Fossil Beast Helps Fill The Backstory of Horses, Tapirs, and Rhinos

There used to be rhinos in North America. In fact, they originated on the continent. The earliest ones didn’t look like the large, thick-skinned beasts we know today. No, if you were to wander through the humid forests of Wyoming or Utah around 50 million years ago, the closest thing to a rhino that you’d see would be a slender, pony-sized mammal that paleontologists know as Hyrachyus. And speaking of ponies, some of the earliest horses wandered the same forests – Eohippus and other horses that stood on several hoofed toes and were the size of a small dog – as well as the first tapirs.

Rhinos, horses, and tapirs are all perissodactyls. Roughly speaking, that means that they’re all hoofed mammals that stand on an odd number of toes. And they weren’t the only such mammals to thrive in ancient North America. Extinct perissodactyls – the “sloth horses” called chalicotheres and the horn-faced brontotheres – thrived on the continent, too. Together, they’re a fantastic example of how many different forms can radiate from very similar ancestors. But where did perissodactyls get their start? An assortment of bones from an enigmatic animal points to an origin not in North America, but prehistoric India.

A restoration of Cambaytherium by Elaine Kasmer.
A restoration of Cambaytherium by Elaine Kasmer.

About 54.5 million years ago, a little before it slammed into Asia, India was an island. And on that island lived a small hoofed mammal that paleontologists have named Cambaytherium thewissi. Exactly which lineage of mammal this little herbivore belonged to was unclear when it was first described nine years ago. Some researchers suspected that it was close to the origin of a group called anthracobunids – themselves related to elephants and manatees – while others hypothesized a closer relationship to early perissodactyls.

Thanks to a new collection of 120 tooth and jaw specimens, as well as 100 bones from the body, Johns Hopkins University paleontology Kenneth Rose and colleagues have finally figured out where Cambaytherium and its kin fit in the mammal family tree. Cambaytherium and its close relative Nakusia come out as the closest-known cousins of perissodactyls. In other words, they weren’t true perissodactyls, but they were very close to the lineage that spawned the last common ancestor of the horses, rhinos, tapirs, brontotheres, and chalicotheres.

Where Cambaytherium fits in the mammal family tree. From Rose et al., 2014.
Where Cambaytherium fits in the mammal family tree. From Rose et al., 2014.

These relationships are underlain by anatomy. While Cambaytherium had some features in common with the first horses and rhinos, such as fused lower jaw bones, the hoofed herbivore also shared some traits with an archaic group of hoofed mammals called phenacodontids, such as the anatomy of its teeth and a skeleton less-suited to running. This doesn’t mean that Cambaytherium was the direct ancestor of the first perissodactyls. Rather, the beast embodies transitional features that give us a more detailed look at how the ancestor of the horse and rhino emerged from more archaic mammals.

Finding an animal so close to the first perissodactyls in prehistoric India also alters the scenario for how these mammals spread around the world. Between 115 and 112 million years ago, still in the Age of Dinosaurs, an island consisting of Madagascar and India split from Africa. By 88 million years ago, India broke off from Madagascar and continued northward. The isolation of these islands spurred the evolution of some weird mammals – like the recently-announced Vintana from Madagascar – but the ancestors of Cambaytherium were not among the odd spinoffs. These hoofed mammals were a more recent addition to India.

After the mass-extinction that wiped out the non-avian dinosaurs and other forms of life, 66 million years ago, India continued to drift northwards toward Asia. And as it got close, Rose and coauthors propose, the ancestors of Cambaytherium and the first perissodactyls somehow dispersed from ancient Afro-Arabia to India. Once on the big island, they continued to evolve in the confines of the island for a few million years, making India a crucible of evolutionary novelty. The island’s eventual collision with Asia brought these new forms in contact with larger, interconnected landmasses where they quickly proliferated, spreading east and west, north and south, in an evolutionary dynasty that is hanging on by the barest of threads today.

[Full disclosure: This research was funded by the National Geographic Society Committee for Research and Exploration.]


Rose, K., Holbrook, L., Rana, R., Kumar, K., Jones, K., Ahrens, H., Missiaen, P., Sahni, A., Smith, T. 2014. Early Eocene fossils suggest the mammalian order Perissodactyla originated in India. Nature Communications. doi: 10.1038/ncomms6570

5 thoughts on “Fossil Beast Helps Fill The Backstory of Horses, Tapirs, and Rhinos

  1. Wait a moment: does the family tree say that Perissodactyla and Afrotheria are more closely related to each other than either is to Artiodactyla, or do I need a new pair of glasses?

  2. Another morphological analysis that nests Paenungulata within Laurasiatheria rather than Afrotheria? Note the recent analysis finding Anthracobunids to be close to Perissodactyla rather than Paenungulata as previously thought. How would these analyses look if molecular characters and more diverse taxa were added?

  3. Dear Bryan,

    Just a brief note to tell you how much I enjoy your posts. They are always well-written, more detailed than the average online report, and thought-provoking. I immensely enjoyed what you wrote about Cambaytherium; this lovely little creature was, I believe, very close to the first Perissodactyls, with characteristics that were already similar to those of the first Eocene equids.

    I look forward to hearing more about Cambaytherium thewissi, the Anthracobunidae and Perissodactyl evolution in general.

    Keep up the WONDERFUL work!

  4. @Don Gibbons

    Adding molecular data would definitely outweigh any signal that morphological characters have for the relationships of living taxa and certainly favor Laurasiatherian monophyly. A better test would be adding fossil and living non-“ungulate” laurasiatherians (and non-paeungulate afrotheres!). The relationships outside of Perissodactyla were clearly not a focus when this data set was constructed and I would take anything you see there with a grain of salt. Morphology only analyses continue to favor Ungulata as a monophyletic grouping, AFAIK.

Leave a Reply

Your email address will not be published. Required fields are marked *