Get a Backbone

by James Morris
Illustrations by Talia Niederman

What’s a vertebrate? That’s easy – A vertebrate is an animal with a backbone.Hagfish

Not so fast. It turns out that not all vertebrates have a backbone. That is, they don’t have the bones that make up the backbone, called vertebrae. Hagfish, for example, don’t have vertebrae, but are classified as vertebrates.

That doesn’t make sense. How can a vertebrate not have vertebrae?

Let’s backtrack, so to speak. Because hagfish don’t have a backbone, they were for some time not considered vertebrates, as shown here –HagfishTree
According to this evolutionary tree, vertebrates include jawless fish like lampreys, jawed fish like salmon and perch, amphibians like frogs and salamanders, reptiles like lizards and snakes, and mammals like us.

But not hagfish.

Although they lack a backbone, hagfish share many traits with vertebrates, particularly lampreys. Hagfish and lampreys both lack jaws, have an enlarged head, and possess well-developed sensory systems, for example.

Due to these and other similarities with vertebrates, we included hagfish in a larger group called “Craniates,” which are vertebrates + hagfish, like this –HagfishTree2
However, recent molecular studies are forcing us to re-examine these relationships. It turns out that hagfish are even more similar to lampreys than we first thought. Furthermore, evidence now suggests that the ancestors of hagfish most likely did have a backbone, but they lost it over evolutionary time.

So, it’s not that hagfish evolved from an ancestor without a backbone. Like lampreys, they most likely evolved from an organism with a backbone – a vertebrate. As a result, we now include hagfish along with the rest of the vertebrates, even though hagfish don’t have a backbone, as shown here –HagfishTree3
That’s how an organism without vertebrae can be classified as a vertebrate.

If hagfish are not our closest non-vertebrate relative, what is? The answer to this question has also been re-thought in recent years.

Vertebrates belong to the phylum Chordata. That is, all vertebrates are part of a larger group called chordates. In addition to vertebrates, this group includes cephalochordates (small marine creatures such as lancelets) Lancelet
and tunicates (which include sea squirts and sea tulips).Tunicate
At first glance, it would seem that neither group shares much in common with vertebrates. They don’t have a backbone, nerves in the head region called cranial nerves, or specialized sensory organs.

But, on closer look, we start to see resemblances. They have anatomical, embryological, and molecular characteristics that unmistakably reveal that lancelets, tunicates, and vertebrates are closely related to one another. All three, for example, are segmented and possess a rod-like structure called a notochord that runs the length of the body.

But which is closest to vertebrates – lancelets or tunicates? For a long time, the answer was lancelets, as shown here –LanceletsTree
But, like the story with hagfish, recent molecular work has up-ended this view. Now the tree is thought to look like this, with tunicates earning the spot of our closest non-vertebrate relative –TunicatesTree
This is perhaps surprising to anyone who has seen a tunicate. They really look nothing like a vertebrate. The key here is that it’s not the adult form that resembles a vertebrate, but the larval form, which looks kind of like a tadpole.Tunicate Larva
Probably one of the most problematic groups to place in the vertebrate family tree is turtles. Hawksbill Sea Turtle
Turtles are famous for their shell, but in fact have many anatomical traits that are unusual. They clearly belong with the reptiles and birds (in a group called Sauropsids), but where exactly?

The anatomy of their skull provided us with what looked like a clear answer. If we consider vertebrates that can reproduce on land (amniotes), including turtles, reptiles and birds, and mammals, we notice that turtles have no holes in the temporal bone of their skull, reptiles and birds have two, and mammals have one. These shared traits help us to organize them as follows –TurtlesTree
But, again, molecular work suggests another picture entirely, with turtles more closely related to crocodiles and birds than they are to lizards and snakes, as shown here –TurtlesTree2
Like hagfish, turtles evolved from an ancestor with a trait (2 holes in the temporal bone) that they lost over time (0 holes in the temporal bone).

What can we learn from these three stories? We sometimes look at early branching groups as living representatives of common ancestors. For example, we think that hagfish must look like early vertebrates, or that having no holes in the skull is the ancestral condition for amniotes. But this may not be the case. Hagfish and turtles are both unusual in some way, and likely adapted for particular ways of life. So, instead of representing ancient forms, they may instead be highly specialized forms.

But there is something else. These changes and revisions help us to see that science is an ongoing process. As we accumulate more information, we change, dismiss, alter, and refine our hypotheses. And an evolutionary tree is a hypothesis – the best explanation we have, taking into account all of our observations and the data we have. So, as we collect more data, we adjust our trees.

Changing trees doesn’t make science tentative or fickle; instead, it is this very willingness to change and alter our models in light of new evidence that makes science such a powerful way to understand the world.

Brandon Amrosino, writing recently in The Boston Globe, explained, “‘The willingness to change’s one mind is among the most necessary of scientific virtues’ said [Christiana] Peppard. Real scientists must have intellectual humility and an openness to surprising conclusions.”

While sometimes it’s good to stick with you what you believe and have a backbone, sometimes it’s best to be flexible.

© James Morris and Science Whys, 2016.

4 thoughts on “Get a Backbone

  1. Nick I.

    Really like the last line – very clever!

    Some of the phylogenetic trees derived from molecular studies like sequencing do seem counterintuitive at first. I am sure there will be many more surprises to come.

    With the growing realization of the importance of horizontal gene transfer, endosymbiosis, etc. (and the possibility that life may have arisen in several places independently on Earth, giving multiple common ancestors), I wonder if “evolutionary web” might be a more accurate term than “evolutionary tree,” e.g. as illustrated here:


Leave a Reply

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