What possessed fish to get out of the water or live in the margins? Think of this: virtually every fish swimming in these 375-million-year-old streams was a predator of some kind. Some were up to sixteen feet long, almost twice the size of the largest Tiktaalik. The most common fish species we find alongside Tiktaalik is seven feet long and has a head as wide as a basketball. The teeth are barbs the size of railroad spikes. Would you want to swim in these ancient streams?
NEIL SHUBIN, YOUR INNER FISH
FROM WATER TO LAND
Ever since Charles Darwin published On the Origin of Species in 1859, scientists have sought fossils that show how one crucial evolutionary transition had taken place: how fish crawled out of the water and became land-living creatures. Of course, an entire class of vertebrates, the Amphibia, are still living in that transition. Some of them spend nearly all their time in the water and rarely go out on land. Others never enter the water at all, but must live in moist habitats. Many have a mixture of the two lives.
Even before the publication of Darwin’s book, some scientists noticed the similarities between amphibians and lungfish, which show many amphibian-like features (especially the lungs), but still are fish with fins. Yet the fins of lungfish and other lobe-finned fish have the same bones as those in the limbs of amphibians. But even that was not so clear-cut. The South American lungfish (
Lepidosiren paradoxa) is so specialized that it has only tiny ribbon-like fins and swims like an eel. When it was discovered in 1837, it was thought to be a degenerate amphibian. Almost the same thing happened when Richard Owen described the African lungfish (
Protopterus) in 1839. A staunch opponent of evolution, Owen ignored the obvious connections between the anatomy of lungfish and amphibians, and emphasized their bizarre specializations, such as the tiny ribbon-like fins. Only when the Australian lungfish (
Neoceratodus forsteri) was discovered in 1870 was it possible to see that some living lungfish have robust lobed fins that have all the same bones as the amphibian limb. This was further confirmed when more and more primitive lungfish fossils showed that most of the lungfish had many amphibian-like features (see
figure 8.2), not the bizarre specializations of the African and South American lungfish.
Still, the gap between lungfish and the earliest amphibians in the fossil record was a large and frustrating one. In 1881, Joseph F. Whiteaves described
Eusthenopteron foordi, probably one of the best transitional fossils. Unfortunately, his description was only two paragraphs, had no illustrations, and made no mention of how this fish showed amphibian-like features.
Eusthenopteron was a large (up to 1.8 meters [6 feet] long) lobe-finned fish that was much more amphibian-like than either extant lungfish or coelacanths (
figure 10.1). It is known from hundreds of beautiful specimens from a famous locality near Miguasha, on Scaumenac Bay, Quebec. Although
Eusthenopteron still had a fish-like body, its lobed fins had all the right bones from which to build the amphibian hand and foot, and its skull had the right pattern of bones to be ancestral to the amphibian skull.
More discoveries of fossils showed that many lungfish and other lobed-fin fish had lived in the Late Devonian (385 to 355 million years ago). By the Early Carboniferous (355 to 331 million years ago), there had been a handful of unquestioned amphibians (in the nineteenth century, called by the now-obsolete names “stegocephalians” and “labyrinthodonts”), although their fossils are much more abundant in rocks of the Late Carboniferous. So where were the transitional fossils? Many Late Devonian localities with fossils of marine fish were found, but few that seemed to be from freshwater and that had much potential for yielding a fossil on the cusp between fish and amphibian.
The breakthrough came through accident and political expediency. In the 1920s, Norway and Denmark were arguing over which country owned East Greenland. Consequently, the Danish government and a foundation established by Carlsberg Brewery (the famous Danish beer maker) funded a three-year expedition to East Greenland that visited the gigantic island in the summers of 1931 to 1933. The members of the expedition hoped to conduct enough scientific research and exploration in East Greenland that Danish territorial rights would be recognized, since Norway had done no exploration there. It was led by the famous Danish geologist and explorer Lauge Koch and featured an all-star cast of Danish and Swedish geologists, geographers, archeologists, zoologists, and botanists.
Figure 10.1
Comparison of the skeletal elements of Ichthyostega and Eusthenopteron. (Drawing by Carl Buell; from Donald R. Prothero, Evolution: What the Fossils Say and Why It Matters [New York: Columbia University Press, 2007], fig. 10.5)
Among the scientists recruited to explore East Greenland was Gunnar Säve-Söderbergh, a Swedish paleontologist and geologist. He had been trained at the University of Uppsala and eventually became a professor of geology there. Only 21 years old at the time he joined the first expedition, Säve-Söderbergh soon found fossils of some remarkable creatures, which he named
Ichthyostega and
Acanthostega, as well as more primitive lobe-finned fish like
Osteolepis, which was much like
Eusthenopteron, as well as many lungfish. All apparently had swum in the same fresh- or brackish waters when East Greenland was near the tropics and the Devonian Age of Fishes was winding to a close (
chapter 8). Through the 1920s and early 1930s, Säve-Söderbergh published short descriptions of these fossils, intending to do a much more detailed analysis later. However, that chance never came, though, because he died of tuberculosis in 1948, at the relatively young age of 38.
Säve-Söderbergh was part of a larger tradition in Sweden of studying early fossil fish. Because the Swedes had mounted polar expeditions to Greenland, Spitsbergen, and elsewhere that had discovered many fossil fish, they soon became a Swedish specialty. The founder of the “Stockholm school” of paleontology (based largely at the Swedish Museum of Natural History) was the venerable Erik Stensiö, who was famous for his detailed studies of armored jawless fish from the Devonian. He had so many good specimens at his disposal that he cut some of them into thin slices (serial sectioning) so he could examine the details of the nerves, blood vessels, and other internal anatomy that are normally invisible in description of fish fossils. Today, high-resolution X-ray computed tomography allows paleontologists to make a “CAT-scan” of a solid fossil without slicing it up and destroying it for other uses.
After Säve-Söderbergh’s death, his Greenland fossils were studied by Stensiö’s successor, Erik Jarvik. He had accompanied Säve-Söderbergh on some of the later trips to Greenland, and then returned to collect more fossils. Jarvik was a careful, methodical worker, never one to rush to publish. He spent years slicing up specimens of
Eusthenopteron to see the details of the internal anatomy of its skull. He worked on Säve-Söderbergh’s
Ichthyostega fossils for
50 years, finally releasing his detailed publication about them in 1996, when he was 89 years old! The profession of vertebrate paleontology is legendary for scientists sitting on important fossils for years without publishing anything for the rest of us to see, but Jarvik takes the cake as one of the slowest workers of all. Although Jarvik’s research was important and his descriptive work was impressive, he proposed many odd notions about different fossil groups that no other paleontologists considered to be plausible. He died in 1998, at the ripe old age of 91.
Since Jarvik’s complete description of
Ichthyostega did not appear until 1996, Säve-Söderbergh’s original reconstructions of the fossils were the only well-documented “fishibian” from the 1920s until the 1980s. Thus
Ichthyostega became the archetypal transitional fossil between
Eusthenopteron and early amphibians (see
figure 10.1). Like amphibians, it had four legs with toes, rather than the lobed fins of its ancestors. However, its forelimbs were not strong enough to do much walking, and the most recent analyses suggest that it could move only by short hops, dragging its more flipper-like hind limbs behind. The forelimbs and, especially, the hind limbs were much better adapted for use in the water, where they propelled the animal along (as newts and salamanders swim).
Ichthyostega had robust ribs with flanges that would help support its chest cavity and lungs out of the water, but they were not capable of the rib-assisted breathing found in many amphibians. The other amphibian-like feature was its long flat snout with eyes directed upward and its short braincase;
Eusthenopteron had a more fish-like cylindrical skull, with a short snout and a long braincase, eyes facing sideways, and big gill covers. Other than the limbs and the bones of its shoulder and hips, however,
Ichthyostega was really fish-like. It still had a large tail fin, as well as many fishy features of the skull, such as large gill covers, hearing adapted for water, and a lateral-line system (canals on the face used to sense motion and currents in the water).
In the 1980s, the locus of research on “fishibians” shifted from Sweden to Cambridge University, where Jenny Clack, Per Ahlberg, Michael Coates, and others were active in collecting more fossils and redoing the work of the “Stockholm school” paleontologists. As Clack describes it:
In 1985, I began to think about the possibility of an expedition to East Greenland, at the instigation of my husband Rob. Along the trail, I met Peter Friend of the Earth Sciences Department across the road in Cambridge, who had been leader of several expeditions to the part of Greenland in which I was interested. It turned out that he’d had a student, John Nicholson, who’d collected a few fossils as part of his thesis work on the sediments of the Upper Devonian of East Greenland between 1968 and 1970. Peter retrieved these specimens from a basement drawer and also showed me John’s notebook from his 1970 expedition. John’s note that on Stensiö Bjerg, at 800 metres [2625 feet],
Ichthyostega skull bones were common was startling, and portentous. The fossils that he’d collected fitted together to make a single small block of three partial skulls and shoulder girdle bits—not of
Ichthyostega, but of its at that time lesser known contemporary,
Acanthostega. Peter suggested I get in touch with Svend Bendix-Almgreen, Curator of Vertebrate Palaeontology in the Geological Museum in Copenhagen. The Danes still administered expeditions by geologists to the National Park of East Greenland, where the Devonian sites are located, so he would be the person to start with in my attempts to mount an expedition there. Peter also suggested I contact Niels Henricksen of the Greenland Geological Survey (GGU). By sheer coincidence, and great good fortune, the GGU had a project in hand in the very place where I needed to go, and their last season there was the summer of 1987. With funds from the University Museum of Zoology and the Hans Gadow Fund in Cambridge and the Carlsberg Foundation in Copenhagen, I, my husband Rob, my student at the time, Per Ahlberg, and Svend Bendix-Almgreen and his student Birger Jorgenson arranged a six-week field trip in the care of the GGU for July and August of 1987. Using John Nicholson’s field notes, we eventually pinned down the locality from which the
Acanthostega specimens had come, and then the exact in-situ horizon that had been yielding them. It was in effect, a tiny, but very rich,
Acanthostega “quarry.”
The discovery of much more complete specimens of
Acanthostega was a big breakthrough. In 1952, Jarvik named
Acanthostega, based on poor material that received little study. But all the new fossils that Clack and her group collected in the late 1980s and the 1990s made
Acanthostega much more complete and informative than the original
Ichthyostega material (
figure 10.2). In most respects, the smaller
Acanthostega was much more fish-like than
Ichthyostega. Unlike those of
Ichthyostega, the limbs of
Acanthostega would not have allowed it to crawl on land—it lacked wrists, elbows, or knees. Instead, its limbs were only capable of only paddling and pulling it through obstacles underwater. Even more surprising, it had as many as seven or eight fingers on its hands, not the standard five fingers that most vertebrates have!
Acanthostega had a much larger fin on its tail than did
Ichthyostega, and its ribs were too short to support its body on land and allow it to breathe without the support of water. Yet it also had a few advanced amphibian-like features: its ear could hear in air as well as in water, and it had strong bones in its shoulder and hip region, four limbs with toes, and a neck joint that allowed it to rotate its head. By contrast, a fish has no “neck” that allows rotation—it must turn the entire front half of its body to change direction or snap at prey.
Figure 10.2
Comparison of the skeletons of Ichthyostega (top) and Acanthostega (bottom). (Drawing courtesy M. Coates, based on research by M. Coates and J. Clack)
YOUR INNER FISH
Jenny Clack’s work revitalized the research on the fish–amphibian transition, and soon many other paleontologist were getting into the act. One of them was an eager and enthusiastic young scientist named Neil Shubin. He was educated in paleontology as an undergraduate at Columbia University and the American Museum of Natural History in New York, where I was a graduate student at the time. There we met in 1980, and together we worked on the evolution of the horse
Mesohippus, his first research project that was published. He went on to earn a doctorate at Harvard, studying the evolutionary and developmental mechanisms that dictate how amphibian limbs and toes form. His first job was teaching anatomy to medical students at the University of Pennsylvania in Philadelphia, where he hooked up with Ted Daeschler of the Academy of Natural Sciences. Together, they searched road cuts of Devonian red beds across Pennsylvania until they found some incomplete fossils of fish and “fishibians.”
But Shubin was looking for bigger fish to fry. As he describes in his book Your Inner Fish, he and Daeschler knew that they had to find rocks older than 363 million years (such as the East Greenland rocks that had yielded Ichthyostega and Acanthostega), but younger than 390 to 380 million years (from which have been recovered most of the lobe-finned fish that are ancestral to amphibians). Shubin and Daeschler predicted that there should be transitional fossils more primitive than Acanthostega but more advanced than Eusthenopteron in Upper Devonian freshwater deposits that filled the gap between 380 and 363 million years ago. They looked at the geologic maps in the first edition of the legendary historical geology textbook Evolution of the Earth (1971) by Robert H. Dott Jr. and Roger Batten. When they studied the map of Upper Devonian outcrops, they saw three likely candidates: eastern Pennsylvania (where they were already working), East Greenland (already collected by the Danes and Swedes and by Clack’s group), and Ellesmere Island in the Canadian Arctic (which no one had studied). Further study of published geological survey reports showed that these outcrops were Upper Devonian, between 380 and 363 million years in age, and the right rock type to preserve freshwater fish and amphibian fossils. These rocks turned out to be about 375 million years old.
By the late 1990s, Shubin and Daeschler and their crew had all the permits and equipment, as well as funding for supplies and helicopter time to take them into and out of the region. Running a major expedition to this harsh region is no picnic! Researchers need a full complement of Arctic gear, especially cold-weather clothing for protection against the freezing summer temperatures and rugged tents that can stand up to hurricane-force winds and provide warmth and shelter during the frequent storms. In addition to rock picks, shovels, and other standard tools for collecting, they carried rifles because polar bears were a serious threat.
Starting in 2000, they made short trips of a few weeks at the peak of the summer to Ellesmere Island, with poor results in the first few years because the rocks were marine, not freshwater, in origin. Finally, they found the freshwater fossiliferous rocks they had been seeking. In 2000, they found what they called Bird Quarry, which by 2003 had yielded abundant fragmentary fish fossils. In 2004, they dug 3 meters (10 feet) below the surface level of the quarry and discovered
Tiktaalik, the fossil that made all the hardships worthwhile. Shubin and his colleagues picked the name
Tiktaalik, which in Inuktitut, one of the Inuit languages, means “burbot,” a freshwater fish of the region. It took two more years before the fossils were properly prepared for study, and all the descriptions and analyses were ready, so
Tiktaalik was announced in two papers published in 2006, with the description of the hind limbs appearing in 2012.
More than 10 individuals of
Tiktaalik have been recovered, ranging in length from 1 to 3 meters (3.3 to 10 feet) (
figure 10.3). Even better, the best specimen of
Tiktaalik is nearly complete, with just portions of the hind limbs and tail missing, although the hind limbs are known from other specimens. As one would expect for a specimen that is 12 million years older than
Ichthyostega or
Acanthostega, Tiktaalik is more fish-like in many ways. Its lobed fins had all the elements ancestral to the amphibian limb, but still had fin rays, rather than toes. It had fish-like scales, a combination (as do most of the “fishibians”) of both gills (shown by the gill-arch bones) and lungs (shown by the spiracles in its head), and a fish-like lower jaw and palate. But unlike any fish, it had amphibian features, too: a shortened, flattened skull with a mobile neck; notches in the back edge of the skull for the eardrums on the back of the skull; and robust ribs and limbs and shoulder and hip bones. Like
Acanthostega, its fins were not strong enough or flexible enough to allow it to drag itself across land for very far or walk with its belly off the ground; instead, they were probably used to paddle in shallow water and to support the animal so it could see above the surface. Like the other “fishibians” (and many modern amphibians, especially newts and salamanders), it probably spent most of its time in water, hunting on the margins of the streams in which it lived.
As Robert Holmes wrote in New Scientist:
After five years of digging on Ellesmere Island, in the far north of Nunavut, they hit pay dirt: a collection of several fish so beautifully preserved that their skeletons were still intact. As Shubin’s team studied the species they saw to their excitement that it was exactly the missing intermediate they were looking for. “We found something that really split the difference right down the middle,” says Daeschler.
Figure 10.3
Tiktaalik: (A) skeleton; (B) reconstruction of its appearance in life. (Courtesy N. Shubin)
And Clack commented, “It’s one of those things you can point to and say, ‘I told you this would exist,’ and there it is.”
The search for even more transitional fossils continues. But one thing is clear: making the transition from water to land is not the gigantic leap that paleontologists and biologists thought it was for more than a century. You need look no further than the huge radiation of the ray-finned fish (Actinopterygii), which include 99 percent of the fish in fish tanks, fish markets, and big aquariums. Except for lampreys, hagfish, sharks, rays, lungfish, and coelacanths, all extant fish are ray-finned fish. They do not have the robust bones of the lobe-finned fish, but long thin rods of bone or cartilage to support their fins.
Ray-finned fish have found a number of ways to use their flimsy fins to move about on land. For example, mudskippers live half in and half out of the water, propped up in hallow mudflats or mangrove roots and using their front fins to crawl slowly on the air–water interface (
figure 10.4). The “walking catfish” is a major pest in the southeastern United States because it can wriggle across land from one pond to another to find food or escape from a drying pool. The climbing perch can also drag itself across land in search of better pools and can even crawl up trees. Many fish, such as gobies and sculpins, adapted for tide-pool life spend part of their time in the air during low tide, and have modified their front fins for crawling along and for pushing up against rocks. Other mostly aquatic fish have modified their front-fin rays into “fingers” that can be used to dig into the surface underwater and pull the fish forward.
None of these groups of ray-finned fish are closely related to one another, so all these adaptations for land life evolved completely independently. Clearly, there are strong pressures and big advantages for fish to exploit land habitats (even if for only minutes to hours), and they have found different solutions to what was once thought to be an insoluble problem. Thus the gradual changes in lobe-finned fish to become first semi-aquatic and then fully terrestrial animals are not the near-impossibility that scientists once imagined.
Figure 10.4
Mudskipper feeding on worms on a mudflat in Japan. (Photograph by Alpsdake; from Wikimedia Commons)
Figure 10.5
The evolution of amphibians from fish. (Drawing by Carl Buell; from Donald R. Prothero, Evolution: What the Fossils Say and Why It Matters [New York: Columbia University Press, 2007], fig. 10.6)
Recently, a group of scientists led by Emily Standen published a study that showed just how easy it is for a fish to leave the water. Their experiment focused on a very primitive bony fish, the bichir (Polypterus) of Africa, which is distantly related to such primitive ray-finned fish as the sturgeon and the paddlefish. Its fins are not unlike those of the earliest lobe-finned fish, and thus it is almost like a link between lobe-finned and ray-finned fish. The researchers raised bichirs on land, rather than in their normal watery habitat (they are good air breathers). Sure enough, after a few generations of breeding, their fins became more robust and better suited for crawling on land through a mechanism called developmental plasticity, which allows animal bodies to modify themselves during embryonic development to adapt to new challenges. As Standen pointed out, developmental plasticity may explain not only why so many kinds of ray-finned fish have adapted to crawling on land or in water, but also the mechanisms that allowed lobe-finned fishes to do the same.
Thus we now have a continuous sequence of “fishibians,” from unquestioned fish-like creatures (such as the lobe-finned fish), through intermediates like
Tiktaalik and
Acanthostega and
Ichthyostega, to animals that are even more amphibian-like (
figure 10.5). Anyone who cannot imagine how fish crawled out of water and became land animals need only look at these incredible fossils to see the answer.
SEE IT FOR YOURSELF!
To my knowledge, fossils of Ichthyostega and Acanthostega are housed in only the University Museum of Zoology, Cambridge University, and the Naturhistoriska riksmuseet, in Stockholm, where a few specimens are on display.
Several museums in the United States display replicas of the skeleton and reconstructions of Tiktaalik, including the Academy of Natural Sciences of Drexel University, Philadelphia; Field Museum of Natural History, Chicago; Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts; and Museum of Natural History and Science, Cincinnati. Some of the best displays of lobe-finned fish fossils and early amphibians are at the American Museum of Natural History, New York.
Clack, Jennifer A. Gaining Ground: The Origin and Early Evolution of Tetrapods. Bloomington: Indiana University Press, 2002.
Daeschler, Edward B., Neil H. Shubin, and Farish A. Jenkins Jr. “A Devonian Tetrapod-like Fish and the Evolution of the Tetrapod Body Plan.” Nature, April 6, 2006, 757–773.
Long, John A. The Rise of Fishes: 500 Million Years of Evolution. Baltimore: Johns Hopkins University Press, 2010.
Maisey, John G. Discovering Fossil Fishes. New York: Holt, 1996.
Moy-Thomas, J. A., and R. S. Miles. Palaeozoic Fishes. Philadelphia: Saunders, 1971.
Shubin, Neil. Your Inner Fish: A Journey into the 3.5-Billion-Year History of the Human Body. New York: Vintage, 2008.
Shubin, Neil H., Edward B. Daeschler, and Farish A. Jenkins Jr. “The Pectoral Fin of Tiktaalik roseae and the Origin of the Tetrapod Limb.” Nature, April 6, 2006, 764–771.
Zimmer, Carl. At the Water’s Edge: Macroevolution and the Transformation of Life. New York: Free Press, 1998.
