Field of Science

The Whale that Looked Like a Walrus

Odobenocetops (shown above in an illustration from de Muizon & Domning, 2002) was definitely one of the odder forms of whale. It was found off the coast of Peru in the early Pliocene, about 3 or 4 million years ago, where it seems to have converged in a number of details on the modern walrus (Odobenus rosmarus). Or perhaps it is more accurate to say that the modern walrus converged on it - while Odobenus also appeared in the early Pliocene (Berta et al., 2005), Odobenocetops may have had a slight head start on it - the records in the Paleobiology Database suggest that the earliest records of Odobenus (from Belgium and New Jersey) are of somewhat uncertain age.

The walrus feeds on molluscs, particularly soft-shelled bivalves. The thick upper lip bears a strong array of sensitive bristles for finding molluscs buried in the bottom sediment. However, unlike most molluscivores the walrus doesn't directly crush the shells of its prey. Instead, the shell of the mollusc is held in the thick lips while the tongue is used like a piston to remove the animal from its shell by suction. Odobenocetops had a broadened palate and evidence of a thickened upper lip (probably with similar bristles) like a modern walruses, and probably fed in the same manner. The interesting thing is that Odobenocetops also evolved enlarged tusks like the walrus. Before the discovery of Odobenocetops, most authors had believed that the walrus' tusks were unrelated to its unusual feeding method, having probably evolved instead through sexual selection. That Odobenocetops also combined suction feeding and tusks suggests that their combination might not be as accidental as previously thought. What exactly the role of the tusks may be in feeding is more uncertain. De Muizon & Domning (2002) suggested that they might function as guides for the sensory bristles, or they may have been used to guide prey animals towards the mouth.

One of the most unusual features of toothed whales (in a group well-provided with unusual features) is a tendency towards a loss of bilateral symmetry, with one side of the skull being more developed than the other. The reasons for this asymmetry are uncertain, but the most popular theory is that it is related to the development of the sonar system. Odobenocetops possessed one of the most dramatic examples of skull asymmetry in the cetaceans. The left tusk of the type specimen of O. peruvianus is estimated to have been about 20 cm in length (because the tusks were quite fragile, fossilised specimens are invariably broken). However, the right tusk was over twice as long, at least 50 cm. The type of the other known species, O. leptodon, is even more dramatic - the left tusk is about the same size as known for O. peruvianus at 25 cm, while the right tusk was over one metre! Oddly enough, the remainder of the skull was rather less asymmetrical than in other odontocetes, as Odobenocetops had lost the melon and therefore the sonar of other toothed whales. Sonar was probably unnecessary for a diet of less mobile animals than fed on by other whales.

Relationship-wise, Odobenocetops is included in its own family, but was closely related to the Monodontidae, the family including the narwhal and white whale (also known as beluga - image above of white whale from Whale Trust). Monodontids also have rather mobile lips compared to other whales. Even more significantly, Odobenocetops and monodontids both have a completely mobile neck with unfused vertebrae, in contrast to other cetaceans which have the first two to five vertebrae fused into a single mass. This extra head motility would have doubtless been critical in allowing Odobenocetops to become an efficient sediment feeder.


Muizon, C. de, & D. P. Domning. 2002. The anatomy of Odobenocetops (Delphinoidea, Mammalia), the walrus-like dolphin from the Pliocene of Peru and its palaeobiological implications. Zoological Journal of the Linnean Society 134: 423-452.


  1. Very cool -- I'd heard of this beast but didn't know much about it.

    The fact that two members of this "pan-monodontid" clade developed asymmetrical tusks independently of each other--and in completely opposite directions!--is quite interesting. I wonder what the developmental basis is.

  2. I love that wierd reverse-narwal thing it's got going on. If you have it, Chris, could you send me the referenced paper? I'd love to restore this walwhale.

  3. Funny I was just thinking about "odo" last week...

    A few random thoughts -

    Suction feeding in Odobenocetops seems likely...other cetaceans do it including Delphinapterus (Werth 2000), but "(probably with similar bristles)" !? Muizon (1993) speculated on the presence of vibrissae but I wasn't aware of any good evidence to suggest their 'probable' presence. If there is soft-tissue preservation or osteological evidence for bristles on any of the odo fossils I'd love to know!

    I'm not going to go into speculation on the role of tusks for benthic foraging, but I do wonder if walrus feeding is always solely a lips-and-tongue affair as it's so often presented. I know there is some good video footage to suggest this, but looking at the interior dentition on walrus it really looks to me like it's serving some function...tenderizing the clam after it's removed from the shell maybe? I don't know. It's interesting to note that both Monodon and Odobenocetops lack teeth aside from the tusks.

    On asymmetry: I suppose that Monodon is even more asymmetrical than Ododbenocetops from a tusk perspective (i.e. 0m/3m v. 25cm/1m). However, it raises interesting questions about balance and movement in "odo", having mass distributed that unevenly about the skull.

    River dolphins, or at least Inia, also have unfused cervical vertebrae. This is usually chalked up to the need for superior mobility when navigating fluvial environments, but riverine delphinids, Sotalia etc., seem to get around just fine. While neck mobility may have functional significance for foraging styles in the monodontidae , it may not be a novel adaptation but rather retention of a primitive character? Fusion may have happened after the monondontid(+odobenocetops) - delphinid split? I'm not sure what the case is in Physeter or the ziphids so I could easily be wrong... rather than look into this I'd better get back to Triassic biogeochemical nutrient cycling though...yay.

    Muizon, C. (1993). "Walrus-like feeding adaptation in a new cetacean from the Pliocene of Peru." Nature 36(6448): 745-748, doi:10.1038/365745a0

    Werth, A. (2000) "A Kinematic study of suction feeding and associated blah blah blah"
    Marine Mammal Science 16 (2) , 299–314 doi:10.1111/j.1748-7692.2000.tb00926.x

  4. de Muizon's evidence for bristle in Odobenocetops was, I believe, largely the enlarged maxilla with a more 'spongy' texture and increased vascularisation, suggesting that there was at least some sort of sensory apparatus attached to it. Bristles are unusual for cetaceans, yes, but not impossible - a number of species have bristles as juveniles, and Inia has a bristly beak.

    Cervical fusion is found in mysticetes and ziphiids as well as delphinids (I'm not sure about sperm whales). Whether the lack of fusion in monodontids is primitive or derived I'm not sure. Do all river dolphins have unfused cervical vertebrae?

    As far as dentition in the walrus goes, I suppose it's worth noting that the walrus lineage goes back quite some time (to the Miocene at least), and only the most recent genus looks to have been a suction feeder.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS