured with the findings. Trey the Explainer suggests that Bell et al.'s work doesn't really change what we already knew about tyrant integument, and thus does not invalidate many existing reconstructions. Andrea Cau posits that interpretations of scaly tyrants reflect our prejudices more than science, and that taphonomic factors may explain the absence of filaments. Brian Switek has concerns that the skin patches are too small and spread too widely to give a complete picture of the integument, and echoes concerns about taphonomic interference. The collective response seems to be a defensive one, protecting concepts of filamented tyrannosaurids from a resurgence of a more traditional, scaly model. Would any other dinosaur get this treatment? Perhaps not: as Brian explains in his recent post, this reaction is the T. rex celebrity effect at full bore.

Supermegafluffy Tyrannosaurus, from 2015. They were simpler times.
I've painted many fluffy Tyrannosaurus in the last few years (above) and quite like the idea of everyone's favourite 6 tonne dinosaur bonecrusher being a giant plush toy. However, we also have to concede that our ideas of Tyrannosaurus skin have been largely informed by prediction, not direct data, and that popular, long-held notions are as ripe for scientific revision as any other (lest we forget other famous examples of this - Brontosaurus and Ornithoscelida). Moreover, although some critics are suggesting the papers don't tell us anything new - rumours of scale impressions have been circulating for years - these recent studies give us the first rigorously documented, peer-reviewed glimpse into Tyrannosaurus skin anatomy. This is new, allowing us to form our own opinions on Tyrannosaurus appearance based on actual data, not hearsay. So, rather than putting our gloves up to defend our prior model, I wonder if we should be exploring how this new data might transform our perception of Tyrannosaurus life appearance. That these new studies present conflicting data to our expectations is not grounds to be upset, annoyed or defensive. To the contrary, they allow us to use real data - not predictions - to refine our ideas of tyrannosaurid appearance and evolution. For those of us interested in dinosaurs as real entities, and not movie monsters, that's a good thing.

What, exactly, has been argued about scaly tyrants?

A lot of the popular write ups of these recent papers include errors and misrepresentation, so let's recap what is actually being argued about Tyrannosaurus skin. A common social media reaction to Bell et al.'s work is that they've presented 'a patch' of skin, and are extrapolating from that. We need to debunk that right away: they've not described a single patch, but multiple small patches from the neck (alas, exactly where on the neck isn't reported), the top of the pelvis, and the base of the tail (below). All the samples stem from the 'Wyrex' specimen (HMNS 2006.1743.01). The most extensively represented area is the tail base, which has the largest single piece of fossil skin - 30 cm². The other skin samples are not as large, some being just a few centimetres across. Each patch shows the same skin type: uniform, tiny 'basement scales', each less than 1 mm across (Take note, artists: you would not see Tyrannosaurus scales until you were being eaten by their owner). Similar scale patches, also described by Bell et al. (2017), have been found on the torso and tail regions of other tyrannosaurid species, implying similarly scaled regions in these taxa.

Tyrannosaurus skin patches from the neck, pelvic area and tail of the 'Wyrex' specimen as illustrated by Bell et al. (2017). The scale bars for the scale imagery are 5 mm (b - e) and 10 mm (f-h). These things are tiny, and we can assume the skin of the animal would look smooth or leathery in life.
Some folks are suggesting that the size of these skin patches allows us to dismiss their scaly signal, or that even that they're anomalous, reflecting unusual taphonomic conditions that cloud their significance. I'm unsure about these ideas. Most skin impressions are small patches (even scaly skin gets a rough ride during fossilisation) and the fact they're small doesn't diminish the fact that each records a cluster of scales. We have to assume these are not unusual or 'special' areas on the body but generally indicative of surrounding skin fabrics. The fact that each patch is consistent with regard to scale size and texture hints at them being part of a continuous, unbroken integument, and not isolated scaly pockets in a sea of fluff.

But what about arguments that the scale patches are tissues stripped of filaments before preservation, like so many 'monster' carcasses? Filament/scale combos do have precedent in dinosaurs, being present on the tail of Juravenator and those scales of Kulindadromeus with fibre-like tassels (Chiappe and Göhlich 2010; Godefroit et al. 2014). We know from modern animals that fibrous epidermal structures are especially vulnerable to decay and physical weathering, but is there evidence that this has taken place on the Wyrex Tyannosaurus skin patches? At present, it's hard to say because we have no idea what tyrannosaur skin looks like as it decays. It might be significant, however, that the scale patches look very similar across the Wyrex specimen, and that they resemble other tyrannosaurid skin impressions closely. We might expect some variation if taphonomy was really distorting these specimens in a major way, and we're not seeing that. Moreover, the Wyrex skin impressions, though small, are pretty high-resolution. The scales, and their intervening areas, have sub-millimetre proportions and sharply defined edges. There's no tatty scale margins, no obvious spaces for filament attachment, or linear structures crossing the scales to imply a rogue filament impression. We'll remain uncertain if these are anomalous, taphonomically-altered samples until we find other examples of tyrannosaurid skin, but there's no reason to be unduly suspicious of the the samples we have.

Of course, the adage that 'absence of evidence is not evidence of absence' is always important when dealing with the fossil record, and it applies here as a sensible caveat. However, we shouldn't wield this phrase as a definitive counter-argument to reasonable interpretations of available evidence. Palaeontologists have to work with data, not suspicions or gut feelings, and the data we have does not include, or hint at, the presence of filaments. I'm not arguing that taphonomy isn't worthy of consideration here (indeed, the omission of details about 'Wyrex' taphonomic history is an issue with the Bell et al. 2017 paper) but we must beware the logical fallacies of appealing to probability (i.e. taphonomy could explain the lack of filaments, so it does explain the lack of filaments) or special pleading (excluding Tyrannosaurus from the same logic we would apply to other fossil animals when presented with this data).

Tyrannosaurus skull AMNH 5027 - note the 'hummocky' textures on the side of the snout, above and below the orbit, and atop the rostrum, likely indications of scaly skin. Image in public domain, sourced from Wikipedia.
Carr et al. (2017) present a different form of evidence for scales: osteological correlates. I consider some aspects of their study problematic in that it only looks to crocodylians and birds for comparative tissues, despite the clear value other tetrapods have in deducing facial tissue types (Knoll 2008; Morhardt 2009; Hieronymus et al. 2010); it lacks illustrations of the bone textures correlated to scaly integuments; and the conclusion of tyrants bearing crocodile-like face scales is flawed: crocodylians do not have face scales, but a tight, highly cracked sheet of facial skin - Milinkovitch et al. (2013). Nonetheless, I think Carr et al. (2017) are right in concluding the bony textures of tyrannosaur skulls seem indicative of scaly skin. These findings echo previous interpretations of bosses and rugosities in tyrant skulls (e.g. Brusatte et al. 2012; Sullivan and Xu 2016) and aren't controversial. Scales closely associated with bone either leave a 'hummocky' surface texture, which is seen on tyrant snouts (specifically their maxillae and nasals) or small bosses and hornlets, which are found in all tyrannosaurid skulls above their orbits (lacrimal and postorbital bones) and on their 'cheeks' (jugal bones). Hornlets and bosses represent the locations of specific scales in living reptiles (Hieronymus et al. 2009) and can thus give especially good indications of life appearance (check out chameleon skulls for especially good correlation between skull and scale features). The presence of hummocky bone textures and hornlets is a strong correlate for scales, as they rule out coverings of naked or feathered skin. Such skin types do not alter the underlying bone surface (Hieronymus et al. 2009).

These osteological correlates combine with the skin impressions to collectively show Tyrannosaurus as scaly across much of its face, somewhere on its neck, over the pelvic region and along the tail base (below). So far as we can tell, this picture seems consistent with osteological correlates and skin sampling from wider Tyrannosauridae. That's pretty extensive coverage, ruling out the presence of fibres in places that we know other dinosaurs - including other tyrannosauroids - were fuzzy, and implies that tyrannosaurids were mostly scaly. I'm particularly startled at the scales over the hip region as they curb even the long 'fibre capes' we see in some modern tyrant reconstructions, like the famous Saurian Tyrannosaurus. The fact that the scales occur in places known to be ancestrally filamented for tyrants is also intriguing: Bell et al. (2017) speculate that they may be modified feathers - that is, the same as bird scales - rather than a reversion to lizard or croc scales. Hold that thought, we'll come back to it soon.

Everyone's doing maps of Tyrannosaurus with integument details nowadays, and I want in. Note that this is Tyrannosaurus specific, and does not feature scale data from other tryannosaurids.

What's in the gaps?

The million dollar question is what was present between these scaly regions: more scales, or fibres? This is a major point for many respondents to the Carr et al. and Bell et al. papers, as it decides whether we keep our interpretation of Tyrannosaurus as an - at least partly - fuzzy animal. With our scale distribution map as a starting point, several options are available. The first is that fuzz was present in regions not yet represented by skin remains or osteological correlates. This would mostly imply the top of the torso (Bell et al. 2017), but may also be parts of the back of the head, some aspects of the neck (depending on where the neck skin impression came from) and maybe the end of the tail. Over on Twitter, Patrick Murphy has presented a reconstruction which shows what this might look like. I must admit to finding it quite amusing, sort of like T. rex has put on a shawl to visit the opera.

But how dense could these fuzzy patches have been? Bell et al. (2017) suggest that dense fibrous coverings are doubtful, noting that large living mammals avoid patches of thick insulating fibres to aid heat loss. This has not gone down well with some critics, who cite studies of feathers preventing over-heating instead of facilitating it. An oft-cited study in this regard is Dawson and Maloney (2004), who found emu feathers block virtually all solar radiation from the skin, preventing them from overheating in solar exposure that causes similarly-sized hairy mammals to seek shelter.

Feathers: great at blocking solar radiation, also great at trapping body heat. Note how cooking hot these ostriches are on their necks, heads and legs, while the feathers are mostly ambient temperature. This isn't because the body isn't warm, but because the feathers block the heat signature entirely, trapping all that heat around the body. As surface area:volume ratios drop as animals get larger, it stands to reason that the benefits of blocking solar radiation give way to a need shed heat. Image from Wikipedia user Arno / Coen, CC BY-SA 3.0.
Feathers, however, are not magic structures that defy fundamental physical laws of insulation, nor do they liberate animals from the challenges of heat loss at reducing surface area:volume ratios. Beyond a certain size, shedding excess body heat is difficult for any terrestrial animal, and it gets tougher as they get larger. King and Farner (1961, p. 249) described feathers as having "an extremely high insulating value to the feathered surfaces" and a rich literature of studies on modern birds shows that feathers are as effective at trapping body heat as they are blocking solar rays (e.g. King and Farner 1961; Kahl 1963; Philips and Sandborn 1994; Dove et al. 2007). We can almost see them as a little too effective, leading many birds to develop heat-dumping adaptations to circumvent their own insulation, such as highly vascularised, non-feathery body parts as well as a repertoire of postures and behaviours (maximising exposure of unfeathered body parts; flapping wings; urinating on their legs) that aid cooling (e.g. Kahl 1963; Arad et al. 1989; Philips and Sandborn 1994). So yes, feathers are terrific at protecting birds from environmental heat, but that limits their ability to release metabolic heat from their own bodies.

If living birds find feathers a little warm, despite their relatively high surface area to volume ratios, we have to assume a theropod weighing anywhere between 6-14 tonnes is going to find big areas of dense filaments a challenge to thermoregulation too. It is not unreasonable to assume blankets of fibres could be a problem for big tyrants. The counterargument here is that Yutyrannus huali, a largish tyrannosauroid, does have dense fibres everywhere. But Yutyrannnus seems more lithe than Tyrannosaurus - perhaps just 10-25% of its mass, depending on the estimates (Bell et al. 2017) - and lived in a more vegetated, and thus shadier, habitat (Bell et al. 2017). A neat comparison Bell et al. (2017) make along this line uses living rhinos, where hairier species live in shadier settings than the virtually naked ones. In light of this, the reduction of filamented regions, and perhaps lessening their density, is a reasonable inference for animals of the size and habitat of Tyrannosaurus, and would reflect thermoregulatory responses to scaling and shade availability seen in living animals.

Large tyrannosauroids, like Yutyrannus huali, show that dinosaurs weighing perhaps 1.5 tonnes could be covered in feathers. But does this reflect the fact that this animal lived in shadier, vegetated habitats than the tyrannosaurids? This idea isn't silly: adaptation to specific circumstances has a major role to play in shaping animal skin anatomy, and could well explain why some tyrants are fuzzy, and others seem less so. (If you want to see the rest of this picture, check out this Patreon post)
Could Tyrannosaurus have had extremely fine, widely-distributed filaments - perhaps similar to something like elephant hair? This isn't entirely falsified by the new data, although the skin impressions we have show no evidence of such a covering despite preserving tiny integument details. Granted, animal filaments can be extremely fine, and they might be beyond the preservation potential and mechanics of even these high-res impressions. However, if we're arguing for filaments of this size and patchiness then - certainly for artistic purposes - we should concede that the animal would be essentially scaly, in the same way that most rhinos, elephants and hippos are essentially naked (below). From a thermoregulatory perspective, short, sparse filaments could make sense as these have the surprising ability to draw heat from the body in modern elephants, helping them stay cool (Myhrvold et al. 2012). Given the potential for overheating under dense filament coats in giant animals (Bell et al. 2017), I see this as more plausible than a 'cloak' of fibres between our scaly waypoints.

Scaly, minimally-filamented Tyrannosaurus. There's some tufts on the neck, but that's it. Is this model more consistent with the thermoregulatory requirements of a 6-14 tonne animal?
A last interpretation of this new data is that Tyrannosaurus was actually just scaly, with no fibres whatsoever. This is the most contested suggestion made by Bell et al. (2017), but it's not unreasonable with our current knowledge. Existing skin data, representing seven parts of the body if you pool all the distinct skull correlates and postcranial points (add several more if you want to extrapolate scale patches from other tyrants), shows enough scales and consistency in the scalation pattern that uniform scale coverage is not a ridiculous or indefensible concept. I appreciate that some folks will point to regional fuzziness of animals like Kulindadromeus in response, and its sharply defined areas of different integument types, and that's valid point. But we can also point to plenty of dinosaurs with extensive or entirely scaly hides and - if there's any value to linking body size and thermoregulatory regimes - they're a better match to Tyrannosaurus body mass than any known fuzzy species. For the time being, wholly scale models fit our existing data just as reasonably as partly fuzzy ones so, archaic and counter-intuitive as it seems - a scaly Tyrannosaurus is not an unreasonable interpretation for the life appearance of this animal, given our current data.

Beyond Tyrannosaurus: 'unlocking' dinosaur skin constraints

My take-home from these new papers is that our models of Tyrannosaurus skin have not crystallised, but we're a little more constrained in how we can imagine this animal, and have to concede a scalier appearance than many of us thought likely. But the implications of the Bell et al. study go beyond Tyrannosaurus in implying new ways to think about dinosaur skin evolution. With incontrovertibly fuzzy animals lining much of the the tyrannosauroid tree and its root, our scalier Tyrannosaurus gives us one of the best examples of a dinosaur replacing fuzz with scales. This is a far-reaching conclusion for those of us interested in dinosaur life appearance, complicating the already confusing evolutionary pattern of scale and fuzz distribution within the group. Ideas that some dinosaurs could be 'secondarily scaled' are supported by this discovery, and we have to wonder if classically fuzzy lineages - including many other theropod lines - are as tightly locked into fuzz, fibres and feathers as we once thought. Could large dromaeosaurs be a little lighter on fuzz than we imagine? Did Therizinosaurus look less like a giant pigeon and more like a walking Christmas dinner? We don't know, but now have reason to wonder.

Fluffy Tyrannosaurus juveniles, one of the possibilities created by the idea that tyrannosaurs might have avian-like 'dynamic' skin. The recovery of scales in non-scaly clades is not as simple as it might first appear!
Furthermore, the notion that Tyrannosaurus scales could be modified feathers (Bell et al. 2017) opens possibilities about mixes of filaments and scales. It's important to realise that not all scales are alike: 'reptile'' scales' are developmentally and genetically distinct from those we see in birds, which are actually secondarily modified feathers (Chang et al. 2000; Dhouailly 2009). Reptilian skin cannot be forced to grow feathers or filaments (Chang et al. 2000) and is developmentally static: once scales are formed, they're with them for life. Bird skin, however, is far more dynamic, and allows for all manner of ontogenetic and even seasonal variation in scale:feather ratios, changes to feather types, and modification of scale size (Lennerstedt 1975; Stettenheim 2000). If, as suspected, our tyrannosaurid skin samples represent fibrous integument masquerading as a scaly one, is this a sign of a bird-like 'unlocked' skin configuration where epidermal dynamism was possible? If so, Tyannosaurus could have changed appearance considerably with age (fluffy when small, scaly when big - above) or season (reflecting changes in climate or behaviour)? It must be stressed that we don't have any direct insight into these sorts of changes at the moment, and the hypothesis of tyrannosaurid scales being modified feathers needs testing. But the irony - we might have data indicating Tyrannosaurus could change its appearance readily, vindicating debaters on both sides of the scaly and fuzzy debate - is not lost on me. Maybe, just this once, everyone wins?

Summing up time

Let's tie this all together. A lot of ambiguity remains about the skin of Tyrannosaurus and its relatives, and it's not wise to hold any opinion about their life appearance too strongly at present. However, unduly downplaying the creep of scaly evidence into the tyrannosaurid fossil record isn't useful or logical. The skull skin correlates and fossil skin patches show that scales were present in numerous, widely-distributed parts of the body, and - until we see evidence to the contrary - this is good reason to assume scalier Tyrannosaurus than we might be used to. And yes, this does mean that some of our favourite, fluffier interpretations are now directly contradicted by fossil data, and consigned to our ever growing book of historic, discredited reconstructions. But this is always a possibility in palaeontology: our views of these animals are only ever hypotheses based on a sparse, biased fossil record, and every new discovery risks overturning someone's favourite concept. The fact we're able to move on from these reconstructions is positive, as it means we're a little less uncertain about the past, and a little closer to the truth.

Enjoy monthly insights into palaeoart and fossil animal biology? Support this blog for $1, see bonus content, and get free stuff!

This blog is sponsored through Patreon, the site where you can help online content creators make a living. If you enjoy my content, please consider donating $1 a month to help fund my work. $1 might seem a meaningless amount, but if every reader pitched that amount I could work on these articles and their artwork full time. In return, you get access to my exclusive Patreon content: regular updates on research papers, books and paintings, including numerous advance previews of two palaeoart-heavy books (one of which is the first ever comprehensive guide to palaeoart processes). Plus, you get free stuff - prints, high quality images for printing, books, competitions - as my way of thanking you for your support. As always, huge thanks to everyone who already sponsors my work!

References

    • Bell, P. R., Campione, N. E., Persons, W. S., Currie, P. J., Larson, P. L., Tanke, D. H., & Bakker, R. T. (2017). Tyrannosauroid integument reveals conflicting patterns of gigantism and feather evolution. Biology Letters, 13(6), 20170092.
    • Brusatte, S. L., Carr, T. D., & Norell, M. A. (2012). The osteology of Alioramus, a gracile and long-snouted tyrannosaurid (Dinosauria: Theropoda) from the Late Cretaceous of Mongolia.
    • Carr, T. D., Varricchio, D. J., Sedlmayr, J. C., Roberts, E. M., & Moore, J. R. (2017). A new tyrannosaur with evidence for anagenesis and crocodile-like facial sensory system. Scientific Reports, 7.
    • Chang, C., Wu, P., Baker, R. E., Maini, P. K., Alibardi, L., & Chuong, C. M. (2009). Reptile scale paradigm: Evo-Devo, pattern formation and regeneration. The International journal of developmental biology, 53(5-6), 813.
    • Chiappe, L. M., & Göhlich, U. B. (2010). Anatomy of Juravenator starki (Theropoda: Coelurosauria) from the Late Jurassic of Germany. Neues Jahrbuch für Geologie und Paläontologie-Abhandlungen, 258(3), 257-296.
    • Dawson, T. J., & Maloney, S. K. (2004). Fur versus feathers: the different roles of red kangaroo fur and emu feathers in thermoregulation in the Australian arid zone. Australian Mammalogy, 26(2), 145-151.
    • Dhouailly, D. (2009). A new scenario for the evolutionary origin of hair, feather, and avian scales. Journal of anatomy, 214(4), 587-606.
    • Dove, C. J., Rijke, A. M., Wang, X., & Andrews, L. S. (2007). Infrared analysis of contour feathers: the conservation of body heat radiation in birds. Journal of Thermal Biology, 32(1), 42-46.
    • Godefroit, P., Sinitsa, S. M., Dhouailly, D., Bolotsky, Y. L., Sizov, A. V., McNamara, M. E., ... & Spagna, P. (2014). A Jurassic ornithischian dinosaur from Siberia with both feathers and scales. Science, 345(6195), 451-455.
    • Hieronymus, T. L., Witmer, L. M., Tanke, D. H., & Currie, P. J. (2009). The facial integument of centrosaurine ceratopsids: morphological and histological correlates of novel skin structures. The Anatomical Record, 292(9), 1370-1396.
    • Hone, D. (2016). The Tyrannosaur Chronicles: The Biology of the Tyrant Dinosaurs. Bloomsbury Publishing.
    • Kahl Jr, M. P. (1963). Thermoregulation in the wood stork, with special reference to the role of the legs. Physiological Zoology, 36(2), 141-151.
    • King, J. R., & Farner, D. S. (1961). Energy metabolism, thermoregulation and body temperature. Biology and comparative physiology of birds, 2, 215-288.
    • Knoll, F. (2008). Buccal soft anatomy in Lesothosaurus (Dinosauria: Ornithischia). Neues Jahrbuch für Geologie und Paläontologie-Abhandlungen, 248(3), 355-364.
    • Lennerstedt, I. (1975). Seasonal variation in foot papillae of wood pigeon, pheasant and house sparrow. Comparative Biochemistry and Physiology Part A: Physiology, 51(3), 511-520.
    • Milinkovitch, M. C., Manukyan, L., Debry, A., Di-Poï, N., Martin, S., Singh, D., ... & Zwicker, M. (2013). Crocodile head scales are not developmental units but emerge from physical cracking. Science, 339(6115), 78-81.
    • Morhardt, A. C. (2009). Dinosaur smiles: Do the texture and morphology of the premaxilla, maxilla, and dentary bones of sauropsids provide osteological correlates for inferring extra-oral structures reliably in dinosaurs? (Doctoral dissertation, Western Illinois University).
    • Myhrvold, C. L., Stone, H. A., & Bou-Zeid, E. (2012). What is the use of elephant hair?. PloS one, 7(10), e47018.
    • Phillips, P. K., & Sanborn, A. F. (1994). An infrared, thermographic study of surface temperature in three ratites: ostrich, emu and double-wattled cassowary. Journal of Thermal Biology, 19(6), 423-430.
    • Stettenheim, P. R. (2000). The Integumentary Morphology of Modern Birds—An Overview 1. American Zoologist, 40(4), 461-477.
    • Sullivan, C., & Xu, X. (2017). Morphological diversity and evolution of the jugal in dinosaurs. The Anatomical Record, 300(1), 30-48.
    • Xu, X., Norell, M. A., Kuang, X., Wang, X., Zhao, Q., & Jia, C. (2004). Basal tyrannosauroids from China and evidence for protofeathers in tyrannosauroids. Nature, 431(7009), 680-684.
    • Xu, X., Wang, K., Zhang, K., Ma, Q., Xing, L., Sullivan, C., ... & Wang, S. (2012). A gigantic feathered dinosaur from the Lower Cretaceous of China. Nature, 484(7392), 92-95.