The fossil skull sitting on Alexander Kellner's worktable makes no sense. Usually fossils are spark plugs in the engine of the imagination. One look at the long, gently curved bones of a spider monkey's arm and you see a graceful swing from branch to branch; one glance at the weighty leg bones of a mastodon and you hear the rumble of a heavy stride. But this skull greets you with cognitive dissonance: all you see is essentially a flat triangle of bone. "A paleontologist once came in here and said, 'What is that?' " says Kellner, himself a paleontologist at the American Museum of Natural History in New York.
Slowly, though, with Kellner's guidance, the details that will give this triangle meaning emerge. There is a pair of large eye sockets midway along the skull, staring out of a braincase the size of a child's head; the straight, smooth daggers in front now become a set of jaws. Most prominent on the skull is a crest of bone that starts at the tip of the creature's beak; it rises above the upper jaw, passes over and behind the eyes, and shoots out behind the skull like an oar. The crest doubles the length of the skull, from two feet to four. How could any creature support such a burden? "It was hollow," Kellner says simply, picking up a piece of the crest to show how the bone is only a few hundredths of an inch thick in some places. While the crest's interior is now filled with rock, in life there was only air. It's all a little less confusing now but no less bizarre.
This skull belonged to a pterosaur--an extinct flying reptile-- named Tupuxuara, and it is one of the latest additions to an inventory of more than 100 pterosaur species known to paleontologists. Tupuxuara lived 110 million years ago in what is now Brazil. It had wings measuring 18 feet from tip to tip, over 6 feet longer than the wings of the largest living bird, the albatross. Yet since many of its bones were as thin as its crest, it weighed no more than 45 pounds.
For the most part, Kellner says, Tupuxuara soared, but from time to time it would flap its wings to stay aloft. With its sharp eyes it would scan the broad lagoons below for food. The fern-and-pine-lined shores of these lagoons were home to many other species of pterosaurs, as well as dinosaurs, turtles, crocodiles, birds, and amphibians. Most important to Tupuxuara, the lagoons were teeming with fish. With a few precise adjustments to its wings, it would dip down to the surface of the water, pluck a fish, and gulp it as it flew away.
If, in Kellner's telling, this pterosaur's behavior comes off as remarkably birdlike, it is no accident. According to much of the newest research by the small number of pterosaur enthusiasts worldwide, these extinct reptiles in many ways mirrored the habits of their far more familiar successors in the sky. Pterosaur flight, the researchers say, was like that of birds down to subtle details; pterosaurs played many of the same roles birds do in today's ecology, and pterosaurs may even have had birdlike social structures and child-rearing techniques.
Pterosaurs haven't always been seen this way; they weren't even originally seen as denizens of the air. In 1784 the Italian zoologist Cosimo Collini, trying to make sense of the first pterosaur fossil ever found, guessed that he was looking at a sea creature--an amphibious mammal. Early in the nineteenth century the French anatomist Baron Georges Cuvier saw drawings of this fossil and quickly interpreted the animal as a flying reptile, but his colleagues found that concept hard to swallow. Throughout the 1800s researchers suggested pterosaurs were everything from extinct bats to flying marsupials. Eventually they seem to have reached consensus on the image of a lizard that, while it didn't actually fly, did glide through the air on batlike wings.
Part of their confusion stemmed from the pterosaurs' bones. Hollow and thin, they were perfect for flight but not for preservation. Only a few sites scattered around the world have yielded many pterosaur fossils that aren't badly damaged. Generally they are found in slabs of limestone, greatly flattened and most often fragmented.
Now, however, beautiful pterosaur fossils are emerging from a scrubby plateau in northeastern Brazil called Chapada do Araripe. The site itself isn't newfound. Two wandering naturalists discovered it in 1817, and paleontologists have been hauling out fossil fish, amphibians, insects, and plants ever since. But not until 1971 was the first pterosaur discovered at Araripe, and most of the other pterosaur fossils there have been uncovered within just the past ten years. Fourteen new species, including Tupuxuara, have now been found at Araripe, each dramatically different from all previously known pterosaurs.
The Araripe pterosaurs are miraculously preserved. For reasons geochemists still don't fully understand, animals that died in the lagoons of Araripe sometimes began to fossilize unusually quickly. In other ancient lagoons, bones that sank to the bottom were covered in mud and gradually crushed as the mud accumulated and turned to stone. But when an animal died in the Araripe water, it was quickly coated with a layer of sediment. Layer after layer of limestone built up on the bones, encasing them in large, round nodules. How quickly the nodules formed is hard to say, but judging from the presence of fossilized muscle and sometimes even bacteria on the skin of Araripe animals, it could have happened within hours. It's as if nature were packing away fine china for 110 million years. When paleontologists unearth these pale lumps, crack them open, and soak away the limestone, they often find fossils with unprecedented three-dimensional detail.
Kellner is one of the world's experts on the Araripe pterosaurs, and he doesn't conceal his pride in them. "Paleontologists always like to say 'My fossils are the best,' " he says, "but these really are the best." Stafford Howse, a paleontologist at the University of London, recently finished a study of a pterosaur skull from Araripe that is preserved down to its ear bones. "If the thing had died six months ago," he says, "and been buried in the soil for the worms and maggots to get to and had then been exhumed, this is the condition it would be in. It's perfect in every respect."
As far as paleontologists can now determine, pterosaurs descended from a small, lightweight bipedal reptile that lived perhaps 250 million years ago, an animal that also gave rise to dinosaurs (and thus, most paleontologists believe, ultimately to birds) at around the same time. How pterosaurs evolved toward flight is unknown; they may have made their first faltering attempts by leaping from the ground or by tumbling out of trees. All researchers can say with certainty is that 225 million years ago there were pterosaurs, which were fully qualified fliers.
The first pterosaurs were small, ranging from robin to sea gull size. They generally had long, narrow heads filled with teeth. Most notably they possessed a finger on each hand--their "pinkie"--that was longer than their entire body; this outsize appendage supported a wing. Their other three fingers were perfectly normal and tipped with claws. Trailing behind them was a long tail that, like the tail on a kite, stabilized their flight. They probably ate fish as a rule, although some may also have eaten insects. At some point on the evolutionary path from their cold-blooded ancestors they must have become warm-blooded, because otherwise they wouldn't have had enough sustained energy to fly. Bolstering this notion, and giving strong hints of what they looked like when alive, is a pterosaur fossil found in Russia in 1970 that shows signs of a thick coat of fur.
This standard pterosaur model persisted for 45 million years. But around 180 million years ago a new version made its appearance. These newer pterosaurs are called pterodactyloids (from the name given to the first member of the group found, Pterodactylus, or "wing finger"; the older pterosaurs are referred to as rhamphorhynchoids). Pterodactyloids manifested some significant changes: Their long head grew still longer, yet, because it had lost some bones in the skull, it became even more lightweight. Their neck became flexible and birdlike. They lost some or all of their teeth. Most important, their tail shrank to a stump, making it useless for stabilizing flight. The only way that tail loss can be explained, say paleontologists, is by pterodactyloids' having developed more sophisticated brains capable of stabilizing flight with quick, small changes to the wings.
For more than 30 million years pterodactyloids remained relatively rare. But 144 million years ago, for reasons unknown, their primitive relatives vanished, and at that point the pterodactyloids exploded into a strange diversity of species. Some became enormous--one late species named Quetzalcoatlus was 39 feet across, making it by far the largest animal ever to fly. Other pterodactyloids developed extremely peculiar heads: one species had hundreds of bristles for teeth; another had a duck bill, another a spoon bill. Many pterodactyloids had bizarre crests like Tupuxuara's; some had crests shaped like swords, others like keels.
For all the hallucinatory appeal of pterosaur heads, however, it is the creatures' method of flight that remains the central question for paleontologists. Pterosaurs were the first vertebrates to fly, and only two other animals have since joined them in the air: birds and bats. By comparing the bones of all three creatures, paleontologists have tried to draw analogies that might suggest exactly how pterosaurs flew. Judging from the bones in the shoulder region, researchers have concluded that the extinct reptiles could have flapped their wings as powerfully as a bird or a bat.
Yet the analogies can go only so far. Bird and bat wings themselves are very different from each other. In birds, the wings are made of feathers with stiff shafts that are rooted in the arms and the fused fingers of the hands. The wings of bats are made of elastic membranes that run between four elongated fingers and reach all the way down to the feet. By moving their feet and fingers, bats control the shape and tautness of the wings.
Pterosaur wings, supported by that single outlandish finger, could not have been really close to either of these models. Something had to give the wings support, to keep them from flapping uselessly in the wind. Either the wings had to be stiffened internally or they had to be tied down to legs or feet. But we have no perfectly preserved soft-tissue fossil to tell us. The best clues come from 80 fossil impressions of wings left behind in the mud. Unfortunately these tracings do not provide researchers with any definitive answers. Usually the wings became distorted and were pulled into unnatural positions in death, and without exception their impressions are maddeningly ambiguous. In many of the fossils it's not even clear if the impressions were made by a wing or by some other soft tissue folded into a provocative position by accidental events. But since the nineteenth century, paleontologists, observing that pterosaur wings looked as if they were made of a membrane like that of a bat, have assumed that the creatures did in fact sport leathery wings that ran, like a bat's, from finger to feet.
For more than a decade paleontologist Kevin Padian, of the University of California at Berkeley, has been vigorously fighting this view. "If you're choosing between the bat and bird analogy," he says, "there are many more reasons to draw comparisons with birds." Batlike pterosaur wings would show a clear attachment to the foot, he says, perhaps with a notch in one of the bones, but he has seen none. A bat's wing also has a tendon running along its trailing edge, but there is no evidence of one in pterosaur fossils.
There is evidence, however, of a different sort of structural component. Paleontologists have long noted that some pterosaur fossil wings bear thin parallel ridges. Peter Wellnhofer, the curator of the Bavarian State Museum in Munich, argued in 1987 that these ridges represent tough fibers, possibly made of the protein collagen, sandwiched inside the wing to provide stiffness. Recently, Padian and Jeremy Rayner, a zoologist at the University of Bristol in England, found that the tips of pterosaur wings support this idea. Since bats have an elastic membrane that they pull tight, their wing comes to a sharp point. But while studying a particularly well preserved pterosaur in Germany, Padian and Rayner discovered that its wing tip forms a blunted curve. Such a shape requires some internal stiffening, such as could have been provided by fibers.
This pterosaur had been preserved in a belly-up position, exposing the underside of its wings. Padian and Rayner found that at the trailing edge some of the fibers had apparently separated from the wing and were lying on top of it at an angle, leaving behind corresponding grooves on the wing itself. The way the fibers had frayed indicates that rather than being sandwiched inside the wing, as many paleontologists had imagined, they lined the underside like the ribbing of an umbrella.
If fibers were indeed a structural component of the wing, then they, rather than the ambiguous impressions of the wing membrane, may be the best indicator of a pterosaur wing's shape. And according to the fibers, pterosaurs had narrow, birdlike wings that attached at the hip or the mid-thigh. Rayner, whose specialty is animal aerodynamics, says the fibers would have given the wings so much stiffness that a pterosaur wouldn't have needed to keep its wings taut with its feet, as a bat does. Instead the fibers would have acted more like a bird's stiff feather shafts, which, when the bird brings its wing down, transfer the lift generated by the wing to the rest of the body.
The shape of pterosaur wings must have affected not only how they flew but also how they moved on the ground. Birds, with their wings free of their hind legs, can walk easily. But bats have to deal with two problems: their wings connect their arms to their legs, and their legs are designed to swing out to the side. In flight this keeps the wing membrane tight, but on the ground it makes a bat's legs too loose to support the weight of its body. As a result, bats crawl on all fours, with their long arms and flexible legs splayed out to the sides.
Just as pterosaurs' wings were designed for birdlike flight, Padian maintains, their legs were designed for bipedal, birdlike walking. He points out that the head of their femur was oriented to swing the leg forward and backward, not side to side, and their ankle bones had fused into a hinge that couldn't bend out in a sprawl. Far from crawling on all fours, he says, pterosaurs walked on two legs as comfortably as a dinosaur or a bird--which, Padian argues, shouldn't be surprising. "They really are coming from fairly similar ancestral stocks," he says, "and so a lot of what they do is going to be predisposed to be similar." Pterosaurs with wings folded up on either side also leave their hands relatively free for handling food or climbing trees.
But the traditional batlike stance still charms many paleontologists. "I find myself with the fuddy-duddies dotted around the world," says David Unwin of the University of Bristol. "In our humble opinion, the legs sort of stick out sideways and the animals tend to sprawl somewhat on the ground and walk around in a quadrupedal fashion." Unwin, along with Wellnhofer, has argued that the hips of pterosaurs had sockets unsuited for walking on two legs. If a pterosaur tried to walk upright, they maintain, the pelvis wouldn't bear directly down on the femur and the femur heads would tend to pop out of their sockets. "I think you can get pterosaur hind limbs into a bipedal position in the same way I can do a split," says Unwin. "It hurts, but I may be able to do it with a bit of training."
Padian counters that what might seem unlikely judging from only one part of a skeleton becomes possible when you add supporting cartilage and muscle. "What pterosaurs can't do," Padian adds, "is jerk the hind limb out to the side in exactly the way a bat does it."
Recently, Christopher Bennett of the University of Kansas gave some of Padian's ideas a new twist. By tilting the hip 60 degrees from the horizontal, he found that a protruding part of the socket's rim was positioned so that it could bear down on a femur head. A pterosaur with such a hip wouldn't be hunched over like a bird; it would be upright, something between a gorilla and a human. Instead of walking on its toes, it would roll back on its heels and keep its feet flat on the ground. Pterosaurs standing as tall as 20 feet might have strolled along the banks of lagoons with a humanlike gait.
Once you know how an extinct animal moved--whether it crawled or ran, flapped its wings or only glided--you can get an idea of how it fit into the ecological web of its time. If, as Padian argues, pterosaurs flew more like birds than bats, then one would expect them to play birdlike roles in their ecology as well.
Rayner and graduate student Grant Hazelhurst have found evidence for this idea in the geometry of pterosaur bodies. Previous research has shown that the ratios determined by a flying animal's mass, wingspan, and wing surface area predict what kind of flight style it will have. Forest birds share similar ratios, while soarers have their own set, as do the insect eaters, divers, and so on. Rayner and Hazelhurst found that the combination of long, thin wings and slender, lightweight bodies made large pterosaurs very aerodynamically efficient, able to soar on the weakest of rising air currents. Many of the largest pterosaurs may have been like today's frigate bird: perhaps they soared hundreds of miles over the ocean, grabbing fish or harassing other pterosaurs to surrender their catches. Perhaps, also like the frigate bird, they were helpless if they accidentally landed on water. Smaller pterosaurs probably flew like sea gulls and petrels, needing to flap more often to keep themselves aloft. Other pterosaurs were aerial hunters like falcons. And the smallest ones were the swallows of their day, chasing insects and maneuvering to match their prey's unpredictable flight.
Rayner and Hazelhurst's work is just as interesting for what it says pterosaurs didn't do as for what it says they did. Their long wings disqualified them from diving for fish like a duck or loon because they would have caused too much drag in the water. The researchers also say those long wings would have been unwieldy in forests. "If you've got a bird living in trees or nesting in twigs, it has a real problem if it's got very long wings because they can knock against the environment," Rayner explains.
Rayner is the first to point out that the fossils may be skewing his view. The fossils of small land animals are most often preserved in watery environments, such as oceans or lagoons, so creatures that kept to forests, deserts, or plains left behind much sketchier traces. Rayner didn't consider many pterosaurs that have left only fragmentary remains. For example, paleontologists know the giant Quetzalcoatlus only from the arm bones of a single animal. As a result, it's still not clear exactly what its physical proportions were, and paleontologists suspect that its wings may actually have been small for its body. Quetzalcoatlus is also unusual because it wasn't found by an ocean but instead in what were once seasonal wetlands far inland in Texas. The biggest birds that live in these kinds of habitats today are egrets and herons, which (possibly like Quetzalcoatlus) have relatively small, squarish wings that let them keep clear of the trees and other plants around them. Imagine Quetzalcoatlus as a spectacularly oversize egret, wading silently and delicately through the marsh, its long neck pulled back, and then suddenly plunging its head into the water to snatch fish with its chopsticks-style jaws.
Other pterosaurs apparently foreshadowed birds' feeding habits in surprising ways. One, named Pterodaustro, had hundreds of long bristlelike teeth in its lower jaw. Flamingos have similar ridges in their bill that they use to filter out algae and insects from water. Pterodaustro was exactly suited for that kind of feeding and supremely unqualified for anything else. In London, Stafford Howse has been studying the jaw of a pterosaur he refers to as the Purbeck spoonbill, after the geologic formation in England where it was found; like the living spoonbills of Africa and Asia, it had a long, narrow beak ending in two horizontal disks. Spoonbills today move their beak through mud at the bottom of streams and lakes to stir up the animals they feed on. The Purbeck spoonbill seems to have been even better adapted to this kind of feeding because it had several dozen sturdy curved teeth on both jaws. Although useless for chewing, since they jutted away from the mouth, they would have been perfect as a mud rake. Moreover, part of the roof of the Purbeck spoonbill's mouth was lined with horn; when it dug up snails and other shelled animals, it could have broken them open like a nutcracker.
In 1991 it occurred to University of Miami ecologist Thomas Fleming that some pterosaurs ought to have eaten fruit. Fruit-bearing plants now depend on bats, birds, and primates to eat their fruit and spread the seeds in their droppings. Yet paleobotanists believe that for 40 million years after fruit-bearing plants first appeared, these dispersers either hadn't yet evolved or were rare. There were plant-eating dinosaurs, of course, but they probably destroyed the seeds in their enormous, slow- moving digestive systems. However, Fleming hypothesized, the seeds would probably have survived a passage through a pterosaur's small gut. Furthermore, it made sense that pterosaurs would have taken advantage of a food that could provide them with the energy they needed for flying. The only problem with Fleming's idea was that at the time there weren't any pterosaurs known that might have been fruit eaters.
Fleming decided to publish his suggestion anyway, and only after his paper had been typeset did Wellnhofer tell him about a pterosaur from Araripe that he and Kellner were beginning to describe; they had named it Tapejara (meaning "the old being" in the language of the Tupi Indians). This pterosaur had an eight-inch-long head dominated by a high, thin prow in front of its nose that tapered back to a narrow prong above the eye. Its toothless jaws were like a precise pair of tweezers, sharp and slightly bowed.
The combination of crest and beak rules out many of the usual pterosaur feeding techniques. The fish eaters had either gripping teeth or long jaws like a pelican's. Tapejara's beak might have been good for picking at carrion, but its crest would have gotten in the way when it tried to poke into a carcass. Yet both its crest and its beak made Tapejara well suited for picking fruit. It could have used the crest to push aside thick foliage and could then have plucked fruit from the stem with its beak, which is precisely what hornbills and toucans do today. If so, then perhaps pterosaurs like Tapejara were crucial to the evolution of fruits like avocados and mangoes.
Flying and eating undoubtedly took up a healthy part of a pterosaur's day but not all of it. What did pterosaurs do the rest of the time? What was their social life like? Christopher Bennett has recently shed some light on the question for a pterosaur known as Pteranodon, which lived from 115 million to 70 million years ago along the coast of a sea that ran down the middle of North America. It was among the best of the soaring pterosaurs, with a wingspan ranging from 10 to 25 feet. Pteranodon fossils have been found at sites that were 100 miles or more from the coastline of the ancient sea, which suggests that these pterosaurs regularly flew great distances in search of food.
Bennett compared some 1,100 Pteranodon fossils, plotting on a graph the length of their individual leg bones, finger bones, and so on. In each case the measurements bunched into two distinct groups. The average individual in the small group, he calculated, had a wingspan of 12.5 feet, while the larger ones averaged 19 feet. The smaller pteranodons seemed to be more common, outnumbering the larger creatures two to one.
Beyond their size--and presumably their weight--there were only two other differences Bennett could find between the groups: the pelvises of the large animals were proportionately narrow, while those belonging to the smaller ones were wide; the smaller animals also had small crests, while the large pteranodons had extravagant ones.
Paleontologists have long speculated on the purpose of pterosaur crests. Did they improve the aerodynamics of the body? Did they act as a rudder? Kellner points out that Tupuxuara's fossilized crest was covered with the impressions of blood vessels. A dense mesh of capillaries on the crest, he proposes, might have acted like an air conditioner during flight, bringing hot blood close to the skin.
Bennett doesn't think either aerodynamics or cooling makes sense. If a small crest is sufficient to do the job, then why do a third of the pteranodons have much bigger ones? The only explanation that made sense to him was that he was looking at male and female pteranodons. The crests were primarily male mating displays, like antlers on an elk or a long tail on a bird, and the wide hips of the smaller, female pteranodons were designed to pass eggs.
To understand this combination of sexual differences in size, displays, and population ratios, Bennett turned to living animals as analogues. He found several creatures with these traits, ranging from elephant seals to boat-tailed grackles. And, he learned, they all court, mate, and raise their young in a similar fashion. Bennett thinks Pteranodon probably did the same. He envisions male pteranodons on crowded rookeries lining the ocean, competing for the attention of females. Usually crest and body-size display would be enough for males to ward off challengers, but occasionally they might have to fight. The winners would maintain big harems of the smaller females, but they'd offer no help in raising their young. The unlucky males could only look on enviously.
The effort of attracting females would have taken a heavy toll on the males. A wing broken in combat would have meant being grounded and eventually starving. Big crests might have made flying more difficult. While the males had a 50 percent wider wingspan than the females, they might well have been twice as heavy, and flying might thus have been far more taxing for them. This kind of built-in hazard could explain the skewed sex ratio Bennett found. "Apparently it increases the reproductive success of the ones who survive," he says, "so it can be selected for, even though in other respects it gets you into trouble."
In studying these bones, Bennett was able to get an idea of what it was like to grow up as a pteranodon. Along with measuring the size of the various bones, Bennett looked for signs of immaturity, such as unfused bones and spongy tissue at the joints where the animals still had youthful cartilage. After separating the creatures by age, Bennett found that typical bone from young pteranodons was similar to the bone of human children and other warm-blooded young in that it was full of blood vessels. Older pteranodons had bones of less uniform composition: the inner and outer surfaces of the hollow bones were covered with a layer of hard, smooth, less vascularized material. These findings suggest that pteranodons went though a massive growth spurt that ended at adulthood. Once they reached full size their bones hardened over. "In other words," says Bennett, "they weren't like living reptiles, like alligators, that continue growing for many, many years. If you could let alligators live a hundred years, they'd still be growing."
Bennett also found that all the immature pteranodons represented in the sample--about 14 percent--were already the same size as the adults. "If they were going out fishing and feeding themselves and slowly continuing to grow," says Bennett, "you would expect to find young animals in the middle of the seaway, but they are just not there. There you find only animals that are virtually adult size."
This would tally with their fast growth. Pteranodons must have started life with a wingspan of a few inches at most, since their wings had to fold inside an egg that could in turn fit inside a mother. "But there's no way that a pterosaur could grow rapidly to large size and feed itself at the same time," Bennett says, because flying is so metabolically costly. After pteranodons hatched, they stayed in their nests, waiting for their mother to bring them food from the ocean--just as modern albatrosses do. "While the birds are being fed by their mother, they just sit around and grow till they are virtually adult size," Bennett explains. "And then they say, 'Okay, I'll go out and take care of myself now.' " Albatrosses need only a year to grow to full size, and judging from their bone growth, Bennett