Lungfish nearly walk, shed light on the invasion of land

December 13, 2011 • 6:14 am

The conventional wisdom about how our tetrapod ancestors invaded land (“tetrapods” are four-footed land animals that include birds, mammals, reptiles, and amphibians) was that the evolution of limbs with digits occurred about the same time as the the walking gait evolved, perhaps when a lobe-finned fish (“sarcopterygian”) like Tiktaalik began frequenting shallow waters.  Those ancestors might have propped themselves up in the shallows, and eventually made forays onto land for food, creating selection pressures on both morphology and behavior to move about on the land.  As this scenario goes, the typical land-animal leg with toes evolved along with the typical land-animal gait, which is an alternation of fore- and hindlimbs that push off the ground.

This scenario may have to be revised, though, in light of a new paper by Heather King and colleagues in the Proceedings of the National Academy of Sciences. What they found is that the closest living relative of tetrapods, the lungfish, seem to have a precursor of the alternating-limb gait and the ability to push off the ground, even though they don’t venture onto land.  And this suggests that the common ancestor of lungfish and modern tetrapods, which lived about 400 million years ago, was “preadapted” or “exadapted” to walk.  That is, that ancestor might have had its own adapted form of movement that could be co-opted for walking when its descendants invaded land.  It’s a gait-first, limb-next hypothesis.

King and her colleagues, who work in the building right next to mine, studied the movement of one of the rare remaining lungfish, Protopterus annectens. It’s from Africa, and looks like this:

Protopterus annectens

The lungfish were filmed from various angles as they moved about in water tanks and, surprisingly, the researchers found what for all the world looks like the precursor of a tetrapod-like gait.  The lungfish “walk” along the substrate using their pelvic fins (the evolutionary source of hindlimbs) as the sole source of locomotion. (They didn’t “walk” on their front, or pectoral fins, which are the ancestors of forelimbs in tetrapods.) Moreover, they were capable of raising themselves above the bottom of the tank with their pelvic fins.

All this is shown in four movies in the paper’s supplementary material. I’ve provided the links below along with the author’s descriptions. Movies 3 and 4 (especially 4) clearly show the beast lifting itself off the substrate with its pelvic fins.

Movie 1:  This movie shows the lungfish Protopterus annectens locomoting underwater in ventral view. Note that the pelvic fins alternate, and the pectoral fins do not move rhythmically. This movie corresponds to Figs. 1A and 2 A and C. Each square of the grid in this movie is 1 cm.

Movie 2:  This movie shows the lungfish P. annectens locomoting underwater in ventral view. Note that the pelvic fins begin by alternating, then make a discrete transition to a synchronous gait. This movie corresponds to Figs. 1B and 2 B and D. Each square of the grid in this movie is 1 cm.

Movie 3:  This movie shows the lungfish P. annectens locomoting underwater in simultaneous lateral and ventral views. In lateral view, the lifting of the body is evident, as is the range of motion of the pelvic fin, including movement in front of and above the articulation with the body. Each square of the grid in this movie is 1 cm.

Movie 4:  This movie shows the lungfish P. annectens locomoting underwater in simultaneous lateral and ventral views. Here we show an example of theeffectiveness of the pelvic fins in lifting the body. Each square of the grid in this video is 1 cm.

While we don’t know whether other lungfish do this, nor whether this condition was present in the common ancestor of lungfish and tetrapods, it’s tempting to speculate that the alternating-gait was present in this ancestor and then became coopted for walking when its descendants developed sturdy limbs with digits. Or, as the authors note, “That P. annectens uses its paired appendages for substrate-associated locomotion provides evidence for this trait arising in sarcopterygians before the evolution of tetrapods, and before the evolution of digited limbs.”

This finding also casts doubt on some previous work: the discovery of fossil trackways from the Devonian—”footprints” that have been interpreted as some of the earliest evidence for tetrapod locomotion on land.  Here’s one of them, the Valentia Island trackway from Ireland:

Since trackways like this lack unambiguous evidence for digits, the authors consider their status as evidence for tetrapod locomotion “now open to question.” They could have been made by lungfish/tetrapod ancestors, perhaps creatures like lungfish.  Obviously, these conclusions are tentative and subject to revision when more fossils are found.

If you want an in-depth look at how these results fit into the historical context of work on the evolution of land-dwelling animals, Carl Zimmer has a nice piece at The Loom, “A long walk to land.”  Zimmer, of course, has been writing about the water-land transition for a long time, most notably in his excellent book At the Water’s Edge : Fish with Fingers, Whales with Legs, and How Life Came Ashore but Then Went Back to Sea.

h/t: Matthew Cobb


King, H. M., N. H. Shubin, M. L. Coates and M. E. Hale. 2011.  Behavioral evidence for the evoluton of walking and bounding before terrestriality in sarcopterygian fishes. Proc. Nat. Acad. Sci. USA. Published online, doi: 10.1073/pnas.1118669109.

22 thoughts on “Lungfish nearly walk, shed light on the invasion of land

  1. Look at the Caves episode of Planet Earth: The cave angelfish shows how tetrapod locomotion came about: It walks (just like a lizard) up rocks in the face of the flowing water. Seeing that motion (borrowed from swimming) instantly sets off the lightbulb: Aha! There is how Tiktalik (and friends) learned how to walk.

    Permian tetrapod trackway in Road Canyon, Cedar Meas, in southern Utah.

  2. I checked with a tetrapod palaeontologist friend who did not much like the paper. But I’ll let experts argue it out!

  3. Cool movies. Lungfish are very cool indeed.

    This seems to me like another ‘revolutionary’ finding that is only ‘surprising’ if you buy or bought the strawman it topples.
    I mean, what else did people think lungfish were doing with their (lobed) fins?
    a) We already knew that coelacanths swim by alternating their (lobed) paired fins (though they apparently do not bottom-walk with them).
    b) the primitive tetrapod gait is clearly derived directly from swimming (see Shubin).
    c) there is nothing ineluctable about alternate-fin/foot locomotion and the possession of digits.
    d) ancestors of these African lungfish certainly had more robust fin/limbs than today’s (cf the Australian version). African lungfish in particular live in shallow, weedy floodplains. A snakelike slither coupled with supplementary pelvic-fin pushes might be the only good way to get around much of the time.

    The movies are cool, and careful measurements and descriptions–data–are what it’s all about. I just think this study’s evolutionary importance is being overhyped. That alternate-lobed-fin locomotion in shallow water preceded the evolution of the particulars of the tetrapod limb should surprise nobody.

    Am I missing something?

    1. That was my outsider’s reaction as well. Moreover, I thought that even the tetrapods that came well after Tiktaalik – like Acanthostega – had fully formed tetrapod limbs with all three of the basic elements, and yet were still largely aquatic in shallow ponds. So we don’t have to posutlate that the walking gait was developed on land at all.

      1. Me three.

        Nice work, tho. Too bad there’s so much emphasis in science on trying to appear ‘revolutionary,’ as you put it, Chas.

    2. Not being a biologist, I have to ask: is this innate behaviour or is it learned as being more efficient? Do very young lungfish use an alternating gait? I have seen examples of human young using a variety of modes before settling on the standard model.

      1. Ah yes, I was going to comment on how this parallels the similar discussion on apes gaits vs bipedalism. As here, bipedalism is “exadapted” and useful way before it became exclusive.

        I think using humans as a comparative model may be out, at least for a while. A few weeks ago I read a post on a paper where they may have caught neural imagining of our walking patterns as learned. We would do that in parallel with and later somewhat overriding the natural pattern generators of some or all tetrapods.

        That was IIRC proposed as explaining why for example rats generally (?) can walk with severely damaged spines but humans generally can’t. (Happily leading up to new proposals for therapies.)

        The reason for this modification in behavior was proposed, as I remember it, primarily because bipedal walking is ‘damned difficult’. (Maybe social and sexual benefits are added later in life as well, making us modify gait as we, well, goes.)

    3. I was going to raise the point of Coelacanth motion too. IIRC, I encountered it in either Peter Forey’s book on the History of the Coelacanth Fishes or Jenny Clack’s Gaining Ground: The Origin and Early Evolution of Tetrapods, both published around the turn of the millennium (Clack has a new edition ; Forey isn’t on the roster at the NHM any more, so may be retired.)
      But in either case, it’s hardly news.
      Those Valentia trackways are by all accounts weird. Very weird. But unless someone has clearly demonstrated that they’re not underprints (compressed sediment from several laminae below that actual skin-sediment interface), then I don’t attach excessive weight to the apparent absence of digits. I’ll have to arrange a bog-trot and wave-dodge to have a look at them next time I’m in Co.Kerry.
      I’ll keep my eyes peeled for further tracks underwater – since I’m most likely to be there with the SCUBA club.

  4. In the post, the link to the second movie is the same as link to the first movie. You should replace the filename sm01.avi with sm02.avi to get the correct link.

  5. I actually thought evolution supported alternate gait from the start, as fish move through water with alternating use of paired paravertebral muscles, snakes move similarly along the ground, and ground lizards, like my Argentinian tegu, who looks at times like his stumpy limbs are actually in the way more than useful propellers of motion.

    On the other hand, were motion not alternating early on, I’d expect more of us tetrapods to hop like kangaroos.

    Glad to see supportive research that fits this evolution-based logic.

  6. Bichirs (Polypterus sp.) have stout pectoral fins, with a dibasal bony endoskeleton and muscles. They also have lungs. Despite this, they are not lobe-finned fish or lungfish, but the most primitive of the rayfinned fish (which shows that lobed fins and lungs are primitive for bony fish). They certainly prop themselves up on the substrate with those lobed pectoral fins, but I’m not sure they ‘walk’ on them. I have two downstairs in an aquarium, perhaps I’ll go have a look.

  7. Awesome stuff. I keep fish and I did notice that they alternate their pectoral fins for locomotion. One of the species I keep are pearl gouramis, trichogaster leeri. They have these amazing filament style pelvic fins that they like to touch everything with, including fingers pressed up against the glass. The males also build bubble nests using their super neat labyrinth organs. 😀

    And speaking of fish nearly walking:

        1. Aw, it looks like mother and baby going for a stroll. That’s adorable! Frogfish have a very tetrapod like gait when they walk. They are fascinating creatures.

          1. I thought so, too, when I first encountered that vid, but IIRC, it actually shows a small male pursuing a female. (I believe I learned that on some SW forum; it’s been a few years…) Yet another YouTube vid that leaves you desperate for more info!

            I certainly agree–they are fascinating!

Leave a Comment

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