Spying on Whales

In the past 250 million years, many backboned animals converted from living in terrestrial ecosystems to living in oceanic ones. The first wave happened throughout the time of the dinosaurs, when many different reptile lineages invaded ocean ecosystems from 250 million to 66 million years ago. Since the mass extinction at the end of the Cretaceous, the ecologically dominant ocean invaders have been mammals—including everything from whales to sea otters—although penguins and Galápagos marine iguanas are also more recent reentrants. All of today’s marine mammal lineages are distantly related to one another, whether it’s a whale, a sea otter, a seal, a sea cow, or a polar bear (yes, technically polar bears too, which eat seals and hop across ice-covered seas).

What makes early whale evolution so important is that the completeness of the fossil record from the early stages—Pakicetus, Ambulocetus, Maiacetus, and all others like them during that first phase—is unmatched by any other group in the fossil record. We simply don’t have the range of fossils showing the specific anatomical transformations from land to sea for any other mammal or reptile the way we do with whale origins.

Even so, the evidence for whale origins has only recently been uncovered. Until about forty years ago, we had no idea what the hind limbs of the earliest whales in the first evolutionary phase really looked like. The discovery of Pakicetus in 1981 gave us mostly bones from the neck up—paleontologists discovered a small W-shaped braincase exhibiting, among other features, an involucrum, but it otherwise looked like any other land mammal’s. They found the skull—pinched and delicate, like a handheld vase—in river deposits, and concluded that the earliest whales lived some part of their life on land. Without more of a skeleton, at the time they could only speculate about what these whales looked like from the neck down.

In 1994 the discovery of Ambulocetus clarified this picture, showing that the earliest whales had weight-bearing fore and hind limbs, with separate phalanges perhaps connected in life by webbing. Relatively large feet in Ambulocetus were a clue about its swimming style, which likely involved flexing its spinal column along with its broad feet, in one motion. Mechanically this style is somewhere between paddling with hands and feet (using drag for forward motion) and employing a hydrofoil, as modern whales do with their tail fluke (using lift, instead of drag). Our pelvis is rigidly connected to our backbone, whereas in Maiacetus, the pelvis was only partially connected to the backbone, permitting a lot of flexibility for the whole spinal column to undulate up and down. The shape of a few tail vertebrae can reveal a lot about locomotion—in Ambulocetus the fact that the tail vertebrae are longer than they are tall tells us that these early whales had long, thickened tails, although we still don’t have enough bones to know what direction these powerful tails might have moved.

Ambulocetus still didn’t provide enough evidence to help answer the big questions about whale origins: Where did they fit into the mammalian family tree? Who are their closest relatives? By the 1990s, DNA studies had shown that hippos are the closest living relatives to whales. Hippos and other even-toed hoofed mammals, such as cows, deer, and pigs, are seemingly unlikely relatives, until you look at their stomachs. Even anatomists in the nineteenth century knew that living whales had multichambered stomachs like these ungulates, pointing to a possible evolutionary relationship. Paleontologists, however, had other extinct fossil mammals in the running for whale’s closest relatives: mesonychids, which had strikingly similar teeth and were wholly carnivorous, as whales are today, but left no descendants. Without more skeletal material from four-legged whales, especially from their limbs, there was no way to parse the stories of DNA versus fossils for the deepest origins of whales.

Then, in 2001, two competing groups of paleontologists reported the same pivotal piece of evidence from different species of early whales: they each had discovered that the anklebone of ancient land-dwelling whales was exactly like those of living even-toed ungulates. This bone, called the astragalus, looks like two 35 millimeter film canisters taped together like a raft; in your hand it feels like some kind of board-game piece. Cows, goats, and camels all have it. Living whales don’t because they have no feet, and the only traces of hind limbs are reduced to nubbins of bone next to free-floating pieces of their pelvis, wrapped deeply in their body walls—making fossil hind limbs in early whales the only source for this information. Mesonychids didn’t have these double-pulley anklebones, which meant their tooth similarities with early whales were the result of convergent evolution—something that has happened frequently in mammal evolutionary history. The discovery that early whales had a so-called double-pulley astragalus confirmed the DNA findings: whales were just highly modified even-toed hoofed mammals, minus the hooves.

Since finding Pakicetus, paleontologists working in remote parts of Egypt, Pakistan, and India have discovered a rich variety of early land-dwelling whales that lived about fifty million to forty million years ago, toward the end of a geologic epoch called the Eocene. These early whales seem to have been experimenting with ecological modes that have parts both familiar and strange: Ambulocetus looked crocodile-like; Maiacetus more like a sea lion, which had not yet evolved; still other strange early whales such as Remingtonocetus were an amalgam of zoological categories, something like a long-snouted otter; and Makaracetus, named after a mythological South Asian creature that is half fish, half mammal, had a downturned snout, perhaps for eating clams. All of these early whales belonged to extinct branches at the base of the whale family tree; our expectation about what makes a whale is hindsight biased, based on how we see them today—a great challenge paleontologists face when trying to understand the biology of these extinct whale relatives.

Knowing how whales turned out makes our retelling of their evolutionary pathway a tidy, preordained story. It’s easy to imagine Pakicetus, looking something like a lost dog dipping its toes in the water, followed then by intermediate stages of creatures each spending more time in the water: Ambulocetus, which could hear underwater and lunge at prey with its powerful limbs, like an ambush predator; followed by Maiacetus, whose pelvis was less strongly coupled to its spinal column, permitting the first kind of flexibility for tail-driven propulsion in whales. Fossils belonging to relatives of Maiacetus extend over a far greater geographic range than those of previous ur-whales, suggesting that this still-quadrupedal animal was seafaring, though it still returned to the shore to give birth, like sea lions today. In this view, Maiacetus represented the last of the earliest whales; all subsequent whales, in the second phase of the evolutionary chronicle, had no weight-bearing limbs and were totally separated from land.

The problem with this linear narrative is that we know the final result, which lets us pick and choose the likely path of least resistance toward the whales we recognize today. But evolution doesn’t work like that: it makes no concession for the future; it’s about what’s good enough in the moment. Selection operates on what’s available, sorting biological variation based on the demands of the immediate world. If you were somehow able to return to a late Eocene shoreline in the Tethys sea and happen upon the entire assemblage of early whales in one lineup—all of the early whales, four-legged and odd, scattered on the shoreline—you wouldn’t be able guess the eventual winner of the evolutionary sweepstakes. In its own time and habitat, each early whale was as well adapted as any crocodile, sea lion, or otter living today. It’s just that when we work with the fossil record, we’re afforded a view of the very long run, and the relative successes and failures in any particular group over millions of years. The eventual winners of the evolutionary sweepstakes were early whales that completely severed their ties to land, becoming fully aquatic, eventually yielding descendants that filter feed and echolocate.

These first whales were merely semiaquatic mammals with specializations for life near the water to one degree or another. There was nothing predetermined about some of their descendants becoming fish-shaped leviathans many millions of years later. Retrospection, however, does cue us into specific features that show incremental transformations: shell-shaped ear bones being repurposed for underwater hearing; or the pelvis becoming unlinked from the backbone, allowing the whole back end to serve as a propulsion device. If you focus on tallying which species go extinct and which ones survive, you might lose sight of the important lessons about major evolutionary change told through bones over geologic time.

Of all the two-hundred-odd bones in a whale’s body, skulls are probably the most important part to examine if you’re interested in the big picture of whale evolution. Like the skull of any vertebrate animal, whale skulls past and present conveniently house the primary organs for taste, smell, sight, sound, and thought all in one unit. Skulls are thus rich sources of functional information about the lives of whales and their transformation over time—after all, these senses are tweaked, enhanced, or diminished when lineages undergo major ecological transitions, such as the one from land to sea, over the course of evolution. Despite their durability, skulls are challenging objects for study. Their individual bones interlock with one another in complex and hidden ways, with blind corners, overlapping parts, and delicate connections. Soft tissues such as the eyes and brain all rest across several bones, like fruit sitting in a bowl made of interconnected puzzle pieces. To make things even more interesting, whale skulls are not only intricate but big. I’ve stared at whale skulls long enough that they feel familiar to me, but I always have to remember that whale skulls are, in very clear ways, unlike those belonging to any other mammal.

Take the skull of a bottlenose dolphin, which rests comfortably on a desk but would require two hands to move carefully. It consists of two basic parts: a paddle-shaped beak, formed of elongate bones with rows of teeth like pencil tips; and a cranium of layered bones that cover a bowling ball–shaped braincase. About those teeth: You won’t find the traditional lineup of incisors, canines, premolars, and molars that mammals usually possess. At some point in their evolutionary history, toothed whales gave up chewing for merely seizing their prey with a snap of their jaws and then swallowing it whole. A bottlenose dolphin may flash what looks like a welcoming toothy grin, but I wouldn’t put my hands anywhere near it.

Moving from the beak to the rest of the skull, the next-most-obvious feature is the orbit, the bone roofing where the eye would be, like a heavy eyebrow, still very much like that of other mammals. But behind the orbits, differences begin to accumulate. First, there’s the aperture that leads to nostrils, or the blowhole. You can peer down the curved passageway formed by these nostril bones to the underside of the skull, where you’ll see an origami construction of delicate, folded bones with paper-thin edges. The bones leading to the blowhole are actually behind where the eyes would be located, the complete opposite of any other mammal, where nostrils are located at the tip of the snout. If your nostrils were positioned like those of a dolphin, you’d blow your nose from the top of your forehead.

Pakicetus, Maiacetus, and Remingtonocetus had nostrils toward the tip of their snout, and these structures slowly migrated backward in other stages of fossil whales, up to the bottlenose dolphins that we see today, with nostrils displaced well behind the eyes. Interestingly, whales aren’t alone in nostril migration: sea cows and manatees, also full-time aquatic mammals, have nostrils positioned high on the skull (though not behind the eyes), whereas their early fossil relatives have nostrils positioned more forward. This parallel migration of the nostrils lets fully aquatic mammals, such as whales and sea cows, orient their body in a more energy-saving horizontal position in the water, as opposed to doggy-paddling with their nose out of the water. But swimming horizontally is just part of the reason for the strangeness of the bottlenose dolphin skull before us.

When viewed from the side, the top of a bottlenose dolphin’s skull is shaped like a scoop—if you were a dolphin, imagine having a skull with a dishlike forehead, right above your eyebrows. In life, for a dolphin this cavity contains a cone of fat called the melon, which gives toothed whales a domelike forehead. Tucked behind the melon, and underneath the blowhole, are empty sacks: air sinuses underlain by muscles and sealed with an organ that looks like a pair of lips. When these lips buzz like a trumpeter’s, they generate sound, which bounces around the inside of the head and then gets focused by the melon into a discrete path out of the head, as a high-frequency sound beam, like an acoustic beam emanating from a searchlight strapped to the dolphin’s forehead.

By coordinating this process across a specialized set of anatomical parts to generate sound, toothed whales create a form of biological sonar—or echolocate—to see their underwater world in sound. All toothed whale species alive today echolocate, whether they are sperm whales, beaked whales, river dolphins, porpoises, or true dolphins. It’s how they wayfind, or hunt, in murky rivers or at ocean depth, sometimes a mile deep, with no light. Echolocation itself has evolved only a handful of times in other vertebrates; toothed whales are the only animals that do it underwater.

Making high-frequency sound, however, is just one part of echolocation; after sound bounces off an object, it produces an echo that the animal needs to hear. (The classic dolphin chirps and squeaks are how they vocalize to communicate with one another at lower frequencies.) We can’t hear directionally underwater, but whales can because their ear bones float inside hollow pockets. In life, the ear bones of a dolphin hang in a sinus cavity full of spongy tissue, which acoustically isolates each ear, letting the brain detect small differences in the arrival time of sound between the right and left ear, helping to pinpoint the source in three dimensions. Whales have had the ability to hear like this since bthe time of Ambulocetus.

But how does sound get to the ear bones, especially when whales don’t have external ears? Fats are conductive of sound, and along with the melon, toothed whales have large, fat bodies that fit hand in glove into hollows in the backs of their jaws and then branch out, backward, into lobes that directly connect to the ear bones. Akin to the way our ear canals funnel sound, these fat bodies provide a pathway for sound to reach the ears, although researchers still debate whether there aren’t other acoustic pathways in the head—testing the anatomical basis for echolocation requires clever experimental work, and captive animals are not easily procured. This hallmark sense in toothed whales was poorly understood until the late 1950s, when scientists, supported by strong interest from the U.S. Navy, resolved the enigmatic echolocation abilities of toothed whales in a simple way: after covering a captive dolphin’s eyes with suction cups, they discovered that the animal was still able to navigate a maze to find a small object.

We still know little about how echolocation in toothed whales truly works, especially in wild, free-ranging animals. It surpasses even the best military technology, and we’re just starting to understand the basics. There is marked variation, from species to species, in terms of the arrangement of air sacs, melon shape, and even ear bones. What any one of those differences means for sound frequency and how whales perceive it remains a broad mystery, hopefully the work of future dissertations. We can say, however, that a quick tour of a dolphin skull shows us how evolution refashions or modifies existing parts—like nostrils displaced behind the level of the eyes—while other times it generates completely new structures, such as a biosonar apparatus on the forehead. How evolutionary novelty happens remains one of the most important unanswered questions in biology today.

Transformation and novelty are recurring themes in whale evolution as much as they are for any group in the tree of life, and it is sometimes difficult to parse the distinction between them. Consider how the feathers that most birds use today for flight are elaborations—transformed many, many times over—from basic scales that once covered their dinosaur ancestors. Or how turtles, early in their evolutionary history, acquired shoulder blades inside their rib cage; turtles’ ribs later fused with other bones so that every turtle thereafter, for the past 200 million years, has had a shell with its shoulders tucked firmly inside. In both of these cases, evolutionary novelty does come from somewhere in the body—it’s an extreme flavor of transformation that starts with some available parts—but novelty is different because of its consequences. Whereas a transformation can be the reduction of parts or a change in size or proportions, a novelty is a one-time evolutionary appearance of a new structure that enables the evolutionary success of all of its descendants. Today’s whales are exemplars of the great success of evolutionary novelty, with echolocation in toothed whales and filter feeding in baleen whales enabled by biological apparatuses present in none of their forebears.

Evolution is the intellectual glue that connects living whales, in all of their seeming weirdness, to their deep ancestry, which is both incomplete and still not entirely known. Skulls point us to these evolutionary clues in a tangible and clear way; and without these insights, and the millions of years that brought these changes about, it would be very hard to illuminate the connection whales have with other closely related mammals.

My way back to whales, as a scientist, has always involved skulls in one way or another. In college I puzzled over a half-rotten dolphin head when our field class, on a barrier island off the coast of Georgia, discovered its carcass buried in the sand. The putrid smell dispersed the rest of the class, but I held fast. I wasn’t riveted by the grotesque allure; instead I was captivated for the first time by the thought of how, exactly, a whale might become a fossil.

In graduate school, I sought fossil whale skulls peeking out of seaside cliffs or crumbling out of rock formations in badlands that were once seafloors. Cetaceans became my vehicles for understanding life over geologic time, across scales so vast that we cannot truly comprehend them, even if paleontologists casually discuss geologic markers as if referring to last week’s dentist appointment. And it all led me to the Smithsonian, where I tend to the world’s great collection of fossil whale skulls.

However, part of the deal with being a museum scientist, especially at the Smithsonian, is that people ask you for tours. It’s more than a fair trade. First, I like giving tours. They give me an excuse to try out new ways to talk about ideas—the big ones, like the evidence for evolution and extinction—that not only excite me but also explain how whales came to be. Also, tours often involve children, and children are usually the toughest audiences of all, whether they’re whale huggers or fossil fiends. If I can figure out how to keep them interested, even for a short show-and-tell, then I think I’ve done my job. And who knows, I might even be convincing enough to ignite more than a passing interest.

My friend and colleague Megan McKenna first taught me about whale heads and their inner biosonar anatomy when we were both in graduate school, so when she brought her family recently for an early-morning tour before the museum opened, I was excited at the opportunity to settle a debt. Her four-year-old daughter was just a bit older than mine. The museum’s halls, especially when they’re empty, can be intimidating spaces, so I started slowly—I didn’t want to overwhelm or manufacture too much excitement.

As we walked into the Sant Ocean Hall, we stopped short underneath the imposing right whale model, and turned to face the eel-like skeleton of Basilosaurus, grimacing from above. Basilosaurus is an early whale several million years younger than Pakicetus and Maiacetus but it looks worlds apart in size and shape. Its dinosaur-sounding name literally translates from Latin as “king lizard,” in a nod to its serpentine, bus-length body. The first fossils of Basilosaurus were collected from the chalky marls of rural Arkansas and Alabama in the early nineteenth century. With a skull over three feet long, jaws bearing palm-sized, saw-shaped teeth, and individual vertebrae large enough to serve as stools, a Basilosaurus skeleton gives every impression of belonging to a sea monster—and its name has stuck, if only for the conventions that scientists use to name species.

Unlike today’s whales, Basilosaurus has a head that does not dominate its body, arms that crook at the elbows, and a surprising, diminutive set of legs that could never have supported its large weight on land. But Basilosaurus has an involucrum like all the other early land-dwelling whales, and its ear bones floated in a sinus space below its skull. Basilosaurus did not echolocate, nor filter feed, leaving it caught somewhat in an evolutionary middle ground for whale evolution: one of the first fully aquatic whales, completely unreliant on land, yet still carrying many of the biological apparatuses of its terrestrial ancestors.

A Basilosaurus tooth

As I walked with Megan under the skeleton, I pointed to its hind limbs and joked aloud about how they dangled like poorly placed, miniature landing gear. Megan then leaned down to her daughter. “Hey, Etta, Nick studies whale bones just like those ones.” Etta looked at me intently, mulling the assertion. “Why?”

I opened my mouth to deliver a canned response involving school, science, and curiosity, more boilerplate than authentic. I knew I could do better. I waited a few beats.

“Their bones all tell stories,” I said, “about where whales came from.” She glanced up at Basilosaurus looming from the ceiling, like a giant, flippered, macabre snake. “And if you become a scientist, you can learn to read them and know their stories.” I knew I had her attention. “But they sure don’t just show up here in the museum all put together,” I smiled. “You have to find them first.”

I walked with my eyes trained on a long, gray road cut through a hillside on a cattle ranch. In the late-afternoon California sun, the waving golden grass gave the hill the look of an enormous, shaggy animal, revealing its sedimentary flank. I followed a small path, parallel to the exposed fossil-rich layer. Walk, stare, walk some more, scrape the exposure, and then walk again; maybe you get lucky.

A few yards away from me, my colleague Jim Parham was doing the exact same thing. I’ve known Jim since graduate school, and we don’t need to say much to each other in the field. Jim’s an expert on turtles and other reptiles. As it is for me with whale bones, Jim has seen enough specimens that even the smallest fragments of fossil shell can help him solve riddles about turtle origins, which stretch back even deeper in geologic time than whales, though we tend to find fossil sea turtles in the same type of rock as fossil whales. Jim and I reliably fall on the same page, by temperament and by rock units.