Whence the beaver? They’re kangaroo rats, not squirrels!

March 9, 2017 • 1:00 pm

Of course the title is clickbait, but it does express a new finding: that, among Rodentia (yes, beavers are rodents), whose phylogeny was till now a bit unclear, we now learn that beavers are more closely related to kangaroo rats than to squirrels. For a long time, beavers had been thought to be closely related to squirrels (the “sciurid rodents”) because of the similar arrangement of their masseter muscles—the muscles that close the jaw. Recently, there was some slight but not completely convincing evidence, however, that beavers may be more closely related to kangaroo rats: those cute hopping mice in the family Heteromyidae. (Heteromyids also include pocket mice, kangaroo mice, and spiny pocket mice.) The molecular evidence was based on a similar piece of DNA in beavers and heteromyids: a single “retrotransposon,” a “jumping gene” that moves around the DNA by being transcribed from its RNA and then stuck in different places in the genome.

So we have a muscle similarity coming up against a single molecular similarity. Well, a new paper in Nature Scientific Reports by Liliya Doronina et al. (reference below; free download), using a lot more molecular data, shows that the kangaroo-rat affinity wins. This is based on a phylogeny constructed from both DNA sequences as well as the presence and position of retrotransposons.

It turns out that beavers, compared to other groups of rodents, share seven new retrotransposons with the kangaroo rat, and none with other groups of rodents. This shows that beavers and kangaroo rats are monophyletic: they have a common ancestor that is not a common ancestor with any other rodent. Below you can see what the new rodent phylogeny looks like, and you can also see, along the right, the similarity of the muscles between squirrels and beavers.

Note that they used the Eurasian beaver (Castor fiber), rather than its good old New World counterpart, the North American beaver (Castor canadensis). But that doesn’t matter, for the two species of beavers—there are only two and they diverged about 9 million years ago—are more closely related to each other than to kangaroo rats or any other rodent.

Before I give the reveal, here are the animals:

A Eurasian beaver:

A North American beaver (much cuter!):

A kangaroo rat (Dipodomys sp.):

It turns out that a similar arrangement of the masseter muscle evolved three times independently in rodents, so that’s not a good character to use for making evolutionary trees; it’s an “evolutionary convergence” that doesn’t tell us much about ancestry. DNA is much better, and here’s the final tree:

(From paper): 3,780 potential phylogenetically informative retroposons were extracted from the beaver reference assembly and projected onto sequence information of other rodent genomes and onto PCR-amplified orthologs from Anomaluromorpha. These newly revealed markers are shown as enlarged red balls. Previously identified phylogenetically diagnostic retroposon markers are indicated by black and two conflicting yellow balls. The two screening strategies and the resulting diagnostic presence/absence patterns are indicated for Castorimorpha and also the mouse-related clade. The myomorphous, sciurimorphous, and hystricomorphous zygomasseteric systems are illustrated to the right (blue and red lines show anterior parts of medial and lateral masseter, respectively; for details of zygomasseteric systems in rodents see Potapova27). The mandible types20 are noted: sciurognathous and hystricognathous. For the squirrel-related clade, only the zygomasseteric system of Sciuridae is presented. The rodent paintings were provided by Jón Baldur Hlíðberg.

At last we can rest easy, knowing that the beaver is not a close relative of the squirrel. The similarity of their muscle configuration undoubtedly comes from their similar habits of gnawing tough stuff, which led to a convergent arrangement of strong jaw muscles.

h/t: Matthew Cobb


Doronina, L., A. Matzke, G. Churakov, M. Stoll, A. Huge, and J. Schmitz. 2017. The beaver’s phylogenetic lineage illuminated by retroposon reads. Nature Scientific Reports 7, Article number: 43562 (2017) doi:10.1038/srep43562

29 thoughts on “Whence the beaver? They’re kangaroo rats, not squirrels!

  1. Totes adorbs!

    USA has the Eagle.

    Russia has the Bear.

    England has the Lion.

    Canada, my country, has the…beaver!

    Beavers are nothing if not hard-working and industrious, eh?

  2. So cool. It always feels immensely satisfying to see a concrete phylogeny. It’s a snapshot of the present through the lens of the past.

  3. Not quite sure why the “rest easy” not closely related to the Beaver? Something wrong with the Beaver, maybe?

    Apparently not so close to another animal as I would think, the Muskrat (Ondatra zibethicus). They are both mostly water animals and have survived much human interference.

    1. Clearly adaptation to a partially aquatic habitat is a suite of traits that can evolve rapidly and relatively easily together.
      Have I just set myself up for a lecture on the bloody “Aquatic Ape” hypothesis again? It must be a couple of months since it last got flogged to death.

  4. Note too the good example of evolutionary convergence seen in the jerboa and kangaroo rat.

  5. Note that bipedal hoppers are shown to have evolved in three of the major clades [jerboa, springhare, kangaroo rats]. All desert critters as well.

    I imagine the more mesic-adapted kangaroo mice are still within the mouse radiation?

  6. At last we can rest easy, knowing that the beaver is not a close relative of the squirrel. The similarity of their muscle configuration undoubtedly comes from their similar habits of gnawing tough stuff, which led to a convergent arrangement of strong jaw muscles.

    The more genetically distant morphologically similar animals turn out to be, the more evidence there is of convergent evolution.

  7. Actually, this phylogeny shows that the sciuromorphous/sciurognathus conditon of the skull and jaw is the primitive condition, and that the myomorphous and hystricomorphous conditions evolved independently from it. No need to specifically explain why beavers are similar to squirrels in their skull anatomy and musculature.

    1. + 1. I was just going to warn Prof. Coyne that with this disparagement of our Eurasian beavers, and without a trigger warning to boot, he has performed a meso-aggression.

  8. It is always nice to get confirmation of things that had been found previously. In your post, though, you have undersold the previous evidence and, thus, oversold the importance of the new. It is correct that only a single *retroposon* had been found previously to support Castoridae + Geomyidae, but there was strong to very strong support for that hypothesis from DNA sequence data (Huchon et al. 2002, Mol Biol Evol 19:1053; Blanga-Kanfi et al. 2009, BMC Evol Biol 9:71, and Fabre et al. 2012, BMC Evol Biol 12:88).

    1. We, the authors completely understand the concerns of Ron DeBry, who appears to base his comment on this layman’s post, whereas the actual peer-reviewed publication correctly honored the pioneering, sequence-based analyses that he cited. The news is that we found significant phylogenetic signals for the close relationship of beavers to Geomyidae by a completely independent and virtually homoplasy-free marker system; thus, confirming the relationship supported by signals of classical sequence analyses of diagnostic nucleotide changes with an even securer marker system.

  9. I’m assuming that the beavers being officially introduced into Scotland (and hopefully, those being unofficially reintroduced at other locations) are Eurasian beavers, not North American ones.
    How did we get so far in the thread without mentioning Leslie Nielson and the young lady’s nice, recently-stuffed beaver?

  10. Why are they called kangaroo-rats? The European colonization of North America began in the early 1600s, that of Australia in the late 1770s. I would not think that the kangaroo-rat was unknown, or remained nameless to the North American colonists until after the English settled Australia one hundred and fifty years later. The knowledge of what a kangaroo was would have taken further time to percolate across the ocean. What was the kangaroo-rat’s original name, and who changed it?

  11. Thank you for this – and you are forgiven for the clickbait.

    We have lots of beavers here in Colorado, and they can saw through a thick Aspen in no time. Wonderful and beautiful species. And yes, North American Beavers are by far the more adorable.

Leave a Reply