For a medium sized sabertooth like Megantereon, remaining unseen was of the essence. With its strong, muscular physique it was not nearly a long-distance runner, so it needed to stalk within a few meters of its prey in order to catch it after a couple of spectacular leaps. But its prey, including medium sized antelopes, pigs or even horses, had a lofty ally: the giraffe.
Since the late Miocene, members of the modern giraffe lineage evolved their long necks and tall forequarters, and inevitably became the sentinels of the savannah. From the privileged viewpoint of their 5 meter height, they miss little of what goes on around them

Some 7 million years ago during the Turolian (late Miocene) the giraffid Bohlinia attica had already developed the large size and unique body porportions of modern giraffes. It inhabited the open woodlands of Europe, giving them a striking “African” touch.
bohlinia attica low res

Any good wildlife tracker in Africa takes advantage of the presence of giraffes in order to locate the big cats. As soon as a predator moves in the vicinity, giraffes stop their browsing or casual walking, and stand motionless pointing with their stare in the direction of the carnivore. That habit is as convenient for the less tall herbivores as it must be annoying for the predators.

Back in the Pliocene and Pleistocene of Africa, the giraffid Giraffa jumae was a contemporary of the sabertooth Megantereon whitei. While the cat tried to take advantage of every element of the landscape to conceal itself, it could hardly escape the stare of the giraffe, which surely ruined many an attempted stalk!
megantereon stalking


All cats devote a large chunk of their time to fastidiously taking care of their hygiene and if anything, sabertooths had even more reason to do so.
Right after killing a large prey, as it should do at least once a week, a sabertooth of the genus Megantereon would look like a bloody mess -literally. Unlike modern cats, its killing bite relied mostly on causing massive blood loss to ensure a rapid death of its victim, and that implied a somewhat untidy spectacle. But if we could find the sabertooth at any random moment, chances are we would find a pretty clean animal, which is only what you would expect from a cat.

If we could time-travel and come across the sabertooth Megantereon, it is very likely that it would be grooming itself
megantereon self cleaning

A sabertooth smeared in old blood would be pestered by flies, it would be prone to skin conditions, and would be more easily smelled by potential prey, so it needed to be quite conscientious about its grooming.
But of course there are other benefits to grooming, incluiding the fact, obvious to any cat owner, that it feels good and decreases stress! Social species would benefit from mutual grooming but a solitary animal, as Megantereon most likely was, would need to twist a bit around. But then its long and flexible neck would be a welcome aid in reaching those difficult spots!


The most usual time for me to find long-lost sketches is when I am looking for something else. A few days ago while searching my old folders I came across a few drawings which I thought were lost for good. These included some discarded sketches for the murals of the 1993 exhibit “Madrid antes del Hombre” (in an earler post I shared a few sketches which I did for that exhibit).
The scene which probably changed the most during conceptual sketching was the reconstruction of the Miocene fossil site of Paracuellos. Looking at the site’s faunal list I first envisioned a forest scene where a pack of bear-dogs harassed a chalicothere mother and her young. I just chose from the list the species which looked most appealing to me.

Here is my first sketch for the Paracuellos site reconstruction. The bear-dog-chalicothere interaction is the undisputed centre of attention.
boceto paracuellos version 1 a

But the scientific advisors tought that the woods actually occupied only a small fraction of the area where the fossils accumulated, so they advised me to show a more open environment. Also they asked me to show other species which were more abundant as fossils at the site, because the bear-dogs and especially the chalicotheres were quite rare finds.

My second version kept the “stars” of the first sketch (the bear-dogs), but they were now cornered in the right-hand section of the scene, leaving room for the more abundant species. Also in this case the victims of the bear-dog attack were not chalicotheres but primitive rhinos.
boceto paracuellos version 2

Unfortunately the scientists found that even this second version did not show clearly enough the inferred environment around the site. Alluvial fans were an important feature of the arid, seasonal landscape, but it was too difficult to properly show them from a ground level perspective. So in the end an aerial view was favoured, and both the chalicotheres and the bear-dogs were totally left out of the scene!

Here is the sketch of the final version, with the reference lines I drew in order to transfer the drawing to the large canvas.
boceto paracuellos vers 3 baja

And here is the finished painting, alluvial fans and all, but no chalicotheres!

I won´t deny that all these changes implied some degree of frustration for me. I missed the opportunity to turn that chalicothere scene into a full-fledged, large format oil painting. Over the decades I made attempts to include that scene in other projects, but it simply never happened. Who knows, some ideas may be destined to remain forever at the sketch stage!


Modern dogs differ from other carnivores in one interesting anatomical feature: they have a “nuchal ligament”, a string-like structure that runs along the dorsal part of their necks and allows them to support the weight of the head with little muscular effort. Dogs share this feature with ungulates, but there are differences in detail: while in ungulates the ligament extends from the spines of the thoracic vertebrae to the back of the head (hence the name “nuchal”, meaning “attaching to the nape”), in dogs it only runs as far ahead as the second cervical vertebra (the axis), so the term nuchal ligament is somehing of a misnomer here. But, at any rate, what would a carnivore want a nuchal ligament for? Such an adaptation makes sense in a cow or a sheep, who spend many hours grazing with their heads down, an activity that calls for some passive mechanism to save the muscular effort of supporting and then lifting the weight of the head. But carnivores don’t graze, do they?
Well, what dogs do is track scent trails. They walk and trot for long distances, nose close to the ground, as they search and follow their prey’s smelly paths. And they also happen to be relatively long-legged carnivores, which implies they need a long neck for their snout to reach the ground as the animal trots. So the neck of a wolf, jackal or coyote is proportionally very long, but some of the muscles that turn it to the sides and pull it up are relatively reduced, compared to other carnivores, partly because their role is taken by the nuchal ligament.

Here is a drawing of the neck of a wolf (Canis lupus), showing the skull and vertebrae (top) selected deep muscles (middle) and more superficial muscles. See how the “nuchal ligament” actually doesn´t reach the nape, just the back of the axis vertebra.
dogs Wolf-neck low

But, have the necks of dogs always been like that? Several years ago, while working on the reconstructions of fossil dogs for our book “Dogs: their fossil relatives and evolutionary history”, Xioaming Wang, the late Dick Tedford and myself looked in detail at the anatomy of the fossil dog Aelurodon, from the American Miocene. Fortunately there is an amazing collection of fossils of these animals at the American Museum of Natural History in NYC, so we had all the information we could hope for. While studying the cervical vertebrae I found something strange about them: they somehow resembled the vertebrae of a big cat, such as a leopard, more than they did the same elements in a wolf. Concretely, the vertebrae were relatively short, and the processes for muscle attachment were proportionally larger, projecting farther away from the vertebral body. When I assembled the bones to create a reconstruction, the neck looked suprisingly short, and when I reconstructed the musculature of the neck on the basis of the shape and position of attachment areas, it was evident that this animal had a more powerful neck than a wolf of comparable size. There is no obvious evidence for the presence or absence of a nuchal ligament, but the morphology of the back of the axis, where the ligament would attach, is rather different from that of modern dogs. Also, the short neck and its powerful muscles would make the function of such a ligament rather irrelevant.

Here is a reconstruction of the head and neck of Aelurodon. The morphology of the cervical vertebrae (top) implies a relatively short neck, while the shape of muscle attachment areas speaks of very strong musculature (center). Both features resemble the necks of modern big cats. When external layers are added, we see that the animal’s head and neck woudl look powerful and socky (bottom)
dogs Aelurodon-neck low

These anatomical differences must have implied differences in behavior, but it is not clear what differences those would be. Given the simmilarities with a cat’s neck, it is tempting to assume a more cat-like hunting style for Aelurodon, implying that the predator was more able to handle its prey individually, using its paws to restrict its struggles, and using its neck as a base for delivering a more precise killing bite, a bit like big cats do. Also it is possible that trotting for miles in search of scent trails as some modern dogs do was a less important part of its behavioral repertoire. But it is also possible that the predatory behaviour of Aelurodon was essentially similar to that of modern wolves and it simply had not evolved some of their anatomical refinements.

A broad comparison of neck morphology in fossil dogs revealed to us that the first taxon to clearly show a modern wolf-like neck anatomy was the late Miocene and Pliocene genus Eucyon. This animal also developed proportionally longer forelimbs than any of its earlier relatives, probably reflecting an adaptation to drier, more open environments, and wider foraging areas. So it would make sense to think that a modern foraging stlye in these members of the dog subfamily caninae was accompanied by the development of a modern forelimb and neck anatomy.
Only additional research will take us closer to solve these riddles, but one thing is certain: just like in the cat family, fossil dogs reveal a diversity of adaptations that we could hardly suspect by looking only at the living species. And the dramatic difference that we see today between dogs and cats could have been a bit more blurred in the distant past!

To learn much more about dog evoution and fossil record, check our book:


In 1960, at a time when little was known about the anatomy and body proportions of bear-dogs, American paleontologist Stanley Olsen described a wonderful collection of postcraneal fossils of Amphicyon longiramus, from the Miocene site of Thomas Farm in Florida. Olsen’s paper profiled a kind of predator with no living counterpart. With a body size comparable to that of a modern brown bear, Amphicyon had a longer and more flexible back and a long, heavy tail. Its dentition resembled that of a dog more than that of a bear, and was better suited for consuming meat and bone, while still allowing the animal a varied diet.

Here is my reconstruction of Amphicyon longiramus, based mostly on Olsen’s descriptions of the Thomas Farm material
a-longiramus-recons-2 low res

In the middle Miocene, when sabertooth cats hadn’t yet attained their full size and dominance, amphicyonine bear-dogs like Amphicyon were the undisputed ruling predators, both in Eurasia and in North America. Although they could not run especially fast or long, they were capable of ambushing large animals using a vaguely cat-like hunting style, and then they could use their great muscular strength and powerful canine teeth to bring down and kill their prey. But with the late Miocene the rule of the bear-dogs was challenged by the appearance of such powerful felid sabertooths as Machairodus. Later relatives of Amphicyon, such as Magericyon, (best known thanks to the fossil sample from Batallones in Spain)adapted to the new times by becoming somewhat smaller and developing more specialized dentitions for killing and consuming large prey efficiently. But there was no resisting the empire of the machairodonts, and near the end of the Miocene the bear-dogs disappeared for good after many million years of successful evolution.

Here is the reference of Olsen’s papers:
Olsen, Stanley J. 1960. The fossil carnivore Amphicyon longiramus from the Thomas Farm Miocene. Part II, Part II. Cambridge, Mass: Harvard University, Museum of Comparative Zoology.


More than 3 million years ago, somewhere in Northeastern Spain, a volcano exploded and created a large crater. With time the crater was occupied by a lake, known by geologists as a maar. Today, the lake is dry and the crater walls have been largely levelled by erosion, but an exceptional fossil site bears witness to the events that took place there. The site, called Camp dels Ninots, is close to the village of Caldas de Malavella in Girona.
Maar lakes often create remarkably beautiful scenery, with their encircling walls covered by forests, but some of them hide something more sinister. One example is lake Nyos in Cameroon, which in 1986 emitted a large cloud of carbon dioxide that killed thousands of people and animals. Such toxic gas emmissions occur because the pocket of magma beneath the lake leaks the carbon dioxide into the water, and perturbations such as those caused by a landslide can cause the gas to emerge with disastrous effects.
Three million years ago around the Camp dels Ninots maar lake, a rich fauna thrived in a subtropical environment, much warmer than today. Rhinos, tapirs and heavy antelopes crossed the forests and came down the crater walls for a drink at the lake margin. But several of those animals died mysteriously at the shore, and their bodies floated into the lake where they ultimately sank. Their skeletons, completely articulated, got exquisitely preserved as fossils, without any trace of having been disturbed by scavengers. A likely explanation is that the lake belched a cloud of toxic gas, instantly killing all the animals along the shore.
As often happens, the animals’ doom was the paleontologists’ blessing, since the death of so many creatures led to their pristine preservation. As the local authorities planned for an on-site exhibit, I was asked to create several reconstructions of the Pliocene environments and fauna of Camp dels Ninots, including a scene showing the hypothetical origin of the fossil accumulation.

This is a preliminary pencil sketch of the Cap dels Ninots death scene
camp dels ninots tafonomia boceto

And here is the finished painting that can be seen today at the fossil site

Can Argilera

Can Argilera

The scene shows antelopes of the genus Alephis and a rhinoceros of the genus Stephanorhinus on the shore, already showing signs of intoxication, while a tapir (Tapirus) lies on the ground, already dead. The bodies of other antelopes float in the lake while a pair of cormorants that happened to fly too low are already falling.
This and other reconstructions of the fauna and environments of Camp dels Ninots can be seen as part of the outdoors exhibit. Thanks to the findings made at the site, we now know a lot more about the anatomy of several species of Pliocene mammals, but, alas, no sabertooth skeleton has been found there as yet. But I don´t give up hope.

In central France, another Pliocene maar lake fossil site, a little younger in age than Camp dels Ninots, has been known for about a century, and it has yielded the most complete skeletons known to science of the typical sabertooth cats of the Pliocene: Homotherium and Megantereon. For several decades the French site, known as Senéze, was exploited without any serious excavation methodology, so we don´t know much about the taphonomy of those early finds. But since the 1990s, new field campaigns have yielded more accurate data, suggesting that the fossil mammal skeletons from Senéze accumulated as landslides coming down the crater walls trapped the animals and dragged their bodies to the lake.

I took this photo of the mounted skeleton of Megantereon from Senéze, exhibited at the Natural History Museum of Basel (Switzerland), back in 1990. All other findings of Megantereon fossils are more fragmentary, so it is only thanks to the exceptional conditions of the volcanic maar lake that we know nearly every bone of this sabertooth cat.
megantereon skeleton seneze basel

Only a portion of the ancient lake shore at Camp dels Ninots has been excavated this far. Sabertooths like Dinofelis diastemata, whose skeleton is mostly unknown, probably inhabited the area at the time, so it is not impossible that the next excavation will yield an amazingly preserved specimen…who knows?

Volcanic eruptions have provided the right conditions for the preservation of some of the best sabertooth fossils known to science, and not only beacuse of maar lake sites. If you want to know more about volcanoes and sabertooths, read my book “Sabertooth”!

Visit the on-site exhibition at Camp dels Ninots:


It was once thought that the enlarged upper canines of sabertooths evolved as an adaptation to pierce the skin of “pachyderms” such as elephants or rhinos and other gigantic herbivores of the past. But if we need just one proof against such an argument, then the nimravid Eusmilus should be it.
Members of the genus Eusmilus lived in Eurasia and North America in the Oligocene, millions of years before any true felid sabertooth ever evolved. But if we look at the skull of a well known species, such as Eusmilus bidentatus from France, we find that this early animal had taken its sabertooth specializations to a degree not seen even in the late Pleistocene felid Smilodon. Eusmilus not only had very elongated upper canines, but its whole skull was deeply remodeled to increase the biomechanical efficiency of its killing bite. But, while Smilodon was considerably heavier than even the largest modern cats, Eusmilus bidentatus was… smaller than a modern lynx!

Here is a life reconstruction of Eusmilus bidentatus based on Oligocene fossils from France. The animal was heavily muscled and powerful but with a shoulder height of about 45 cm it was shorter than a modern lynx
eusmilus revised low res

With its modest body size, Eusmilus could not even dream of attacking any of the thick-skinned behemoths that roamed the Oligocene woodlands and prairies, including many kinds of relatives of the rhinoceros. Eusmilus‘ elongated canines meant that, even with the additional gape provided by its specialized mandibular articulation, the clearance between the tips of upper and lower fangs was similar to that of a lynx. The size of the animals to which it could apply its killing bite was consequently rather small. It is thus evident that the sabertooth adaptations of this predator were not aimed at hunting giant, thick-skinned herbivores, but rather to the quick and efficient dispatching of small and medium-sized prey thanks to a killing bite that caused rapid death through massive blood loss, thus minimizing the danger of a trashing prey escaping or wounding the predator, or both.

Another consequence of its small size was that Eusmilus, like many other kinds of small sabertooths through the Tertiary, was not nearly the dominant predator in its environment. For millions of years, hyaenodonts, bear-dogs, and even the omnivorous pig-like entelodons, have abused these sophisticated but small sabertooths, and stolen their rightful prey.

This scene set in the Oligocene of France shows the cow-sized, omnivore entelodontid Entelodon evicting a couple of Eusmilus from their kill
Eusmilus-and-entelodon low res

Small as it was, Eusmilus was not the tiniest sabretooth to evolve. Other species of the genus, like the American Eusmilus cerebralis, was even smaller, and so was the Eocene creodont sabertooth Machaeroides, not taller than a house cat. It is funny to think how nice a pet one of these miniature killers would make, but leaving it to roam in the neighborhood could result in more bloody incidents than any modern house cat can cause…

Want to learn much more about Eusmilus and other mini-sabertooths? Get the award-winning book “Sabertooth” and have your fill of long-in-the-tooth predators!


Hyenas have long been the victims of human prejudice and superstition, from ancient tribal tales to “The Lion King”. That is a pity because it gets in the way of our perception of a group of amazing animals with incredible adaptations for their ecological niches.
Lions sometimes have a hard time defending their rightful kills against large hyena clans, but quite often it is the hyenas who lose their own prey to opportunistic lions. Such dynamics are not new, and there is every likelihood that the woodlands and prairies of the Miocene, Pliocene and Pleistocene witnessed comparable conflicts quite often.
In the Old World Plio-Pleistocene, it was the lion-sized sabertooths of the genus Homotherium who had to deal with the challenges of living next to a most impressive hyena: Pachycrocuta brevirrostris. This animal was considerably larger in its linear dimensions than the living spotted hyena, but it was also more robust, so its body mass would have been much larger.

Here is a life reconstruction of Pachycrocuta brevirrostris, based on fossil remains from China and Spain. With a shoulder height of about 1 meter, it was larger and far heavier than any modern hyena
Pachycrocuta life

P. breviorrostris shared all the adaptations of modern hyenas for cracking bones (massive, blunt premolar teeth, robust skull with a domed forehead) and for carrying large pieces of carcasses over long distances (long, well muscled neck, large scapula with a flat articulation for weight transmission, shortened back and hind limbs for stability). But its massive size took those adaptations to a different scale, and certainly it made the giant hyena a rival to reckon with for any competing predator.

The skeleton of Pachycrocuta was massive but essentially very similar to that of modern hyenas
Pachycrocuta skeleton

This illustration shows Pachycrocuta cracking a large ungulate bone, and a schematic view of the anatomical features involved in this action. The massive muscles of mastication (temporalis and masseter) provided the huge force necessary to crack the bone with the premolar teeth, and the domed forehead helped to dissipate the stresses generated during the bite
pachycrocuta bite

But, was P. brevirrostris a scavenger and a kleptoparasite of predators such as the sabertooths, or did it kill much of its own prey? This is a good question and one to which we may never get a final answer. On one hand, its skull and dentition were adapted to process bone at a phenomenal scale, so it was clearly very well adapted to scavenging. In fact, the cutting blade of its carnassial teeth was slightly shorter than in the highly predaceous modern spotted hyena, leaving more room for the crushing section of the dentition, a detail that suggests a more scavenging lifestyle.
On the other hand, the huge body mass of P. brevirrostris made it less efficient for this animal to forage through the enormous distances required in order to come across such a dispersed resource as carrion is. More purely scavenging species, such as the modern brown and striped hyenas, are much lighter, and actually weight considerably less than the more predatory spotted hyena. And while a large body mass can be a problem for long-distance foraging, it can be an advantage for active hunting, since one or several heavy hyenas can be more effective at subduing and bringing down a large prey animal.

Whether it killed or scavenged most of its food, the fact is that Pachycrocuta had the habit of bringing lots of it back to its den sites, a habit which apparently explains the origin of several remarkable fossil sites. If nothing else, paleontologists need to be grateful to this gigantic bone cracker for its efforts to collect hundreds of bones and gather them in the places where they eventually became preserved as fossils.

A family group of P. brevirrostris gather at the den site, where the cubs play with some old bones. Such dens, when placed near seasonal lakes or waterholes, could be buried by mud during floods and the bones would be preserved as fossils
pachycrocuta den site

Conferencia en el Gabinete de Historia Natural, próximamente

El próximo 18 de Marzo, impartiré la charla titulada “Los grandes felinos de África: evolución, pasado y presente”. Desde mi primera visita a Botswana en 1993, mis viajes a África en busca de los grandes felinos han supuesto un contrapunto y un complemento a mi trabajo de investigación sobre la anatomía, evolución y adaptaciones de estos animales. Pero a lo largo de estos años también he tenido ocasión de estudiar en diversos museos los restos fósiles de los félidos extintos de ese continente. En esta charla hago un rápido repaso a la historia evolutiva de los félidos en África, mostrando algunas de las especies más sombrosas que allí han habitado, así como unos breves apuntes de mis observaciones más sorprendentes de los leones, leopardos y guepardos en su ambiente. Finalmente proyectaremos en primicia mi documental “Belleza Salvaje: la visión de un artista de los grandes felinos africanos”.

La conferencia será en el marco del Gabinete de Historia Natural, un espacio de encuentro para todos los aspectos de la naturaleza en el centro de Madrid.

¡Os espero!

Lugar: Calle Victoria, 9. Madrid.

Hora 19:00

Entrada: 5 Euros hasta completar aforo.

anuncio gabinete


As we have seen in previous posts, the Miocene was a time of gigantic hyenas and hyena-like predators. But more than that, it was a time of hyaenid diversity. So, members of the hyaenid family occupied different ecological niches, and we talk of the “civet-like”, “mongoose-like” and the ”dog-like” hyaenas, besides the more familiar “bone crackers”.
Giant bone-crackers like Pachycrocuta, for instance, were no doubt spectacular animals, but my personal favorite are the dog-like hyenas. Comparable in build and body mass to today’s coyotes and wolves, these species combined their elegant, gracile skeletons with a “multipurpose” dentition that allowed them to take a variety of middle sized prey which they would consume to the last bone, but they also could search far and wide for any carcass in the landscape, scavenging both in an opportunistic and in a more determined way.
Years ago, during my visit to Hezheng in China I was fortunate to study first-hand an amazing sample of Hyaenictitherium fossils, including many postcranial bones that gave me a much clearer idea than I had before of the body proportions of these animals.

My step-by step reconstruction of Hyaenictitherium wongii based on the Hezheng fossils, starts with this drawing of the complete skeleton
hyaenictitherium skel low res

The next stage is the reconstruction of the musculature, for which my previous dissections of modern hyaenids and viverrids were enormously useful references
hyaenictitherium musc low res

Finally, here is the reconstructed life appearance of Hyaenictitherium, an animal that would have the approximate size of a modern coyote. The coat pattern is broadly based on that of modern hyaenids, especially that of the striped hyena and the aardwolf, but some reference to viverrids is also made
hyaenictitherium life low res

Hyaenictitherium wongii and similar species somehow filled in the Old World the niches that the true dogs were occupying in North America at about the same time. With time, some lineages of dog-like hyaenids evolved into the “hunting hyenas” of the Pliocene, apparently the only hyaenids that eventually made it to the New World… but that is a different story!


Get every new post delivered to your Inbox.

Join 1,316 other followers