Walking With Sea Cows

Sea cows once walked on land. Pezosiren leaves no doubt of that. This roughly 48 million year old mammal once trod over prehistoric Jamaica, and looked akin to a hippo with the skull of a manatee. Much like Pakicetus in the history of early whales, Pezosiren embodies a critical transitional period in the evolution of manatees and dugongs, yet the place where this amphibious sea cow was found did not match what paleontologists expected.

In the big picture of mammalian evolution, sea cows are paenungulates – members of a group that also encompasses hyraxes, elephants, and extinct branches such as the double-horned Arsinoitherium and the aquatic desmostylians. The earliest members of these lineages first appear in Africa shortly after the end-Cretaceous extinction of 66 million years ago, with the perplexing exception of the sea cows. The earliest, most archaic progenitors of today’s manatees and dugongs, such as Pezosiren, have been found in Jamaica. Anatomical and genetic evidence is clear that sea cows must have shared an African origin with the other paenungulates, but, until now, no one has picked up the fossil trail of the earliest sirenians.

Today, in PLoS One, paleontologist Julien Benoit and colleagues describe a bone from the Eocene of Tunisia that closes the geographical gap in the sea cow backstory.  The skull fragment is a petrosal – part of the temporal bone from the side of the skull. That might not seem like much to go on, yet the intricate, complicated features in this single bone allowed Benoit and coauthors to confirm that it belonged to a sirenian rather than an early elephant or hyrax. The researchers have wisely avoided naming the animal on the basis of such limited material. They simply call the mammal the Chambi sea cow. (A bit of an irony since the mammal was found in a freshwater deposit.)

The petrosal in question is about 50 million years old. If not slightly more ancient, the fossil is just as old as Prorastomus found in Jamaica, until now regarded as the oldest and most archaic sirenian. The ancestry of sea cows may go even deeper. If the Chambi sea cow and Prorastomus are about the same age, then there must have been an earlier dispersal event during which legged sea cows made their way across the Atlantic. Early whales followed a similar pattern. While whales first entered the water in what is now Pakistan and India, forms similar to Georgiacetus – found in the American south – were able to disperse across oceans even before they lost their hindlimbs.

How early sirenians spread around the world, starting in Africa and dispersing across the ancient Atlantic. Image from Benoit et al., 2013.

Since the Chambi sea cow is only represented by a single skull piece, Benoit and colleagues can’t say exactly what this animal looked like. Based on the mammal’s early age and relationships, it was probably a legged sirenian that looked more like Pezosiren than a modern manatee. And even though the sirenian probably had legs able to support the animal on land, the petrosal bone contains clues that the Chambi sea cow was a proficient swimmer.

As Benoit and collaborators point out, the inner ear of their early sirenian is similar in some ways to the inner ears of early whales. Part of the mammal’s middle ear is thickened, and, as in aquatic mammals, the size and curvature of the semicircular canals appear to be have been convergently adapted for keeping the sirenian’s head steady while swimming. The Chambi sea cow probably spent a great deal of time in the water, Benoit and coauthors suspect, similar to early, amphibious whales like Ambulocetus and Maiacetus.

Additional fossil finds will test some of these ideas about the Chambi sirenian. Still, the fact that the mammal lived in Africa confirms what zoologists and paleontologists suspected based upon genetics and anatomical traits shared with elephants and other paenungulates. Characteristics of modern animals, and a few fossil ones, hinted at a hidden history that paleontologists are just now beginning to uncover. Slowly, bone by bone, paleontologists are beginning to assemble the skeletal outline of how sea cows evolutionarily earned their name.


Benoit, J., Adnet, S., El Mabrouk, E., Khayati, H., Ben Haj Ali, M., et al. 2013. Cranial Remain from Tunisia Provides New Clues for the Origin and Evolution of Sirenia (Mammalia, Afrotheria) in Africa. PLoS ONE 8, 1: e54307. doi:10.1371/journal.pone.0054307

5 thoughts on “Walking With Sea Cows

  1. Puzzled. The legs and feet on that reconstruction of Pezosiren don’t look as though it would be up to swimming the Atlantic, even the Eocene Atlantic. And you’d think the earlier Chambi beast would be more terrestrial if anything, if it’s supposed to be ancestral to the Jamaican forms. How does this work?

  2. Dear chris, Good question. First, the hands and feet of Pezosiren are reconstructed on this photo (see http://www.dinosauria.org/blog/wp-content/uploads/2012/09/Figure2.jpg), second, the fact that a mammal looks like a terrestrial one does not imply that it can’t swim (think about elephant or Hippopotamus for example) and finally, rodents (the ancestor of guinea pigs for ex.) and primates (the ancestor of Howler monkeys for ex) have crossed the Atlantic sea at the Oligocene, at a time when it was larger and they were by far, more terrestrial than a stem sea cow. Thank you.

  3. I was taken aback by your dating of the end Cretaceous at 66 mya, since in my mind that number has always been 65… So I googled some, and found Renne (2008), where they improved the error margins in argon-argon dating dramatically, and were able to date that fateful day to 65.95 mya…

    Thanks, Switek! You’ve enhancified my edumacation once again.

  4. It should be mentioned, especially on the NGS blog, that Pezosiren was discovered and collected with the generous, decade-long support of the National Geographic Society — although they’ve never taken the credit for it that they deserve. Thanks, NGS!

Leave a Reply

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