In its modern incarnation, South Carolina's picturesque Charleston Harbor hosts a wide variety of marine birds—from the pelicans and cormorants that forage in its estuaries to the gulls and herons that breed and nest on its offshore islands and the songbirds that pass through en route to warmer climes for the winter months. Around 25 million years ago, however, dragons ruled the Carolina skies. These beasts were not the monsters of medieval folklore, of course, but rather evolution's closest facsimiles, fearsome in their own right: giant flying birds with wings longer than those of some light aircraft and beaks equipped with deadly, spearlike choppers.

The evidence for these terrifying creatures comes from fossils found at Charleston International Airport, appropriately enough. In 1983 a team led by paleontologist Al Sanders, then at the Charleston Museum, unearthed the bones and recognized that they belonged to a large bird. But the researchers had their hands full with other fossils, and the avian bones went into storage. Three decades would pass before an analysis carried out by one of us (Ksepka) revealed just how remarkable the forgotten animal was. Sanders and his colleagues had recovered the largest flying bird on record, a never before seen species belonging to an enigmatic group known as the pelagornithids. Ksepka named the creature Pelagornis sandersi, in honor of its discoverer.

For more than 150 years paleontologists have recognized that pelagornithids once patrolled the air. But with only a handful of fragmentary specimens available for study, little was known about how these animals flew, what their lives were like or why they evolved such extreme proportions. Recent analyses of the biggest of them all, P. sandersi, along with other studies of colossal avians carried out by the other of us (Habib), have filled in many gaps, helping to paint the most complete picture to date of these astonishing animals. The latest evidence indicates that pelagornithids rose to prominence in the aftermath of the asteroid impact that doomed the dinosaurs and their close relatives the flying pterosaurs and that they may have developed their impressive size as an adaptation to foraging over the open ocean. Whatever the driving force behind their gigantism, they were able to achieve sizes beyond the limits of what some researchers thought was possible for a flying bird.

Enigmatic Bones

The study of pelagornithids has a long, rich history. In 1857 French paleontologist Édouard Lartet described a very large wing bone of one of these birds, which he believed might have belonged to an ancient albatross. He dubbed it Pelagornis miocaenus, meaning simply “Miocene seabird.” Although the name was uninspiring, the fossil was electrifying and mysterious. The wing bone, a humerus, measured nearly 0.6 meter (two feet) long, indicating that its owner had been a bird twice the size of some modern albatrosses—unthinkable in Lartet's day. Unfortunately, with just that piece of wing to go on, paleontologists had no real clue about what the rest of the animal looked like.

A hint that the owner of the huge bone was not a supersized albatross emerged more than a decade later, in 1873, when English anatomist Sir Richard Owen described the skull of another giant bird, which he assigned to a new species, Odontopteryx toliapica. His work made clear that the skull was so distinctive that it could not belong to any of the modern bird groups. Instead it represented a previously unrecognized group of huge extinct birds. The eventual discovery of more complete specimens revealed that Lartet's humerus belonged to this same group.

Additional discoveries came to light slowly over the next century, sometimes vanishing soon afterward. In 1910 one of the most complete pelagornithid skulls ever found was attributed to a new species, Pseudodontornis longirostris. The University of Königsberg in Germany had purchased the skull from a Brazilian sailor. But during World War II, Allied bombing devastated Königsberg, which was annexed by the Soviet Union and renamed Kaliningrad at the end of the war. Today the fossil's whereabouts are unknown; no one is sure if it was destroyed in the conflict, stolen or removed to another location.

In the decades that followed, fossil hunters discovered more pelagornithid species, including Pelagornis orri from California and Pelagornis chilensis from Chile. Whereas most of the earliest finds were scrappy, partial skeletons from these new species allowed scientists to begin piecing together a more detailed understanding of how these animals were built and what kinds of activities they were adapted to perform.

The emerging picture defied imagination. Foremost on the long list of unusual features of pelagornithids are the serried ranks of toothlike structures that line their upper and lower jaws. Birds lost the ability to form teeth more than 65 million years ago. But pelagornithids evolved a work-around. Unlike true teeth, which are composed of enamel and calcified tissue known as dentine and are set in sockets, the so-called pseudoteeth of pelagornithids were hollow projections formed directly from bone. These pseudoteeth were arranged in orderly, repeating sets of size classes in the best-known species. A pair of short, thin, needlelike pseudoteeth flanked each of the medium-sized projections, and a pair of these three-tooth packages in turn flanked the tallest, conelike pseudoteeth. In life, a thin layer of the same material that sheathes the beak of modern birds probably covered the bony teeth. The overall effect was that of a menacing grin rippling with spikes adapted to nabbing and holding onto prey.

Other weird traits further enhanced the hunting prowess of pelagornithids. The skull of these birds was uniquely flexible. Its midpoint had a strong hinge that permitted bending at the spot where the braincase met the upper beak. Additionally, the lower jaw had a joint built into the midpoint of the left and right sides. Instead of solid bone, the jaw was held together at the “chin” by a flexible ligament. Together these traits would have enabled substantial bending of the jaws, perhaps to accommodate large prey.

Bones from below the head also distinguish pelagornithids from other avians. The wing bones of these birds are so flattened that some paleontologists actually arranged one of them, the humerus, upside down in past skeletal reconstructions. Although all flying birds have hollow bones, which make their skeletons structurally efficient, pelagornithids took that trend to extremes. All their wing elements exhibit exceptionally thin bone walls. This thinness means the birds retain the bone stiffness they need with a minimum of weight—critical for giant flying animals. But lightly built bones have a downside: an unexpected collision might spell doom because such bones are easier to fracture. A break in one of them would ground the bird, leaving it unable to feed.

The leg bones are arguably the most normal part of pelagornithids, at least in terms of their shape. Yet they are almost comically small compared with the wing bones. Nevertheless, the hind limb bones had reinforced walls and a stout shape that would have made them relatively strong. Like many living seabirds, pelagornithids were probably somewhat awkward when it came to crossing large distances on land. But all they needed, presumably, was the ability to sprint effectively for short distances to initiate takeoff.

Record Breaker

By the time P. sandersi was finally described in 2014, scientists had already established that the pelagornithids were highly unusual birds. But P. sandersi one-upped even that strange company. Its humerus alone measured nearly a meter (around three feet) in length—more than a third longer than Lartet's original pelagornithid humerus and even longer than the entire arm of an average person. It seemed inconceivable that fossils of such enormous size could even belong to a bird. Indeed, some research suggested a theoretical limit of 5.1 meters for wingspan in a marine soaring bird, beyond which an animal simply would be too heavy to remain aloft by flapping. Yet the limb bones found at the Charleston airport clearly represented an avian wing and leg, as indicated by their telltale wafer-thin walls, which needed careful treatment with chemical hardening agents to keep them from crumbling into shards. And there was no mistaking the accompanying skull, with its trademark pseudoteeth, for anything other than a pelagornithid.

The excellent preservation of these skeletal elements, combined with insights from other pelagornithid specimens, allows for a detailed reconstruction of P. sandersi. In life, the feathered wings of this bird would have measured an estimated 6.06 to 7.38 meters (20 to 24 feet) tip to tip—the largest wingspan of any bird on record, living or extinct, and more than double the average wingspan of the largest modern flying bird species, the wandering albatross. Extrapolating from the circumference of the weight-bearing leg bones, P. sandersi would have tipped the scales at somewhere between 21.9 and 40.1 kilograms (48 to 88 pounds)—the weight of a golden retriever. Although massive compared with modern fliers, the animal was dainty for its wingspan, thanks to its small body and ultralightweight skeleton.

Proceeding from those parameters, we have worked out how this magnificent creature and other giant pelagornithids flew. Estimating the locomotor capabilities of extinct animals is a tricky exercise, but researchers today have better tools than ever before to do so. Key observations from living birds, along with fundamental physical principles from aerodynamics, informed our proposed flight scenario.

Today's flying birds exhibit a wide variety of flying styles, such as the hummingbird's hovering and the seagull's slower flapping flight. Right away the incredibly long wings of P. sandersi and other pelagornithids suggested that their primary mode of flight was soaring, in which the wings do not flap to generate lift but instead are held outstretched to use energy from wind or rising air. Modern soaring birds have a few different ways of remaining aloft for long periods, though, and figuring out what strategy pelagornithids employed required deeper analysis.

Species such as condors and vultures possess broad wings relative to their body weight, which creates what is termed low wing loading—that is, each square centimeter of wing is required to support relatively fewer grams than would be needed in a bird of comparable mass but less expansive wings. The wings of these birds also have slotted tips, meaning that the feathers at the tip of the wing can spread apart, reducing drag. The combination of low wing loading and slotted wing tips enables these animals to surf currents formed by warm air as it wafts up from land. And it allows them to do so with relatively shorter wings than seabirds have, which comes in handy when navigating environments with obstacles such as cliffs and vegetation.

Frigate birds pursue a second type of soaring, traveling on thermals that form over ocean rather than land. They have more slender, tapered wings with pointed, rather than slotted, tips. They are also among the most lightly built of all birds and thus exhibit exceptionally low wing loading. These traits aid frigate birds in traveling long distances while cruising high up in the sky, ready to swoop down to capture prey near the sea surface.

At the other end of the marine soaring spectrum are the albatrosses, which also have very long, narrow wings with pointed tips. Albatrosses, however, are heavier relative to their wing area, which means they need strong, fast winds to power their flight. Albatrosses fly by harnessing the wind gradientabove the waves. They fly into the slower wind near the surface of the water to gain altitude and then curve around to ride the stronger winds back down to sea level, endlessly looping to gain altitude and trade it for distance in a maneuver called dynamic soaring. In 2004 an albatross outfitted with a tracking device was clocked moving an average of 127 kilometers an hour for nine hours straight—the record sustained soaring speed for any living animal. It was riding winds from an Antarctic storm.

Improved knowledge of pelagornithids from spectacular specimens like that of P. sandersi suggests that these birds specialized in a form of soaring not seen among today's soaring birds. Their wings were narrow but still large in area thanks to their great length. In other words, evolution gave these birds the best of two worlds: their large overall size would have allowed them to use dynamic soaring when winds were strong, and with their large wing area and high aspect ratio, they would have also excelled at cruising over quiet oceans for thousands of kilometers at a time. The biggest pelagornithids would have been able to cover those distances relatively rapidly: we calculated that the speed of optimal efficiency for these giants would have been more than 40 kilometers an hour, putting them well ahead of the pace attained by world record holder of the 100-meter dash, Usain Bolt, who broke the ribbon at 9.58 seconds—equivalent to running 37.6 kilometers an hour. Moreover, P. sandersi could have maintained that pace with relatively little effort: after gaining 45 meters of altitude, the bird could glide for more than a kilometer without any flapping or assistance from winds.

Although P. sandersi probably spent most of its time on the wing, it would have to land occasionally (to nest, for example), which would also mean taking off again. The tiny legs of large pelagornithids originally led some researchers to question the ability of these large birds to launch effectively. But with the discovery of more complete behemoths, including P. chilensis and P. sandersi, it became apparent that the hind limbs were actually well proportioned to the relatively compact bodies of these giant birds. The first ever quantitative analysis of launch mechanics in giant pelagornithids, presented by Habib at a premiere international paleontology meeting, found that the short, stout hind limbs of Pelagornis were appropriately shaped and positioned for brief sprints, especially over water surfaces (the feet in Pelagornis were most likely webbed). The bones of the hind limbs were also sufficiently strong to support significant muscle mass, able to propel the modest-sized bodies (with their oversized wings) up to launch speeds. These leg traits would have made P. sandersi an excellent water launcher, even if it was probably relatively poor at walking over land.

A Vacant Niche

The discovery of P. sandersi—a titan among what were already considered to be exceptionally large birds—raises the question of why giant size evolved in flying avians. Gigantism is not universally advantageous in biology. Big animals need more food than small ones, they may require larger areas for nesting and they tend to have smaller population sizes than modestly proportioned species. Yet despite those drawbacks, multiple successful lineages of giant fliers have evolved over the course of the earth's history. In fact, the lack of truly enormous fliers today is the exception to the rule: giant flying animals darkened the skies for most of the past 120 million years.

It turns out that large size has considerable upside. For one, it improves the efficiency of long-distance flight because bigger fliers use less energy per unit distance covered than their small counterparts do. Larger animals can also capture (or steal) prey that smaller fliers cannot handle. Furthermore, large flying animals have limited predation risk—once airborne, a big flier is almost immune to attack from predators.

For millions of years the winged reptiles known as pterosaurs ruled the airspace over land and sea. Those living over the oceans probably fed on invertebrates and fish, and they had body plans well adapted to long-distance ocean voyages. They were very successful. But the same asteroid impact that extinguished the dinosaurs (apart from birds, which are living dinosaurs) also did in the pterosaurs. With their extinction, competition in several realms suddenly plummeted, and the ecological “niches” they had occupied opened up. One of these niches was that of the large, marine soarer.

Pelagornithids appear to have filled this role, debuting approximately 10 million years after the last pterosaurs. Their fossils come almost exclusively from sedimentary deposits in ocean environments, indicating that marine prey formed the mainstay of their diet. Because their pseudoteeth were not very strong compared with true teeth, some paleontologists speculate that soft-bodied animals such as squid and eels found near the ocean surface may have been the primary food source. Other, more ill-gotten morsels may also have been on the menu. Today large marine birds often bully other species into relinquishing their food, sometimes even harassing other birds in flight until they vomit up their prey, as the skua does. By far the largest birds in their ecosystems, pelagornithids may well have harangued smaller seabirds to rob them of meals. They also could have snatched chicks from their nest, a predation behavior practiced by modern giant petrels, skuas and even some pelicans.

Pelagornithids were not the only large birds to help fill the roles vacated by pterosaurs: another group of large flying birds, the teratorns, appeared about 23 million years ago and survived all the way up to the end of the Pleistocene epoch, 11,700 years ago. With their shorter, broader wings and heavier bodies, they probably flew and hunted more like condors.

After soaring over the seas for more than 50 million years, pelagornithids vanished completely roughly three million years ago during the Pliocene epoch. The root cause of their disappearance remains a mystery. The Pliocene witnessed profound changes in the oceans as the Panama land bridge closed, sundering a major connection between the Atlantic and the Pacific and radically altering currents. Yet it is hard to imagine even this event ending a lineage that had survived so many previous shifts in climate, ocean circulation and fauna.

Perhaps overspecialization played a role in the demise of pelagornithids. Early in the radiation of this group, several “small” species, which reached the size of modern albatrosses, evolved. Over time these diminutive forms died out, and for the last half of pelagornithid history only giant species remained. These behemoths may have relied more heavily on specialized feeding strategies and global wind currents than smaller marine birds did—and ultimately became victims of their own success.