by Brian Switek
There were many types of marine reptiles in the past, all of which evolved from terrestrial ancestors, but they were shaped in different ways by historical contingencies. The long- and short-necked plesiosaurs, the lizardlike mosasaurs, the barrel-bodied placodonts, and sea turtles all became independently adapted to marine life. One particular group, however, looked like they could have been an early version of the dolphin prototype: the ichthyosaurs. Their long snouts and streamlined bodies made them look superficially similar to living marine mammals, but the convergence was not perfect.
One of the most obvious differences between dolphins and ichthyosaurs is that dolphins have flukes that are held horizontally while ichthyosaurs had caudal fins oriented vertically, meaning that dolphins swim by beating their tails up and down and ichthyosaurs by moving their tails from side to side. These differences have everything to do with how the early terrestrial ancestors of both groups moved. The ancestors of ichthyosaurs were reptiles that had a sprawling posture in which their bent legs were held out to the side. This means that when they walked they moved in a side-to-side motion and upon entering the water used their legs and tail in a similar way to swim. Eventually the tail became broader and flattened for propulsion, and their limbs became more useful for steering than swimming. This is just the type of condition seen in the “proto-ichthyosaurs” called icthyopterygians. The terrestrial ancestors of these animals are still unknown, but like Ambulocetus the ichthyopterygians are intermediate forms that represent the early stages of life in the water.
Small, early Triassic members of this group, like Utatsusaurus and Chaohusaurus had long bodies, stubby limbs, no dorsal fin, and just a bump for a caudal fin. This is known from body impressions of these animals from China and British Columbia, and they were similar in form to dogfish sharks. They probably swam much like eels do, and they may have been ambush predators that relied on quick bursts of speed to catch prey, as they would not be able to sustain swimming at high speeds for very long. From these creatures the first true ichthyosaurs evolved, typified by forms like the miniscule Mixosaurus and the enormous Cymbospondylus. These genera had even larger fins than their predecessors and swimming abilities more comparable to modern requiem sharks, like the tiger shark.
By the late Triassic ichthyosaurs had become more speedy and streamlined than any of their predecessors. These tunalike ichthyosaurs diversified and spread all over the globe, and the Jurassic genera Opthalmosaurus and Ichthyosaurus represent the classic long-snouted, torpedo-bodied type. They moved through the water by moving their tail fins from side to side as their ancestors did, but they had more developed half-moon-shaped caudal fins akin to those seen in fast-moving fish like mako sharks, marlin, and tuna.
The reason ichthyosaurs evolved this way was determined by their ancestry. The way their forerunners moved on land was adapted for life in the water; what already existed was co-opted for new uses. The different modes of locomotion exhibited by the paddling plesiosaurs and mosasaurs illustrate other evolutionary alternatives, but the way all these reptiles swam was constrained by attributes of their ancestors. The same is true of cetaceans, and the reason for their distinctive method of swimming is just as fortuitous.
The ancestors of the early archaeocetes were terrestrial artiodactyls that carried their legs underneath their body. When they walked or ran their spine would have moved up and down, as in other quadrupedal mammals, and this simple fact of mammalian locomotion constrained the way the early ancestors of whales could have swum. When Pakicetus entered the water, with its long legs and relatively inflexible spine, it could not have swum like a fish or a lizard; it would have to have done the doggie paddle to get around.
Ambulocetus was a more proficient swimmer. Its expanded hands and feet were almost certainly webbed, and it probably swam by undulating its spine to a limited degree and paddling with its feet. For the earliest archaeocetes, each swimming stroke would have two parts, a power stroke that propelled them through the water and a recovery phase in which the limb was brought back into position for the next stroke.55 This type of swimming does not allow for quick pursuit of prey, but by undulating its spine Ambulocetus could have kept moving even during the recovery phase of a stroke. Living otters provide a good model for this type of swimming, as they use both the sinuous undulations of their spines and their limbs to move through the water.
Ambulocetus was probably more reliant on paddling than movements of its spine, though, because of a feature inherited from its land-dwelling ancestors. In mammals, the area of the spine that is connected to the hip is called the sacrum, and it consists of vertebrae that have fused together to give strength to the hips. For archaeocetes beginning to undulate their spines to move through the water, the sacrum limited their spine’s range of motion. Increasing flexibility in this part of the spine would have been an advantage. A greater range of up and down motion, and thus more powerful propulsion, could be achieved by unfusing the sacral vertebrae.
This vertebral adaptation is exemplified by Rodhocetus. The unfused vertebrae of its spine even functionally resemble those of the tail, indicating their role in locomotion. Rodhocetus still had its sacral vertebrae fused to its hips, however, and this means that archaeocetes began undulating their spines to move through the water before they stopped using their limbs to propel themselves. Indeed, Rodhocetus and related genera like Maiacetus still had large, paddle-shaped feet, but another relative named Georgiacetus may reveal how early whales began using vertebral undulation more than paddling.
FIGURE 61 - The partially restored skeleton of Rodhocetus (the limbs were incomplete in the original description).
FIGURE 62 - The sacral vertebrae of Rodhocetus . Normally these vertebrae are fused in mammals, but in Rodhocetus they were unfused to allow greater flexibility of the spine.
Georgiacetus had relatively large paddle-shaped rear limbs, but its spine was disconnected from its hips. Without the spine providing an anchor the hind limbs might not have been useful in swimming, but they could have worked as rudders to help direct the whale through the water. The fact that Georgiacetus and the bones of related archaeocetes have been found in the United States, thousands of miles from the epicenter of early whale evolution in Pakistan, indicates that it was a proficient enough swimmer to cross oceans, and it sat close to the transition from the protocetid-type whales to the oceangoing basilosaurids. By the time these whales evolved, the sacral vertebrae were entirely unfused, the hips were completely disconnected from the spine, and the hind limbs had become so reduced that they may have been functionless. 56 Ultimately cetaceans lost all external remnants of the hind limbs, probably because the heavy vestigial structures increased drag. Any such encumbrances would have been detrimental to predators that relied on speed to chase after their prey.57
The adaptation of early whales to the water also involved more subtle changes. Buoyancy is a major problem for terrestrial animals in the water; it is difficult to submerge or stay submerged with light bones filled with spongy tissue. Light bones require an animal to exert a lot of energy to stay under water, and this makes it very difficult to swim. In order for archaeocetes to fully adapt to aquatic life they would have to become neutrally buoyant, so that they would neither sink nor float when at rest in the water.
Some animals achieve neutral buoyancy as a result of swallowing stones (called gastroliths once they enter the stomach) that are the functional equivalent of ballast on a ship. Living crocodiles are among the animals that benefit from gastroliths in this way, but archaeocetes became neutrally buoyant due to changes in their bones. The cavities in limb and rib bones of terrestrial mammals are typically filled with marrow, but in many archaeocetes the cavities are filled with trabecular bone. Trabecular bone itself is not as dense as other types of bone, but this in-filling of bone cavities that otherwise would be empty made the bones as a whole denser and functioned as a kind of bone ballast.58
The ribs of Indohyus and Pakicetus show just this kind of bone restruct
uring. The development of bone ballast preceded other aquatic adaptations, but it had a trade-off. As the bones in the limbs and ribs of these animals became denser they would have become more brittle. Running on land with heavy, brittle bones would have been more energetically expensive and even risky, thus providing another reason for these animals to spend more time in the water. The amount of hypermineralization seen in the bones of Pakicetus suggests that it was doing just that.
The bones of later archaeocetes, like Ambulocetus and Kutchicetus, were even denser, and the heavy bone ballast in archaeocetes like Rodhocetus allowed them to devote more energy to swimming. Eventually, though, this trend would be reversed. Georgiacetus had a lighter bone density and marked the beginning of an osteoporotic trend, in which bone is thinner and lighter. This was carried even further in later marine whales like Dorudon, whose bone density more closely corresponded to that of living cetaceans. The reason for this reversal is attributable to life at sea. An extra heavy skeleton is a disadvantage during deep dives, where a large amount of energy is expended in search of food. The denser the skeleton a whale had, the more energy it would take to return to the surface before running out of breath. Making the skeleton lighter would allow whales to roam into new niches, not only feeding at the surface but also deeper down. Thus, natural selection favored a lightening of the skeleton.
Thus the skeletons of extinct whales have filled in our understanding of how whales became adapted to life at sea, but not all the evidence comes from the fossil record. If living cetaceans are truly the descendants of terrestrial ancestors, then they should retain some vestiges of their ancestry. This is most certainly the case.
As with ichthyosaurs, the pectoral flippers of dolphins were adapted from arms and hands that were once used for terrestrial locomotion. Many cetaceans have additional bones in some of their fingers (primarily the second and third fingers) that lengthen their pectoral paddles, but the bones within their flippers are only modified versions of the front limbs of their terrestrial ancestors. Living whales retain vestiges of their hips and hind limbs, too. In cetaceans, the muscles of the tail important to propulsion connect to muscles in the abdomen, but the vestiges of the pelvis function as an anchor for muscles that attach to reproductive organs (an anatomical feature that is seen in other mammals, including artiodactyls). In this case the altered and reduced hips retain some of their original function, but some whales have exhibited archaic traits that are of no use whatsoever. In 1881, John Struthers published a paper called “On the Bones, Articulations, and Muscles of the Rudimentary Hind-Limb of the Greenland Right-Whale (Balaena mysticetus),” in which he discussed the vestigial hips, femur, and tibia found inside a right whale. The rudimentary leg bones were cartilaginous (which is why you don’t usually see them in museum mounts) and appeared to have no function.
Why should these traits be preserved or even reappear if whales lost their external limbs long ago? While the morphology of whales has certainly changed, some of the genes that code for lost traits have been retained. This is because there is no single gene that could have been eliminated that controls the formation of legs. Instead, there are many genes involved in limb formation, some of which are involved in the development of other parts of the body. The multifunctionality of these genes helps to explain their endurance as well as why, for a short time, whales begin to grow legs in the womb.
At twenty-four days old, the embryos of spotted dolphins have both fore- and hind limb buds. By forty-eight days, however, the forelimb is well developed, with the beginnings of digits, but the hind limb bud is little more than a speck. It is resorbed into the body. What this suggests is that after whales started propelling themselves by oscillations of the tail and external hind limbs began to be selected against, there was a relatively sudden change in development that caused the cessation of growth and resorbtion of those limbs. Some of the genetic instructions for limb development remained, but they have been prevented from progressing far enough to create external hind limbs. An exception is a bottlenose dolphin with external pelvic fins described in 2006. It did not have fully developed hind limbs, but it illustrates how genetic instructions for “long lost” traits can be preserved and expressed due to changes in development.
In a similar fashion, baleen whales develop the beginnings of teeth only to have them resorbed during their early development. Even though adult baleen whales do not have any teeth, they still retain some of the genetic instructions for their formation. As we all learn in elementary biology, in DNA genes are made up of three-paired combinations of four letters, A, T, G, and C. Combinations of these letters have different functions, from coding for a protein to turning other genes on or off, but not all of the genome of any given organism is functional. If there is a mutation where a new letter is inserted into the code of a gene, it might create a stop codon, or a genetic stop sign that prevents that gene from being fully expressed. These stop codons might have different effects based upon where they occur along a gene, but when a gene loses its function it may pick up more and more mutations, becoming a “fossil gene” that slowly degrades away. In the case of living mysticetes, Thomas Deméré and colleagues found that they possessed two such fossil genes related to tooth production, called AMBN and ENAM. Both contained stop codons that prevented their expression. Even though mysticetes have not had teeth for millions of years, the genes for creating teeth are still there, slowly being eroded away by mutations.
As strange as it might seem, there once was a time when there were baleen whales with teeth. Some of them, like the recently discovered Janjucetus , superficially resembled Basilosaurus more than a humpback whale. These frightening predators were the early relatives of today’s filter-feeding giants, and the clues to this connection can be seen in the skull.
As odontocetes and mysticetes evolved, their skulls became longer, and certain bones became elongated in a trend called “telescoping.” This is what caused the nasal openings of the whales to move from the front of the skull back over the eyes. If we were to line up a selected group of fossil whale skulls in the order in which the species appeared, from oldest to youngest, we would see that the early forms had a relatively short premaxilla and maxilla, the bones of the upper jaw that hold the teeth. Right behind the maxilla are the nasal bones, or the bones surrounding the nasal opening. Moving from the skulls of geologically older cetaceans to younger ones, the maxilla becomes longer and longer, pushing the nasal bones back on the skull until they are sandwiched together with other parts of the skull over the eyes.
FIGURE 63 - A simplified evolutionary tree of whales over the past fifty-five million years. While archaeocetes persisted until about twenty-six million years ago, by that time the first representatives of modern whales (the odontocetes and mysticetes) had appeared.
The maxilla is important for another reason. In baleen whales the maxilla has extended so far back that part of it scoops downward and backward under the eye socket. This is not seen in odontocetes. It is a characteristic feature of baleen whales. The skull of Aetiocetus, a baleen whale with teeth, shows just this condition where the maxilla extends backward to make up part of the margin of the eye. Stranger still, Aetiocetus probably had baleen.
Modern mysticete whales do not have teeth. Instead, they sift their food from the sea with a comb of hairlike structures that sprout out of plates attached to their jaws. In the past, however, the early relatives of modern baleen whales had teeth, and when it was first discovered Aetiocetus was an important confirmation that modern baleen whales evolved from toothed ancestors. What has only more recently come to light, however, is that baleen evolved before the mysticetes fully lost their teeth. The hairlike, keratinous baleen of whales hangs from a set of plates that attach to the upper jaw, and these plates are nourished by blood vessels. The blood vessels come through a series of holes, or foramina, in the roof of the mouth, and Aetiocetus has the same telltale foramina seen only in modern baleen whales. How the baleen helped Aetiocetus feed is still unk
nown. (Perhaps it was useful in filter feeding, or straining prey from mud stirred up from the sea bottom.) But the unexpected discovery of an ancient whale with both teeth and baleen is a stunning confirmation of evolution.
FIGURE 64 - The underside of the skull of the fossil baleen whale Aetiocetus weltoni (left) and a close-up of part of the jaw (right) showing foramina that would have nourished baleen with blood.
There are some traits of modern cetaceans that were not direct adaptations to life in the sea, like the increases in brain size seen in dolphins thirty-four and fifteen million years ago (most likely a result of changes in the social lives of dolphins). But the general whale body form can be understood as an extreme modification of an ancestral hoofed mammal type to aquatic life. This change happened extremely fast, only ten million years separate the wolflike Pakicetus and the fully aquatic Basilosaurus, but fortunately the fossil record has preserved an extraordinary diversity of fossil whales. A confluence of fossil and genetic data has allowed us to understand a transition that was once mysterious, but there are still things to learn about lineages that were thought to be well known.
Behemoth
“Behold now behemoth, which I made with thee; he eateth grass as an ox. Lo now, his strength is in his loins, and his force is in the navel of his belly. He moveth his tail like a cedar: the sinews of his stones are wrapped together. His bones are as strong pieces of brass; his bones are like bars of iron. He is the chief of the ways of God: he that made him can make his sword to approach unto him.”
—JOB 40:15-19 (King James Version)
“It appears somewhat extraordinary, at the first view, that we should discover manifest proofs of there having existed animals of which we can form no adequate idea, and which in size must have far exceeded any thing now known upon the earth; and those too, in climates where elephant (the largest animal now in existence) is never found.”