Friday 18 September 2015

Humps, lumps and fatty tissues in dinosaurs, starring Camarasaurus

I like to see fossil animals restored as if they belong in the world they're depicted in. That is, not just as basic, conservative reconstructions of ancient species in an certain landscape, but instead with colours, integument and soft-tissue adaptations suited for their possible lifestyles and the environments they frequented. To this end, last year I published an illustration of the Late Jurassic, North American sauropod Camarasaurus supremus as an species well adapted for life in arid settings. As a common part of the famous Morrison Formation dinosaur fauna, dry conditions would be familiar to Camarasaurus, and especially because it occupied the drier, desert-like southern extent of the Morrison palaeoenvironment. I rendered Camarasaurus as a dinosaurian camel, complete with several common cranial adaptations to resisting dry conditions and, most obviously, a fat hump on its back.

2014 restoration of Camarasaurus supremus, published in Witton (2014). Painted to make a point about palaeoart (as well as plugging the awesomeness of All Yesterdays), here's what the caption read. "Reasoned speculation in palaeoart. The sauropod Camarasaurus supremus depicted with adaptations for living in a very dry environment: enlarged nasal cavities to aid resorption of moisture, sealable nostrils to reduce evaporation, wrinkled skin to enhance heat dissipation, white and tan colouring to resist heat soaking, and a fat hump to store energy. Such features are speculative, but do not contradict any data we have for this taxon, and are consistent with the adaptations of modern desert-dwellers."

I decided to revisit this image this week to boost the sauropod content of Recreating an Age of Reptiles (coming soon, I swear!). In doing so, I decided to conduct some more research into the likely nature of non-avian dinosaur fatty tissues. I wanted to keep the fat store on Camarasaurus, as equivalent structures provide energy and water reserves for many modern desert species, and there's no reason to think that extinct dinosaurs would not have developed fat stores for similar purposes. However, is a camel-like hump really likely in a dinosaur? Can we credibly restore any details of dinosaur fats? These were questions I sought to investigate more thoroughly before jumping into my revisions.

Yo extant diapsids so fat

If we're thinking about how to restore dinosaur fats, we need to investigate what the reptile lineage is capable of when it comes to producing and storing fatty tissues. The composition of diapsid fats is a little different to our mammalian ones, although we share functionally comparable approaches to fatty tissue makeup in many respects, including responses to endothermic demands (Goff and Stenson 1988; Saarela et al. 1991; Azeez et al. 2014). Amniotes, as a whole, have fairly similar approaches and uses for fatty tissues, which is great, because that allows us to make some reasonable inferences about fossil species.

Modern reptiles generally have lower fatty tissue fractions than mammals because of their lower energy requirements (Birsoy et al. 2013; Azeez et al. 2014). However, this is not to say that they are incapable of storing large quantities of fat, or even putting on weight rapidly. Some reptiles are indeed lean species, but some - most famously certain geckos, but also some iguanas, skinks and snakes - periodically or permanently hold large stores of fat in case of hard times, or to prepare themselves for energy-intensive feats (e.g. reproduction or long distance travel). Reptiles generally sequester fatty deposits within their torsos or in their tails, but some species also store them in their armpits and in fat pockets located at the back of the head. Individuals of many lizard species are considered healthy when these regions are literally bulging with fatty mass. To my knowledge, these masses are not directly supported by the skeleton or other tissues: it is simply the cohesive nature of fatty tissues and dermis which keeps them in place. It is known that some lizards can pack their tissues with fat rapidly when necessary, some experiments finding geckos can increase their body mass by 50% in four days (enough fuel to sustain them for over half a year!) (Mayhew 2013). Indeed, reptiles are so good at packing on fat, and maintaining it, that owners pet reptiles will know that obesity can be a real issue for captive lizards.

What about living dinosaurs? As with other diapsids, birds can rapidly generate fatty tissues in anticipation of stressful periods, and frequently do so before, for instance, migrating (Lindström and Piersma 1993). 10-15% body fat is considered low for a migrating bird, with the bodies of some species comprising 50% fatty tissues before embarking on their travels - seasoned ornithologists recognise birds as positively emaciated when they finish their journeys (Alerstam and Christie 1993). However, birds are not fully reliant on fatty tissues as energy stores, some species routinely using their muscles and organs as fuel sources during long migrations. It seems only their lungs and brains are safeguarded against being turned into energy (Battley et al. 2000): everything is fair game for fuel or other components needed to maintain a functioning body. Avian fatty tissues are, like those of lizards and crocs, deposited within their torsos but, in lieu of large tails, they also store them across the surface of the chest and abdomen. Bird skin has some transparency, and field ornithologists interested in avian fat tissue fractions can determine their extent by simply checking the amount of yellowish fat tissue visible underneath bird feathers (e.g. Rogers 1991).

The dinosaur hump controversy

Is there any direct indication of fatty tissues in Mesozoic dinosaurs? The answer is probably 'no', except for the controversial idea that the elongate dorsal neural spines if some dinosaurs are indicative of a camel-like 'hump' morphology. Spinosaurus, Ouranosaurus and Deinocheirus are key species here, these animals being depicted sometimes as humpbacked creatures. These interpretations are not the sole remit of artists, either: Bailey (1997) proposed that the tall neural spines of certain dinosaurs supported masses of tissue acting as energy stores or heat buffers - in other words, a heap of fat.

I must admit to being very sceptical that neural spine anatomy is linked to fat humps. For one,it seemingly violates what we see in the extant phylogenetic bracket for dinosaurs, where no species (to my knowledge) have substantial fat deposits on their backs. Of course, it might be queried how meaningful phylogenetic bracketing is for this issue. Fatty tissues seem quite pliable in an evolutionary sense, being chucked around animal bodies with ease as lineages adapt to new conditions (Birsoy et al. 2013). It isn't crazy to think that dinosaur bodies are different enough from those of modern diapsids that they could not have their own take on fat distribution, and there are certainly functional constraints on extant diapsid fatty tissues which are unlikely to apply to non-avian dinosaurs. However, that's only speculation, and one which conflicts with a big pool of direct data on this issue.

Another approach might be to look at animals which do have fatty humps on their backs - several types of mammal - to see if their composition is analogous to anything we see in non-avian dinosaurs. What do their humps look like internally?

A collection of animals with humpbacks and sails. Fatty humps are not directly supported by skeletons in modern species including (B) lowland gorillas (Gorilla gorilla), (C) dromedaries (Camelus dromedaries) and (D) white rhinoceros (Ceratotherium simum). Vertebral spines anchor sails in some modern lizards, such as crested chameleons (Trioceros cristatus; E), and withers anchor powerful neck muscles as in American bison (Bison bison; F). Cropped figure from Witton (2014); B–D and F from Goldfinger (2004); E historic x-ray (1896) by Josef Maria Eder.

Turns out that most mammalian humps are akin to those bulging reptile fat masses mentioned above: they tend to exist without internal support or even osteological correlates. Where humps do correlate with bone, they are comprised of powerful musculature, not fat: the shoulder humps of rhinos and bison show this well. These structures might have subcutaneous fat on them, but this is not their primary composition, nor does fat storage seem to be a principle adaptive purpose. In several species, like camels and rhinos, the longest neural spines do not align with soft-tissue humps at all, these actually being located over dorsal vertebrae with smaller neural spines (camels) or short-spined cervical vertebrae (rhinos). Taking our attention away from mammals, and turning to reptiles, we see that elongate neural spines anchor laterally compressed sail-like structures, not masses of fat. It thus seems that we have no modern correlation between fatty humps and skeletons at all, and that there is no link between elongate neural spines and fatty deposits - quite the opposite actually seems true. It was this suite of observations which led to my 2014 humped Camarasaurus image: bizarrely, it is more consistent with modern data (though still extremely speculative) to put a camel-like hump on something without long neural spines, like Camarasaurus, than it is to put one on Spinosaurus, Ouranosaurus or Deinocheirus. Sail-like structures or (at least for the lower regions of the spines) muscle attachment seem more parsimonious interpretations of their strange vertebrae - if we're being scientific (as we should be in palaeoart), we really shouldn't be looking at those tall neural spines and thinking 'fat hump correlate'.

Tying all this together

Although we may lack direct evidence of them from fossils, data from extant animals suggests it is sensible to restore dinosaurs with noticeable, prominent fatty tissues, especially if we're reconstructing animals associated with extremes of behaviour, climate or environment. Animals about to undertake migration should look well fed and bulky, and those at the other end might look leaner and less nourished. We certainly have good precedent for restoring desert-dwelling Mesozoic dinosaurs - of which there are many - with energy and water reserves, given that even energy-limited ectothermic diapsids take such precautions, as do some endotherms. We should probably not limit fatty tissues to bulky energy stores, either: as in modern lizards, some extinct reptiles may have housed pockets of fat in prominent places to serve as advertisements of health and virility.

Where should we locate those big energy stores? With no direct indication from fossils, I suggest we err on the side of caution and follow the diapsid condition, principally locating them around the tail base and abdomen. Most Mesozoic dinosaurs had well-developed, powerfully muscled tails, and were thus likely capable of supporting a wad of adipose tissue at the tail base. We could start restoring humps in other places, but it seems sensible to keep speculative anatomy grounded somewhere. Besides, it's not like a fat-tailed dinosaur is boring concept!

Combining all this together, I'll leave you with the completed, revised version of my desert-adapted Camarasaurus image, now with fatty tissues fully consistent to those of modern diapsids. This meant chopping off the back hump (I'm not going to pretend I wasn't disappointed to do that), but it's worth it for a more defensible image. Note that the adult is sporting not only a fat tail, which is meant to represent sustenance for wandering through harsh desert settings, but also a pair of natty fat pockets behind the skull. It looks fairly happy with them.

Camarsaurus supremus, queen of the desert, not a member of Weight Watchers.

This post was brought to you by Patreon

Production of this image and article was sponsored by my awesome Patrons, who back me at Patreon. As a bonus, they were privy to production of this image and ongoing commentary on the changes made to it. If that sounds like something you're interested in, you can get involved for as little as $1 a month. You'll also get access to other exclusive content, discount print prices and other rewards for your troubles. A huge thanks to those who have signed on already!


  • Alerstam, T., & Christie, D. A. (1993). Bird migration. Cambridge University Press.
  • Azeez, O. I., Meintjes, R., & Chamunorwa, J. P. (2014). Fat body, fat pad and adipose tissues in invertebrates and vertebrates: the nexus. Lipids Health Dis, 13, 71.
  • Bailey, J. B. (1997). Neural spine elongation in dinosaurs: Sailbacks or buffalo-backs?. Journal of Paleontology, 1124-1146.
  • Battley, P. F., Piersma, T., Dietz, M. W., Tang, S., Dekinga, A., & Hulsman, K. (2000). Empirical evidence for differential organ reductions during trans–oceanic bird flight. Proceedings of the Royal Society of London B: Biological Sciences, 267(1439), 191-195.
  • Birsoy, K., Festuccia, W. T., & Laplante, M. (2013). A comparative perspective on lipid storage in animals. Journal of cell science, 126(7), 1541-1552.
  • Goldfinger, E. (2004). Animal Anatomy for Artists: The Elements of Form: The Elements of Form. Oxford University Press.
  • Goff, G. P., & Stenson, G. B. (1988). Brown adipose tissue in leatherback sea turtles: a thermogenic organ in an endothermic reptile?. Copeia, 1071-1075.
  • Lindström, Å., & Piersma, T. (1993). Mass changes in migrating birds: the evidence for fat and protein storage re-examined. Ibis, 135(1), 70-78.
  • Mayhew, W. W. (2013). Biology of desert amphibians and reptiles. In: Brown, G. W. (Ed.). Desert biology: special topics on the physical and biological aspects of arid regions (Vol. 1). Elsevier.
  • Rogers, C. M. (1991). An Evaluation of the Method of Estimating Body Fat in Birds by Quantifying Visible Subcutaneous Fat. Journal of Field Ornithology, 349-356.
  • Saarela, S., Keith, J. S., Hohtola, E., & Trayhurn, P. (1991). Is the “mammalian” brown fat-specific mitochondrial uncoupling protein present in adipose tissues of birds?. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry, 100(1), 45-49.
  • Witton, M. P. (2014). Patterns in Palaeontology: Palaeoart-fossil fantasies or recreating lost reality. Palaeontology Online, 4, 1-14.


  1. Not sure if you see comments on older posts, so I am re-posting what I wrote on the original Recreating an Age of Reptiles post.


    I was surprised when I found out how much cheaper it is to use a printer and get a batch of books printed than to offer print-on-demand. (Also: shipping is a tiny fraction of the ludicrous price that Lulu charges for folding a book into a bit of cardboard and bunging it in the post.)

    It might be worth getting an online quote from somewhere like Lightning Source:
    I'm getting £700 including shipping for 100 copies colour printed in 70lb white paper. That goes up to £978 for hardbacks, so a bit under a tenner per copy. It's more hassle, and you're gambling that you'll be able to shift all the stock, but it's WAY better value.

  2. While I accept the idea of fatty deposits and other lumps of tissue exist on dinosaurs, is it really accurate to be assuming this based upon mammals? 5/6 out of those pictures are mammals and the one that is a reptile is rather close (in relation to the mammals being close) to its skeleton. Wouldn't basing off birds and reptiles be a better option? Apologies if I'm mixing something up, I'm just an amateur to paleontology.

    1. 4/5* pictures, my apologies.

    2. This comment has been removed by the author.

    3. Quote tags not goin', back to ye olde quotation marks:

      "Wouldn't basing off birds and reptiles be a better option?"

      Isn't that what this blog post was about...?

      "I decided to revisit this image this week to boost the sauropod content of Recreating an Age of Reptiles (coming soon, I swear!). In doing so, I decided to conduct some more research into the likely nature of non-avian dinosaur fatty tissues...
      I suggest we err on the side of caution and maintain the diapsid condition, principally locating them around the tail base and abdomen."

      If you want an idea of just a few of those diapsids with obvious fat stores, that Mark is talking about, look up eublepharid geckos, especially Eublepharis (plenty of pet photos online) and Hemitheconyx. (The clue is in the name ;) ) Also Heloderma, Tiliqua, and Uromastyx. As mentioned, other lizards and crocs have wide tail bases, but I think the effect's especially pronounced in these genera due to their desert habitats and the fact their tails don't go too much further than 'base'.

      On the blog post: nice! The old camarasaur hump didn't offend, but I definitely like the diapsid option. (Is the fact it's caring for a single offspring a more controversial depiction?) :) After reading some other bits about reotilian tail recently I start to wonder, do those mammals have to put fatty humps on the back and neck because their caudal 'quipment is just not up to the task?

  3. I'm skeptical of whether any gorilla actually has that much space between the top of it's skull and the top of it's head. It looks like the illustrator for the figure used a female or young male gorilla skull and outlined a large male silverback's head on to it. The hump in silverbacks is composed of neck musculature in the back and temporalis muscles on top of the cranium, but these are supported by insanely large sagittal and nuchal crests on the skull -- thus having clear osteological correlates. Gorillas without massive sagittal and nuchal crests (like the skull in the above figure), also seem to lack a large hump on their head and neck. Or, at least, I've never seen anything to suggest the contrary. Has anyone ever seen a published gorilla dissection that shows otherwise?

  4. As much as I love the All Yesterdays movement I get the impression that too many artists have taken it as a go ahead to do anything despite plausibility. This sort of well researched and reasonable speculation is exactly what All Yesterdays was about. I just wish depictions such as this would find their way into mainstream paleontology media in some form or another and influence popular ideas about prehistory to some degree.

  5. Camarasaurus are so cute.. I like them!

  6. This comment has been removed by a blog administrator.