DNA suggests whales descended from land mammals

If you were asked what kinds of animals a whale most resembles, what would you say? Fishes? Sharks? Manatees? Perhaps even seals and sea lions? Whales live their whole lives in the water, and have a host of anatomical and physiological features that make them feel right at home there. So if DNA encodes these characteristics, you might think that whales should have DNA most similar to other aquatic animals.

Yet evolutionary biologists have long suspected that whales descended from land mammals. Whales are classified as mammals, having hair (albeit not much) and producing milk for their young, and nearly all other mammals are land dwelling. But is there any more evidence of this idea? Indeed, paleontologists have uncovered fossils that seem to document a transition from land dwelling whale-like mammals to fully aquatic whales. The distribution of these fossils further suggests that these mammals were geographically restricted until they made the plunge into water. Developing whale embryos even have tiny hindlimb buds that eventually disappear, implying that these beasts still retain some of the genetic machinery to produce four limbs.

So if fossils and developmental biology seem to be telling the same story, what does DNA say? To the initial surprise of paleontologists, whales have DNA that is extremely similar to even-toed hoofed mammals (artiodactyls), so-named because they always have an even number of toes. This group include animals like cows, deer, giraffes, hippos, pigs and camels.

In fact, whales aren’t just genetically similar to these mammals, they are genetically nested within them. By this I mean they are more genetically similar to some hoofed mammals than these hoofed mammals are to each other.

You can see this in the phylogeny below coming from a study [1] that compared 164 different species of mammals using 35,603 letters of DNA. Whales, indicated in part by paintings of a humpback and sperm whale near the middle of the figure, are most genetically similar to hippos (Hippopotamidae), a perhaps unsurprising finding given the semi-aquatic nature of the latter animals. The next closest group includes a batch of other hoofed mammals such as giraffes (Giraffidae; indicated by the okapi painting), deer (Cervidae; also associated with a painting), as well as cows (Bovidae), pronghorn antelope (Antilocapridae) and others. Outside of this whale + hippo + deer, cow, pronghorn, etc. grouping are the pigs and their relatives (Suidae + Tayassuidae) and camels (Camelidae).

Mammal phylogeny
Mammal phylogeny

It can be rather shocking when you first think about this given how completely different whales look from these animals, but there are distinct clues in the fossil record that point to this same conclusion. For example, some of the proto-whale fossils, such as Pakicetus, share some features that are unique to even-toed hoofed mammals, including a special indentation (trochlea) on the bottom of an ankle bone (astragalus).

Astragalus of Pakicetus compared to other hoofed mammals
Astragalus of Pakicetus compared to other hoofed mammals

When considering the evidence from DNA in conjunction with fossils, biogeography, development and now-defunct genes, a compelling picture is painted in which a group of hoofed mammals, against all odds, transitioned to a life in the oceans. Perhaps next time you go whale watching, think of it as viewing a herd of underwater giraffes or oceanic camels. You probably won’t look at whales the same way again.

Questions for Creationists

Why would the Creator design whales to have DNA so similar to hoofed mammals? With all of their adaptations for living in the water, shouldn’t their DNA be more similar to fish, sharks or other aquatic mammals such as manatees or seals? Is it possible that whales and hoofed mammals belong to the same kind, but whales evolved from these hoofed mammals over just a few thousand years? Wouldn’t this involve major evolutionary change at extraordinary speeds? Is it just a coincidence that whale DNA seems to tell the same story as the fossil record, geography and development?


1. Meredith, R. W., Janečka, J. E., Gatesy, J., Ryder, O. A., Fisher, C. A., Teeling, E. C., … & Murphy, W. J. (2011). Impacts of the Cretaceous Terrestrial Revolution and KPg extinction on mammal diversification. Science, 334(6055), 521-524.


3 thoughts on “DNA suggests whales descended from land mammals

    1. Hi g!

      I believed I addressed a similar topic in a previous discussion we had, but in case it wasn’t clear, I’ll try to elaborate on it here. This might sound a bit complicated, which may be even more difficult since English isn’t your native language, so please let me know if you need to be explain anything better!

      So within the world of molecular phylogenetics research, the science of using DNA to reconstruct hereditary relationships among organisms, you basically have three types of results:

      (1) some relationships amongst organisms are always or almost always recovered, no matter how much data you use. From my own experience, for instance, I believe every analysis I’ve ever performed results in tigers and domestic cats coming out more similar to each other than they are to other non-cat animals.

      (2) some relationships are not apparent with small datasets (e.g., too few letters of DNA for comparison, not enough species for comparison) but become more and more robustly supported with larger datasets. One example of this might be the relationship of turtles to other reptiles. Initially they jumped all over the place (e.g., closer to lizards, closer to crocs, closer to crocs + birds, etc.) but with larger datasets, they seem to be consistently coming out as closest to birds + crocs. In other words, it appears to be a statistical power issue, namely that there isn’t enough information to infer the correct relationship.

      (3) some relationships, regardless of dataset size, give you different results depending on the precise methodology or precise sets of genes you use. The base of placental mammals, which the paper you provided a link to refers, is one of these examples. Two major groups of mammals always come out together: Laurasiatheria (carnivores, hoofed mammals, whales, pangolins, insectivores, bats) + Euarchontoglires (primates, flying lemurs, tree shrews, rodents, rabbits and pikas). Together, this large group is called Boreoeutheria. The rest of placental mammals include Afrotheria (elephants, manatees, hyraxes, African insectivores, aardvark, elephant shrews) and Xenarthra (sloths, armadillos, anteaters). The relative positioning of Boreoeutheria, Afrotheria and Xenarthra switches depending on the analysis.

      So examples in category 3 would seem to imply evidence against evolutionary theory. Different dataset, different analysis, and you get different relationships means that these results can’t be trusted, right?

      Well, it’s a little more complicated than that. First off, in my experience, the vast majority of examples fall into categories 1 and 2 described above, meaning there seems to be a consistent signal in the DNA pointing to hereditary relationships between organisms. You never get walruses being more genetically similar to flies than other mammals, or lemon trees being more genetically similar to penguins than other plants.

      But how do we explain examples in category 3? First off, nearly all examples in category 3 result in the inference of what we call short branches in phylogenetic trees. What this means is that there are very few differences in DNA that seem to unite one group of organisms with another. Assuming evolutionary theory accurately describes reality, what would lead to such few shared DNA substitutions? First off, if the lineages split very quickly, then there may not have been enough time for DNA substitutions to be fixed in their common ancestor, or any shared substitutions may have been erased with reversals in the DNA over time.

      Alternatively, perhaps ample time passed, but the organisms evolved very slowly. The speed of evolution depends on a number of things including the population size, whether natural selection is acting on the mutation, and how quickly does the organism reproduce. We think whales, for instance, are notoriously slowly evolving, in part because they typically only have a single baby and it takes a long time for them to become sexually active.

      Assuming the evolution has not happened very quickly or the amount of time that has passed was minimal, then interbreeding may have a major effect. Normally the way we think new species form, corresponding to those branches in phylogenies, is that a population of organisms splits into two, and the two populations stay separate and evolve into different species. If they stay isolated or at least don’t recognize each other as potential mates, then they will eventually be very different from each other. Now imagine two populations evolving separately from one another, and they are each accumulating different DNA substitutions, but for some reason an individual from population A mates with an individual of population B. That hybrid will have a mix of DNA from each population. If that hybrid then mates back with population A, and the population B DNA is neutral or provides an advantage, then it can spread throughout the population. This is a phenomenon called introgression, which you may have read in the news in the context of some humans possessing Neanderthal DNA.

      So going back to the base of placental mammals: if there was a rapid splitting between one population into three (the boreoeutherian, xenarthran and afrotherian lineages) and/or introgression between one or more of the populations, then there would be minimal signal in the DNA to separate the different lineages. In fact, because of the very nature of these phenomena, you can get different results depending on the sets of genes you use.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s