Form and Function of the Skull of the Ice Age Bush Dog
Mammoths, sabre-toothed cats, and giant ground sloths may be among the most iconic animals of the ice age, but the diversity of life during this time in Earth’s history extends well beyond them and includes one of the smallest and most peculiar of canids: Speothos pacivorus, the Pleistocene bush dog. A new study by Juan Ruiz and colleagues, published in Journal of Vertebrate Palaeontology, takes a careful look at how the skull and jaw anatomy of this species differ from that of the extant bush dog (S. venaticus), which was also around in the Pleistocene, through a classical comparative anatomy approach, an approach using a digital analysis of shape known as geometric morphometrics, as well as a virtual biomechanical approach utilising finite element analysis (FEA) to investigate how physical stress distributes across an object when loads are applied on it. Before diving into the new research, we must first explore the evolutionary of history of canids and where the bush dogs fall among its relatives.
In the aftermath of the calamity that befell the dinosaurs 66 million years ago, mammals flourished, with many forms quickly filling the roles of small predators. Among these would have been the first carnivorans, and these weasel-sized, agile predators would have been dwarfed by larger archaic mammal predators like the mesonychids. But the carnivorans wouldn’t remain small forever, as the extinction of the larger, archaic predators at the end of the Eocene, around 33 million years ago paved the way for carnivorans to diversify into the two modern divisions present today. One of which is the Feliformia, that includes the cats, hyenas, and smaller predators like mongooses. The other division is the Caniformia, and includes bears, seals, weasels, raccoons, skunks, and of course, the canids. Modern canids are highly successful, being found on every continent except Antarctica and in a large variety of habitats ranging from deserts to dense forests. Almost all canids superficially resemble a stereotypical wolf or fox, and while thinking of canids as such does not accurately capture their amazing diversity, it does underscore how important their generalised morphology has been to their success.
Within canids, the Cerdocyonina represents a diverse group endemic to South and Central America. Many of its species are fox-like, opportunistic predators, one of which, Cerdocyon thous, has a diet heavy on crabs, hence why it is colloquially named the crab-eating fox. One of most bizarre canids is also part of this South American lineage; Chrysocyon brachyurus, a relatively large canid with exceptionally tall legs that are thought to be an adaptation for traversing the tall grasses of South American savannahs. And curiously, the closest living relative of this species has some of the shortest legs relative to its body size of any canid; the bush dog (Speothos venaticus), which somewhat resembles a mustelid. Although the maned wolf and bush dog are the only living members of their lineage, back in the Pleistocene there were many more forms. The sub-lineage including the bush dog and all the extinct, ice-age species more closely related to the bush dog than to the maned wolf have been characterised as a clade of hypercarnivorous cerdocyonines. Previous research has shuffled around the evolutionary relationships between these South American hypercarnivorous canids, including how the extant bush dog is related to the Pleistocene bush dog (S. pacivorus), and some have even questioned whether they are truly different species.
Ruiz and colleagues set out to test whether the two are synonymous by carefully comparing the skulls of the two bush dog species, and identified several unique features characterising the teeth of the Pleistocene bush dog. The first molar (M1) of each tooth row of the ice age species is more complex than that of the modern form, as in it has 2 more “peaks”. Additionally, the ice age species had an extra socket for the second molars (M2) in its lower jaw. The researchers further supported their argument for the two species being genuinely distinct through the digital geometric morphometric analysis of skull shape they ran by digitising the skulls they studied. All the modern bush dogs were found to have very similar skull shapes, and while the shape of the Pleistocene bush dog was somewhat similar as well, it was still found to be notably distinct both to adult and juvenile extant bush dogs.
Now that Ruiz and colleagues had captured the traits that differentiate the ice age species from the modern one, and found empirical evidence for them having different skull shapes, they investigated whether this variation in form is related to variation in function. Using finite element analysis (FEA), they subjected the digitised models of the bush dog skulls to different mechanical constraints and loads that represent various types of prey capture and feeding. The scenarios they modelled reflect stabbing with different teeth along the tooth row, pulling-back with the skull after biting, and shaking and twisting of the head to tear apart flesh. Overall, the skulls of both bush dog species performed very similarly in all the modelled scenarios, with there being relatively low stresses across the skulls in the pull-back and head-twisting scenarios, and relatively high stresses across the skulls in the stabbing and head-shaking scenarios. In the latter scenarios where greater stress was predicted, the pattern of stress distribution of both species is again highly similar. A common theme of biology is that differences in form are followed by differences in function, but these bush dogs subverted such expectations.
Since the skulls of the two species performed so similarly, this likely means both were hunting in a similar way, and probably had to compete for resources. But lucky for them, the fauna of ice age South America was incredibly diverse, much greater than what is around today. Meaning that environments were bountiful enough to provide enough prey for two distinct, but very closely related small, mustelid-like canids to coexist. This is not to say there weren’t ecological differences between the two species at all, but instead that the researchers did not find evidence for any through the scientific methods they used. There could have very well been differences in geographical distribution, the extent of activity during different times of the day or night, and prey choice, which could have been linked to the differences in teeth of the two species. There is yet more to be discovered about these bizarre, hypercarnivorous canids as well as the ice age of South America more broadly. To learn more about this research, I’ve had the chance to interview the lead author of this new study; Juan Ruiz of Sao Paulo State University (UNESP), who previously featured on a Palaeocast blog post about notosuchian research he led.
Q1. Bush dogs and notosuchians are such different animals, how did you go from studying one to the other?
A1. Well, although I entered the palaeontological world by studying notosuchians in my master’s degree project, I have admiration for other groups as well… sharks, donkeys, ants, and, of course, carnivorans. So, when it was time to write a doctoral project based on finite element analysis, my supervisor and I decided to work with the canids of the genus Speothos, which comprises two species, the living bush dog (S. venaticus) and the Pleistocene ‘cave jackal’ (S. pacivorus). These critters are awesome both biologically as well as historically: they were described by the guy who become the ‘father of Brazilian palaeontology’, the Danish naturalist Peter W. Lund. And the fossil species was described before the living one!
Q2. The differences between the skulls of the two bush dog species you identified are so minute, how difficult was it to draw detailed comparisons between them? How much was this difficulty impacted by where the specimens were accessioned?
A2. One of the main objectives of my doctoral thesis was to investigate if both Speothos species are, actually, just one, something that was assumed by some authors. The diagnosis is based on a few characters, such as the tooth number and relative size (S. pacivorus being fairly larger), and as the last detailed work on S. pacivorus dated from the 1980s, I needed to see the materials for myself. Although collected in Brazil, all the fossil remains of S. pacivorus are deposited in the Natural History Museum of Denmark. When I finally had the opportunity to see it in my own hands, I was thrilled to finally discover how different the Speothos could be. Well… not much: apart from the known characters, the cranium of both species appeared to have different proportions to me, but how much of this represented anatomical differences rather than subjectivity of my judgment? That was the reason to consider the cranial form through quantitative analyses such as geometric morphometrics.
Q3. Considering there is a great deal of complexity to the digital methods you used in this research, how did you learn them?
A3. This work is based on two approaches: a finite element analysis (FEA) was conducted to infer ecological information, and geometric morphometrics (GMM) analyses were made to investigate the possible cranial differences between S. pacivorus and other Cerdocyonina (the South American canids). Through my doctoral years, I was trained in FEA by my supervisor, Felipe Montefeltro, one of the leading zoologists employing this technique in Brazil. So, for me, no problem with FEA. For GMM, I had the opportunity to work with Fabio Machado, a Brazilian researcher located at Oklahoma State University, USA, who is a specialist in taxonomical studies based on statistical analyses. Finally, to work with these analytical tools in a digital model, we CT scanned the holotype of S. pacivorus, a cranium, which was reconstructed by Christina Kyriakouli, an expert on digital modelling based in the Senckenberg Tübingen, Germany. So, in a nutshell, this research was only possible by teamwork, and I believe that I had found the Dream Team.
Q4. Following on from the previous question, what are limitations of those methods?
A4. Speaking for FEA, of which I’m a bit more familiar, it’s a very laborious method. There are a lot of steps: the image acquisition, the reconstruction of the material digitally, the application of the biological properties to it, the creation of test scenarios, all those analyses and the resulting data… this usually takes time, a lot of time, and also machine power. As a result, FEA studies commonly have limited sampling. To conduct FEA you also need good computers, which is not a reality for many researchers around the world.
Q5. What do you think should be the next steps taken to further discover the lost diversity of South American canids? Are you planning on leading more research on them?
A5. For me, this study showed the potential to work with new, quantitative approaches to understanding biological aspects otherwise difficult to test or infer by traditional, qualitative ways. I’m looking forward to expanding my doctoral research to other members of Cerdocyonina as a whole, something that is already on the way. This is a group taxonomically and functionally diverse, which evolved from a common ancestor very recently (the last ~3.7 My). So how they diversified? How could they originate such specialized hypercarnivorous forms, and why most of them went extinct? There are a lot of questions waiting to be answered.
Q6. Do you have any advice for prospective students or researchers interested in studying the ice age of South America?
A6. If I may give one piece of advice to young South American researchers is to use all the capacities available to study this fascinating subject and not limit the research to a local impact. Apart from the possibility of learning new techniques and contacting other researchers, many of the fossils from our own countries are deposited abroad, so is nice to planning international activities. We usually need to look for funding opportunities for that, and this can be very exhausting and never guaranteed, but we cannot let ourselves get discouraged. Another advice is to make good scientific connections: they allow us to expand our research beyond our own limits, which is always a good thing to do.
Q7. For a final, fun question: What would you rather fight, a couple of peirosaurines, or a pack of bush dogs?
A7. Well, this is a good plot for an episode of ‘Animal Face-Off’ (who remembers that show?) I would prefer to ‘fight’ bush dogs, because, well, they’re dogs: I could do tricks and, by the end, I might make some new friends.
Read the paper here: https://doi.org/10.1080/02724634.2024.2438827
And if you’re interested in downloading a digital reconstruction of the holotype of Pleistocene bush dog, you can find it here: https://morphomuseum.com/articles/view/229
Hady George is a palaeontology PhD student at the University of Bristol researching the jaws of the earliest tetrapods among other things, and seemingly always has a pop science book somewhere in his bag.
Cover image: Artwork depicting a life reconstruction of the newly described procolophonid reptile Threordatoth chasmatos by Mark Witton.
Related Posts
Episode 162: Cerney Wick →
The Evolution of Predatory Synapsids →
Modern Vs Archaic Mammals, a match-up decided by brains →
