One of the first people faced with the task of interpreting sabertooth fossils was the nineteenth-century Danish naturalist P. Lund. In the 1830s, Lund devoted a lot of time and effort to exploring the caves of Lagoa Santa, in the Brazilian region of Minas Gerais. He had left Denmark in 1833 (at the age of thirty-two) to pursue botanical research in Brazil, but in 1834 he met his compatriot P. Claussen, a fossil collector who had worked in Argentina before coming to Brazil. Lund was immediately fascinated by fossils and paleontology, so he abandoned botany and moved to Lagoa Santa, then a village with fewer than five hundred inhabitants that was surrounded by numerous calcareous caverns, some of them rich in fossils. Many of the caves were actively exploited for saltpeter, with lots of fossils being destroyed in the process, so Lund set out to salvage as much material as possible (Paula Couto 1955; Cartelle 1994). With admirable dedication, he and his local assistants explored cave after cave, collecting over 12,000 fossils between 1834 and 1846. It was extremely hard work, but Lund’s fascination for the extinct fauna of Brazil led him through all the difficulties he encountered, and whenever he found a rich fossil site, his imagination was set aflame. One of his peak achievements was to find the first remains of perhaps the most spectacular sabertooth cat, the Pleistocene felid Smilodon populator. Lund’s first finds were just a few isolated pieces, and he thought they belonged to a hyena, naming the creature Hyaena neogaea in 1839. But in 1842, with the addition of a little more material, including a few more teeth and some foot bones, Lund – an adept follower of G. Cuvier, the father of comparative anatomy – soon realized that the predator actually belonged to the cat family.

Lund was convinced that the numerous bones of large mammals that he found in the caves had been dragged there by big predators, which retreated to their dark dens to feed at leisure. The identification of Smilodon as the dominant predator of its time rounded out the scenario in his mind:

In the second half of the nineteenth century, remains of Smilodon from the Pleistocene were also discovered in North America, although the fossils there were scantier – but that situation would change. In 1875 Major H. Hancock, then owner of the Rancho la Brea property in Los Angeles County, presented a Smilodon canine tooth, recovered from the asphalt pits on his property, to Professor W. Denton of the Boston Society of Natural History. In spite of the impression that this gift made, it took twenty-five more years for that society to join forces with other institutions and organize the first scientific investigations at La Brea. Then, during the first decade of the twentieth century, an enormous amount of fossil mammal bones were recovered from the site (Harris and Jefferson 1985). Among these were literally thousands of bones of Smilodon, representing hundreds of individuals of the North American species S. fatalis, closely related to but distinct from its South American cousin.

Thus in the early twentieth century, Smilodon was a well-established element of the known prehistoric world. It was not the first sabertooth to be known to science – that privilege apparently belongs to the genus Megantereon, ironically mistaken for a bear by Cuvier. But Smilodon was the first genus to be known from reasonably complete fossils; it remained the best-known one for many decades; and it still includes the biggest and most spectacular species recorded.

The identification of the concept “sabertooth” with Smilodon has inevitably masked one of the main facts of sabertooth history: diversity. This book deals with sabertoothed predators, a broader concept than that of the sabertoothed cats. While the latter indeed belonged to the cat family (including animals like Smilodon itself and closely related genera such as Megantereon and Homotherium), many sabertooths did not, and some of them would look quite uncatlike to a modern observer. Some really were big and heavier than any living cat, but others were smallish creatures, scarcely larger than your average house cat (figure 1.1). Some were contemporary with early humans and mammoths, but much of the evolutionary history of sabertooths took place long before mammoths, humans, or even our earliest hominid ancestors existed.

How, then, can we define the sabertooths? Briefly, we can say that they were a group of fossil vertebrates, most – but not all – of them mammals; they were all predators; and they all possessed a set of anatomical features that define them as sabertooths, including the presence of elongated upper canine teeth and other adaptations in the skull that allowed them to open their jaws in the huge gapes necessary to bite with such enormous teeth. But this is a rather complex definition, and its various parts need to be discussed in more detail.

First of all, sabertooths are fossils. All of them became extinct so long ago that no human being ever saw one alive in historical times – to our knowledge, at least – so everything we know about them is based on their fossilized remains. Claims of modern sightings of sabertooth cats in South America and Africa remain completely unsubstantiated. And although some zoologists have suggested that the living clouded leopard of South Asia qualifies as an incipient form of sabertooth because of its long canines, the similarities between sabertooths and this extant felid are limited and rather superficial.

The history of sabertooths spans an enormous length of geological time. Although the latest species, such as the famous Smilodon fatalis from Rancho la Brea, disappeared “only” ten thousand years ago, the earliest mammalian sabertooths lived some 50 million years ago (or Ma) in the Eocene of North America. Even earlier, the gorgonopsians, a group of so-called mammal-like reptiles with many sabertooth features, lived in the Permian period, long before true mammals or even dinosaurs evolved, thus stretching the history of sabertooths, in the broad sense, as far back as 250 Ma (figure 1.2).

Second, all sabertooths were synapsids – that is, they belonged to the large group of vertebrates that includes both the mammals and the mammal-like reptiles. In anatomical terms, the skull of synapsids is characterized by the possession of a single cranial aperture behind the orbits (or eye sockets), hence their name, which means “a single opening.” In mammals, that opening corresponds to the large area in the side of the skull where the temporalis muscle attaches to the parietal bone. In contrast, the diapsids – a large group of vertebrates that includes the dinosaurs, crocodiles, and birds – originally had two openings behind the orbits. The more familiar-looking of sabertooths were all mammals, and during the last 50 million years there were quite a lot of species of sabertooths, belonging to several unrelated families, orders, and even infraclasses of mammals. Sabertoothed predators include the following groups:

It is not impossible that nonsynapsid sabertooths could evolve, but it is improbable, and certainly we haven’t found any in the fossil record. Carnivorous dinosaurs such as the allosaurs, as well as modern varanid lizards, have sharp, flattened and serrated teeth that resemble the namesake canines of sabertooths to a remarkable degree, but otherwise the animals are just too different to be regarded as sabertooths.

Third, all sabertooths were predators. Apart from their fearsome-looking upper fangs, the shape of the rest of the dentition of sabertooths shows that they fed almost exclusively on meat. They were “hypercarnivores,” meaning that their dentitions were so specialized for cutting meat that they had lost most of their ability to deal with other food items, such as bone or vegetable matter. Many modern mammalian predators, including wolves and bears, are more versatile, with dentitions that are suited to crushing bones, crunching insects, and processing vegetable matter. These animals possess molar and premolar teeth with different shapes adapted to those ...