A Blog by Ed Yong

Fossil tracks push back the invasion of land by 18 million years


Around 395 million years ago, a group of four-legged animals strode across a Polish coast. These large, amphibious creatures were among the first invaders of the land, the first animals with true legs that could walk across solid ground. With sprawling gaits and tails held high, they took pioneering footsteps. Their tracks eventually fossilised and their recent discovery yields a big surprise that could rewrite what we know about the invasion of land. These animals were walking around 18 million years earlier than expected.

The evolution of four-legged creatures – tetrapods – is one of the most evocative in life’s history. It has been illustrated by a series of beautiful fossils that vividly show the transition from swimming with fins to walking on legs. These include Panderichthys, a fish with a large tetrapod-like head and a muscular pair of front fins. Tiktaalik expanded on these themes. Its head could turn about a solid neck. Its limbs had the fin rays of its fishy predecessors but clear wrist bones and basic fingers too. Tiktaalik could support itself on strong shoulder bones, bend its fins at the wrists, and splay out its hand-like bones.

These animals – the elpistostegids – have largely been seen as transitional fossils. Tetrapods supposedly evolved from these intermediate forms and eventually replaced them. You could draw all of their skeletons in the corner of a book and flick the pages to see the move from sea to land happen before your eyes. But the new fossil tracks tell a very different story. As one reviewer writes, they “lob a grenade into that picture”.

The earliest true tetrapods so far discovered were around 375 million years old and the earliest elpistostegids hail from around 386 million years ago. But the Polish tracks are 10 million years older still. These dates suggest that the elpistostegids weren’t transitional forms at all. They weren’t early adopters of new biological technology, but late-surviving relics that stayed in their fish-like state while other species had evolved new bodies and, quite literally, run with them. Per Ahlberg, who led the study, says, “I’ve been working on the origin of tetrapods for about 25 years, and this is the biggest discovery I have ever been involved in. It is enomously exciting.”

It would be tempting to cast animals like Pandericthys or Tiktaalik in the role of biological luddites, outstaying their welcome with outmoded bodies. But that would be an injustice. If these animals co-existed with tetrapods for at least 10 million years, it suggests that their bodies were stable, well-adapted structures in their own right. They weren’t just brief flirtations with sturdiness on the way to full-blown walking.

Ahlberg discovered the fossil tracks along with researchers from Warsaw University, led by Grzegorz Niedzwiedzki. The tracks are found in the disused Zachelmie Quarry, nestled among Poland’s Holy Cross Mountains. The area used to be part of a tidal plain. Rocks from the site and a few rare fossils allowed the team to confidently date the track-ridden layer to around 395 million years ago, the middle of the Devonian period.

Among this layer, Niedzwiedzki found several tracks of different shapes and sizes. He also found several isolated handprints and footprints that clearly show signs of toes and ankles. Many of these tracks (such as in the diagram at the top) were clearly made by an animal walking with powerful, diagonal strides, powered by sprawling right-angled limbs.


They weren’t the work of any elpistostegid, whose straight limbs and backward-pointing shoulders and hips would have created far narrower tracks. Elpistostegids would also have dug long central troughs as they laboriously dragged themselves along. No such troughs exist at the quarry. This tells us that the track-makers strode along using strong hips and shoulders to hold their bodies and tails off the ground.

Niedzwiedzki thinks that they were undoubtedly tetrapods, and big ones too. The animal that created the tracks above was just 40-50 cm long. But some of the footprints were 15cm wide and hinted at creatures that were around 2.5 metres in length. The largest print is 26cm wide and its maker was probably a giant.

The implications of the Polish tracks are so controversial that reactions from other palaeontologists have been, understandably, mixed. Ted Daeschler and Neil Shubin, who discovered Tiktaalik, both find the study intriguing, but not definitive.

For Shubin, the deal-breaker would be identifying the animals that made the trackways and establishing where they sit on the evolutionary tree. He says, “The skeletal anatomy, let alone evolutionary relationships, of a trackmaker is hard to interpret from a track or print.” For example, he says that a model of Tiktaalik‘s skeleton would produce a print much like the one in the paper if it’s mushed into sand, and different consistencies or angles would produce an even closer match. He adds, “There is nothing in Tiktaalik’s described anatomy that suggests it didn’t have a stride.” 

Daeschler agrees that “trace fossils such as these presumed tracks… are a notoriously difficult class of evidence to interpret with full confidence”. Nonetheless, he’s keeping an open mind and a keen eye on future developments. “Paleontology is a lively field in which new discoveries constantly refine our knowledge of the history of life on earth,” he says.  

Jenny Clack, the Cambridge scientist who discovered Acanthostega, has seen the Polish tracks for herself and finds them more convincing. Her only reservation is that the detailed prints don’t have any trackways to show how their maker moved, while the trackways themselves consist of blobs. “But so do lots of previously known tracks,” she says. “If you’d found those in other deposits in the last part of the Devonian, you wouldn’t have any qualms about them.” She’d like to see trackways of the detailed prints but she’s nonetheless excited. “It’s going to change all our ideas about why tetrapods emerged from the water, as well as when and where.”

On the question of where, many scientists have suggested that the invasion of land began at the margins of freshwater – at river banks, deltas, lakes or flooded forests. But the Zachelmire quarry wasn’t any of these. Most likely, it was a shallow tidal flat or perhaps a saltwater lagoon. The first tetrapods didn’t lurk in rivers, but trampled the mud of coral-reef lagoons.

Niedzwiedzki thinks that this revised locale makes a better staging ground for the invasion of land. Twice a day, the zone between high and low tide is awash with stranded marine animals that would provide a feast for marine creatures experimenting with life on land. He argues that life was driven aground by the rich availability of snacks.

This new setting for the rise of the tetrapods also helps to answer a big question raised by the tracks. If tetrapods were walking around 18 million years earlier than we thought, they and the elpistostgids must have large “ghost lineages” – periods when they must have existed but for which no fossils have been found. Actually, very few fossils have been found at Zachelmie Quarry at all. These sites, where tetrapods first marched onto land, may have been good at preserving footprints, but they haven’t been equally kind to bones.

But why do the fossils that have been found make it look a lot like the elpistostegids preceded the tetrapods? That’s more of a stumper, but Niedzwiedzki has a possible answer. He thinks that elpistostegids may have colonised new environments before their tetrapod peers (or at least those environments that would preserve their bones). It’s a nice hypothesis, but for the moment, it’s just that. There’s no clear answer, although the hunt for new fossils or tracks will hopefully provide one.

Clack is certainly excited by the new doors opened by this discovery. “People are now going to start looking in different places from where they traditionally looked,” she says. “The Polish trackways were only discovered by accident. Nobody had ever looked at these Devonian deposits in detail before. Now the same team are starting to look for body fossils and they’ve started to find some, but no tetrapods yet. I’m expecting stuff to come out from other parts of the world too, like China.”

And here, even the sceptics agree. “All scenarios are intruiging, but we simply do not know for sure,” says Shubin. “All the more excuse to continue to go out in the field and find skeletons!”


Reference: Niedzwiedzki. 2009. Tetrapod trackways from the early Middle Devonian period of Poland. Nature doi:10.1038/nature08623

All images: copyright of Nature

More on transitional fossils: 

23 thoughts on “Fossil tracks push back the invasion of land by 18 million years

  1. Ed,
    You say :
    1. “This tells us that the track-makers strode along using strong hips and shoulders to hold their bodies and tails off the ground.”
    and then later you say…
    2. “Niedzwiedzki thinks that this revised locale makes a better staging ground for the invasion of land.”
    Surely, if the first statement is true then the creature that made these tracks wasn’t invading the land – it was already master of it?

  2. Is it sure that these rocks are of terrestrial origin? Also, I am almost certain that the first invaders of the land were bacteria, fungi or algae.

  3. I too found the title of “invasion of land” misleading, thinking instead of animal invasion of land by insects.
    As for the fossils, like Shubin, I am not convinced but it’s a highly facinating topic and I await more evidence and new fossils.

  4. Every time you see “invasion of the land”, the words “by vertebrates” are behind it in small, white letters. Oh okay, I’ll take the criticism. I did think about insects and I remember correcting the text to this effect somewhere, but clearly not everywhere or in the headline. Nuts.

  5. Finally, mainstream support for Henderson’s hypothesis that Acanthostega and Ichthyostega were the secondarily aquatic descendants of earlier, more terrestrial tetrapods…
    Henderson, D. M. 1999. Late Devonian amphibians as secondarily aquatic tetrapods. In Hoch, E. & Brantsen, A. K. (eds) Secondary Adaptation to Life in Water. University of Copenhagen, p. 18.

  6. I wonder if people who work in this area are used to considering rich habitats that existed then but have since been eliminated by seabirds.

  7. I’m not very excited. For some reason it reminds me of the early Triassic “bird” track ways of Santo Domingo formation. Not that it would necessarily be something like that, though.
    And, again an analogy with dinosaur-bird evolution, I think it’s not quite right to say that the species previously thought as transitional forms are “not transitional forms at all”, but, as the text implies, it may be just like the so-called “temporal paradox” for the origin of birds (i.e.: “Deinonychus only appearing much later, 35 million years later, in the fossil record than Archaeopteryx”), which is not a real paradox at all, but the same underlying logic of the answer of “why there are still monkeys”.
    The only thing they may not be are transitional forms in the most strict sense of a phyletic gradualist fashion, with newer forms appearing and completely replacing some earlier form in the same adaptive “trend”, and also neatly disappearing and giving space to the new and more advanced, all pretty and linearly, all the time.

  8. Species named from fossil tracks or other trace fossils are known as ichnospecies. Similarly genera known only from trace fossils are called ichnogenera, see http://en.wikipedia.org/wiki/Trace_fossil#Common_ichnogenera.
    I find the whole business a little unreliable, because for most extant species there is no way congeners could be separated only from tracks. Then again, some paleontologist have weird practices in delimiting species…

  9. Is there a reason for supposing that vertebrates colonized land only once? The idea of Tiktaalik et al being relics implies that they had a common land vertebrate ancestor with the maker of these tracks. Could it be that Tiktaalik and its ilk were simply a later group of land vertebrates? It seems logical to me that many groups of Ray-finned fish would’ve been preadapted to colonizing land

  10. It would be quite easy for those (and there are few) that are well versed enough in the art and science of tracking to determine species from a track alone. Especially a track that has not been seen before that is fossilized in bedrock.

  11. I second Darren Naish’s post. I have never been comfortable with the idea that tetrapod limbs with digits evolved for walking underwater, which is my understanding of the current thinking in the field based on fossils like Acanthostega and Ichthyostega. This earlier fossil of a competently terrestrial tetrapod that predates those species does lend support to the idea that those forms were secondarily aquatic descendants of more terrestrial tetrapods. This may mean that tetrapod limbs did evolve for walking on land, rather than being a preadaptation that originally developed in the water for walking underwater.

  12. The nomenclature of animals is kept strictly separate from that of their tracks. Ichnospecies can be named from tracks – species can’t. :-)

    I second Darren Naish’s post.

    On his own blog (click on his name), he said he was not completely serious. After all, why do both Acanthostega and Ichthyostega retain functional internal gills, tail fins (dragging the lower lobe across hard ground would lead to constant bleeding at the very least), the inability to put their feet on the ground and thus to walk (Ichthyostega could have moved on land like a seal, Acanthostega with its thin bones… hard to imagine it was more terrestrial than a catfish), and so on…

    Is there a reason for supposing that vertebrates colonized land only once?

    Yes – all land vertebrates share a common ancestor exclusive of all other vertebrates.

    It would be quite easy for those (and there are few) that are well versed enough in the art and science of tracking to determine species from a track alone.

    This rock was very soft mud before the limestone hardened between the microscopic grains. No clear footprints can be left in such stuff.

    the idea that tetrapod limbs with digits evolved for walking underwater

    No, for moving through thick vegetation, and for sneaking up on prey without moving the tail fin and thus moving so much water that the prey notices.

  13. We can’t say whether vertebrates colonized land more than once, but we can say that only one lineage left good fossils that we have found. You might take that to mean the other colonists were failures, but luck has a bigger role in evolutionary (as in all) success than we like to think.
    Calling the earliest walkers “terrestrial” may be overstating the case. Today, tidal mud flats are a terribly hazardous environment that demands obsessive burrowing, but wouldn’t have been at the time.

  14. I should have written “limbed vertebrates” instead of “land vertebrates”. It’s not clear how terrestrial certain extinct limbed vertebrates were.

  15. David,
    Thanks for replying to my post. Perhaps secondarily aquatic was not a good term to use, I was influenced by Darren Naish’s sarcasm and my tremendous respect for him influenced my post. I have read Clack’s book and understand the theory of limbs evolving for moving through thick vegetation (although some of the frog-fish used as examples don’t use their “limbs” for moving through vegetation). My point is that the current thinking is that limbs evolved on animals that were fully aquatic.
    I would imagine that the animal that made these tracks would likely have internal gills and tail fin(s)and was amphibious or even primarily aquatic but evolved and used its limbs for walking on land. I think this find opens the door for the possibility that earliest limbed tetrapods did not have the anatomical contraints that Acanthostega et. al. had on terrestrial living. There are fish today (“walking” catfish, “walking” anabantids, and mudskippers) that have internal gills and fins yet still drag themselves around on land. It does not seem hard to believe that Acanthostega and Ichthyostega had ancestors that were more competent on land than they were, and that limbs did evolve for walking on land, and that these later forms resumed a more aquatic lifestyle than their ancestors. A major point made around Tiktaaliks discovery is that it’s forelimb anatomy allowed it to do a “push-up.” I don’t see how that fits into the “limbs evolved for walking underwater (through aquatic vegetation or not) theory.”
    Damn that Darren, always so sarcastic! Where did he comment on this story in his blog?

  16. Ok David, I found Darren Naish’s comment buried in the toad thread. I was hoping for a full post by Darren Naish on the subject but know that it is unlikely since the subject is being covered by so many others.

  17. Is there any way to tell whether the earliest leggy things found lived in fresh or salt water? Those would seem to correspond to very different sorts of origins. I’m presuming the seas were already pretty salty by that point, but maybe they weren’t, yet; somebody must know.

  18. Is there any way to tell whether the earliest leggy things found lived in fresh or salt water?
    In short, no. I mean, you can tell for a specific fossil (for example, I think these track-makers are the first marine examples of fishapods known), but until we know which fossils are “the earliest” that’s not much help. There is no obvious phlogenetic/physiological signal in our current knowledge, if that’s what you’re asking. They might have even been estuarine or anadronous or something.

  19. Yes – all land vertebrates share a common ancestor exclusive of all other vertebrates.
    I understand that, but my question was, couldn’t it be possible that the LCA of all current tetrapods was aquatic, but that the only surviving lineages are terrestrial (or secondarily aquatic) That is, that 400 million years ago or so, there was a wide range of shallow-water lobe-finned fish. Several of these lineages could’ve come on to land and radiated, but the lineages that remained aquatic died out, save for the coelecanths and lungfish. Or that all but one of those terrestrial lineages died out, outcompeted by our clade.

  20. I think I see what Christina is getting at. If “our” common ancestor was obligately aquatic, then its various descendants could have colonized land more than once, each becoming terrestrial in its own way, more or less convergently. My only qualm about this scenario is that definitions of “aquatic” and “terrestrial” get (as it were) muddy at that interface.

Leave a Reply

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