Short of rocketing into space, it is hard to imagine a bolder or more life-changing step than leaving the water for dry land. The two life-zones are different in so many ways that moving from one to the other demands a radical shift in almost all parts of the body. Gills that are good at extracting oxygen from water are all but useless in air, and lungs are useless in water. Methods of propulsion that are speedy, graceful and efficient in water are dangerously clumsy on land, and vice versa. No wonder 'fish out of water' and 'like a drowning man' have both become proverbial phrases. And no wonder 'missing links' in this region of the fossil record are of more than ordinary interest.
If you go back far enough, everything lived in the sea - watery, salty alma mater of all life. At various points in evolutionary history, enterprising individuals from many different animal groups moved out on to the land, sometimes eventually to the most parched deserts, taking their own private sea water with them in blood and cellular fluids. In addition to the reptiles, birds, mammals and insects we see all around us, other groups that have succeeded in making the great trek out of life's watery womb include scorpions, snails, crustaceans such as woodlice and land crabs, millipedes and centipedes, spiders and their kin, and at least three phyla of worms. And we mustn't forget the plants, onlie begetters of usable carbon, without whose prior invasion of the land none of the other migrations could have happened.
Fortunately the transitional stages of our exodus, as fish emerged on to the land, are beautifully documented in the fossil record. So are the transitional stages going the other way much later, as the ancestors of whales and dugongs forsook their hard-won home on dry land and returned to their ancestral seas. In both cases, links that once were missing now abound and grace our museums.
When we say that 'fish' emerged on to the land, we have to remember that 'fish', like 'reptiles', do not constitute a natural group. Fish are defined by exclusion. Fish are all the vertebrates except those that moved on to the land. Because all the early evolution of vertebrates took place in water, it is not surprising that most of the surviving branches of the vertebrate tree are still in the sea. And we still call them 'fish' even when they are only distantly related to other 'fish'. Trout and tuna are closer cousins to humans than they are to sharks, but we call them all 'fish'. And lungfish and coelacanths are closer cousins to humans than they are to trout and tuna (and of course sharks) but, again, we call them 'fish'. Even sharks are closer cousins to humans than they are to lampreys and hagfish (the only modern survivors of the once thriving and diverse group of jawless fishes) but again, we call them all fish. Vertebrates whose ancestors never ventured on to land all look like 'fish', they all swim like fish (unlike dolphins, which swim with an up-and-down bending of the spine instead of side to side like a fish), and they all, I suspect, taste like fish.
To an evolutionist, as we just saw in the example of reptiles and birds, a 'natural' group of animals is a group all of whose members are closer cousins to each other than they are to all non-members of the group. 'Birds', as we saw, are a natural group, since they share a most recent common ancestor that is not shared by any non-bird. By the same definition, 'fish' and 'reptiles' are not natural groups. The most recent common ancestor of all 'fish' is shared by many non-fish too. If we push our distant cousins the sharks to one side, we mammals belong to a natural group that includes all modern bony fish (bony as opposed to cartilaginous sharks). If we then push to one side the bony 'ray-finned fishes' (salmon, trout, tuna, angel fish: just about all the fish you are likely to see that are not sharks), the natural group to which we belong includes all land vertebrates plus the so-called lobe-finned fishes. It is from the ranks of the lobe-finned fishes that we sprang, and we must now pay special attention to the lobefins.
Lobefins today have dwindled to the lungfishes and the coelacanths ('dwindled' as 'fish', that is, but mightily expanded on land: we land vertebrates are aberrant lungfish). They are 'lobefins' because their fins are like legs rather than the ray fins of familiar fishes. Indeed, Old Fourlegs was the title of a popular book on coelacanths written by J. L. B. Smith, the South African biologist most responsible for bringing them to the world's attention after the first live one was dramatically discovered in 1938 in the catch of a
South African trawler: 'I would not have been more surprised if I had seen a dinosaur walking down the street.' Coelacanths had been known before, as fossils, but they had been thought extinct since the time of the dinosaurs. Smith movingly wrote of the moment when he first cast eyes on this astonishing find, to which he had been summoned by its discoverer, Margaret Latimer (he later named it Latimeria), to give his expert opinion:
We went straight to the Museum. Miss Latimer was out for the moment, the caretaker ushered us into the inner room and there was the - Coelacanth, yes, God! Although I had come prepared, that first sight hit me like a white-hot blast and made me feel shaky and queer, my body tingled. I stood as if stricken to stone. Yes, there was not a shadow of doubt, scale by scale, bone by bone, fin by fin, it was a true Coelacanth. It could have been one of those creatures of 200 million years ago come alive again. I forgot everything else and just looked and looked, and then almost fearfully went close up and touched and stroked, while my wife watched in silence. Miss Latimer came in and greeted us warmly. It was only then that speech came back, the exact words I have forgotten, but it was to tell them that it was true, it was really true, it was unquestionably a Coelacanth. Not even I could doubt any more.
Coelacanths are closer cousins to us than they are to most fish. They have changed somewhat since the time of our shared ancestor, but not enough to be moved out of the category of animals that, colloquially and to a fisherman, would be classified as fish. But they, and lungfish, are definitely closer cousins to us than to trout, tuna and the majority of fish. Coelacanths and lungfish are examples of 'living fossils'.
Nevertheless, we are not descended from lungfish, or from coelacanths. We share an ancestor with lungfish, which looked more like a lungfish than it looked like us. But it didn't look much like either. Lungfish may be living fossils, but they are still not very like our ancestors. In the quest for those, we must instead seek real fossils in the rocks. And in particular we are interested in fossils from the Devonian era that capture the transition between water-dwelling fish and the first vertebrates to live on land. Even among real fossils, we would be too optimistic if we hoped literally to find our ancestors. We can, however, hope to find cousins of our ancestors that are sufficiently close to tell us approximately what they were like.
One of the most famous gaps in the fossil record - conspicuous enough to have been given a name, 'Romer's Gap' (A. S. Romer was a famous American palaeontologist), stretches from about 360 million years ago, at the end of the Devonian period, to about 340 million years ago, in the early part of the Carboniferous, the 'Coal Measures'. After Romer's Gap, we find unequivocal amphibians crawling through the swamps, a rich radiation of salamander-like animals, some of them as large as crocodiles, which they superficially resembled. It seems to have been an age of giants, for there were dragonflies with a wing span as long as my arm, the largest insects that ever lived.* Starting about 340 million years ago, we might almost call the Carboniferous the amphibian equivalent of the age of dinosaurs. Before that, however, was Romer's Gap. And before his gap, Romer could see only fish, lobe-finned fish, living in water. Where were the intermediates, and what led them to venture out on to the land?
My undergraduate imagination at Oxford was fired by the lectures of the prodigiously knowledgeable Harold Pusey who, despite his dry and prolonged delivery, had a gift for seeing beyond dry bones to the flesh-and-blood animals that had to make a living in some departed world. * His evocation of what drove some lobe-finned fish to develop lungs and legs, which was derived from Romer himself, made memorable sense to my student ears, and it still makes sense to me even though it is less fashionable among modern palaeontologists than it was in Romer's time. Romer, and Pusey, envisaged annual droughts during which lakes and ponds and streams dried up, only to flood again the following year. Fishes that made their living in water could benefit from a temporary ability to survive on land, while they dragged themselves from a shallow lake or pond that was threatened with imminent desiccation to a deeper one in which they could survive until the next wet season. On this view, our ancestors didn't so much emerge on to the dry land as use the dry land as a temporary bridge to escape back into the water. Many modern animals do the same.
Rather unfortunately, Romer introduced his theory with a preamble whose purpose was to show that the
Devonian era was a time of drought. Consequently, when more recent evidence undermined this assumption, it seemed to undermine the whole Romer theory. He'd have done better to omit the preamble, which was, in any case, overkill. As I argued in The Ancestor's Tale, the theory still works, even if the Devonian was less drought-ridden than Romer originally thought.
Let us, in any case, return to the fossils themselves. They trickle sparsely through the late Devonian, the period immediately preceding the Carboniferous: tantalizing traces of 'missing links', animals that went some way towards bridging the gap between the lobe-finned fishes that were so abundant in Devonian seas, and the amphibians that later slithered through the Carboniferous swamps. On the fish side of the gap, Eusthenopteron was discovered in 1881 in a collection of fossils from Canada. It seems to have been a surface-hunting fish and probably didn't ever come on land, notwithstanding some early imaginative reconstructions. Nevertheless, it did have several anatomical similarities to the amphibians of 50 million years later, including its skull bones, its teeth and, above all, its fins. Although they were probably used for swimming and not walking, the bones followed the typical pattern of a tetrapod (the name given to all land vertebrates). In the forelimb, a single humerus was joined to two bones, the radius and ulna, joined to lots of little bones, which we tetrapods would call carpals, metacarpals and fingers. And the hind limb shows a similar tetrapod-like pattern.
Then, near the amphibian side of the gap, some 20 million years later, at the border between the Devonian and Carboniferous, great excitement was caused by the 1932 discovery in Greenland of Ichthyostega. Don't be misled by thoughts of cold and ice, by the way. Greenland in the days of Ichthyostega was on the equator. Ichthyostega was first reconstructed by the Swedish palaeontologist Erik Jarvik in 1955, and again he portrayed it as closer to a land-dweller than modern experts do. The most recent reconstruction, by Per Ahlberg at Jarvik's old university of Uppsala, places Ichthyostega mostly in the water, although it probably made occasional forays on to the land. Nevertheless, it looked more like a giant salamander than a fish, and it had the flat head that is so characteristic of amphibians. Unlike all modern tetrapods, which have five fingers and toes (at least in the embryo, although they may lose some in the adult), Ichthyostega had seven toes. It seems that the early tetrapods enjoyed more freedom to 'experiment' with varying numbers of digits than we have today. Presumably at some point the embryological processes fixed upon five, and a step was taken that was hard to reverse. Although, admittedly, not as hard as all that. There are individual cats, and indeed humans, who have six toes. These extra toes probably arise through a duplication error in embryology.
Another exciting discovery, also from tropical Greenland and also dating from the boundary between the Devonian and the Carboniferous, was Acanthostega. Acanthostega, too, had a flat, amphibian skull and tetrapod-like limbs; but it too departed, and even further than Ichthyostega, from what we now think of as the fivefinger standard. It had eight digits. The scientists most responsible for our knowledge of it, Jenny Clack and Michael Coates of Cambridge University, believe that, like Ichthyostega, Acanthostega was largely a water-dweller, but it had lungs and its limbs strongly suggest that it could cope with land as well as water if it had to. Again, it looked pretty much like a giant salamander. Moving back now to the fish side of the divide, Panderichthys, also from the late Devonian, is also slightly more amphibian-like, and slightly less fish-like, than Eusthenopteron. But if you saw it you would surely want to call it a fish rather than a salamander.
So, we are left with a gap between Panderichthys, the amphibian-like fish, and Acanthostega, the fishlike amphibian. Where is the 'missing link' between them? A team of scientists from the University of Pennsylvania, including Neil Shubin and Edward Daeschler, set out to find it. Shubin made their quest the basis for a delightful series of reflections on human evolution in his book Your Inner Fish. They deliberately thought about where might be the best place to look, and carefully chose a rocky area of exactly the right late Devonian age in the Canadian Arctic. There they went - and struck zoological gold. Tiktaalik! A name never to be forgotten. It comes from an Inuit word for a large freshwater fish. As for the specific name, roseae, let me tell a cautionary tale against myself. When I first heard the name, and saw photographs like the one reproduced on colour page 10, my mind immediately leapt to the Devonian, the 'Old Red Sandstone', the colour of the eponymous county of Devon, the colour of Petra ('A rose-red city, half as old as time'). Alas, I was quite wrong. The photograph exaggerates the rosy glow. The name was chosen in honour of a benefactor who helped finance the expedition to the Arctic Devonian. I was privileged to be shown Tiktaalik roseae by Dr Daeschler when I had lunch with him in Philadelphia, shortly after its discovery, and the lifelong zoologist in me - or perhaps my inner fish - was moved to speechlessness. Through rose-tinted spectacles I imagined I was gazing upon the face of my direct ancestor. Unrealistic as that was, this not-so-rose-red fossil was probably as close as I was going to get to meeting a real dead ancestor half as old as time.
If you were to meet a real live Tiktaalik, snout to snout, you might start back as if threatened by a crocodile, for that is what its face resembled. A crocodile's head on a salamander's trunk, attached to a fish's rear end and tail. Unlike any fish, Tiktaalik had a neck. It could turn its head. In almost every particular, Tiktaalik is the perfect missing link - perfect, because it almost exactly splits the difference between fish and amphibian, and perfect because it is missing no longer. We have the fossil. You can see it, touch it, try to appreciate the age of it - and fail.
Was this article helpful?