Many people, on first hearing that birds might be surviving dinosaurs, think that such a striking claim must represent a complete discombobulation of received doctrine about vertebrate relationships. Nothing could be further from the truth. All paleontologists advocate a close affinity between dinosaurs and birds. The current debate centers about a small shift in phyletic branching points: birds either branched from pseudosuchians or from the descendants of pseudosuchians—the coelurosaurian dinosaurs. If birds branched at the pseudosuchian level, they cannot be labeled as descendants of dinosaurs (since dinosaurs had not yet arisen); if they evolved from coelurosaurs, they are the only surviving branch from a dinosaur stem. Since pseudosuchians and primitive dinosaurs looked so much alike, the actual point of branching need not imply much about the biology of birds. No one is suggesting that hummingbirds evolved from Stegosaurus or Triceratops.
The issue, thus explicated, may now seem rather ho-hum to many readers, although I shall soon argue (for a different reason) that it isn’t. But I want to emphasize that these twists of genealogy are of utmost concern to professional paleontologists. We care very much about who branched from whom because reconstructing the history of life is our business, and we value our favorite creatures with the same loving concern that most people invest in their families. Most people would care very much if they learned that their cousin was really their father—even if the discovery provided few insights about their biological construction.
Yale paleontologist John Ostrom has recently revived the dinosaurian theory. He restudied every specimen of Archaeopteryx—all five of them. First of all, the main objection to dinosaurs as ancestors had already been countered. At least two coelurosaurian dinosaurs had clavicles after all; they are no longer debarred as progenitors of birds. Secondly, Ostrom documents in impressive detail the extreme similarity in structure between Archaeopteryx and coelurosaurs. Since many of these common features are not shared by pseudosuchians, they either evolved twice (if pseudosuchians are ancestors of both birds and dinosaurs) or they evolved just once and birds inherited them from dinosaur ancestors.
Separate development of similar features is very common in evolution; we refer to it as parallelism, or convergence. We anticipate convergence in a few relatively simple and clearly adaptive structures when two groups share the same mode of life—consider the saber-toothed marsupial carnivore of South America and the placental saber-toothed “tiger” (see essay 28). But when we find part-by-part correspondence for minutiae of structure without clear adaptive necessity, then we conclude that the two groups share their similarities by descent from a common ancestor. Therefore, I accept Ostrom’s revival. The only major impediment to dinosaurs as ancestors of birds had already been removed with the discovery of clavicles in some coelurosaurian dinosaurs.
Birds evolved from dinosaurs, but does this mean, to cite the litany of some popular accounts, that dinosaurs are still alive? Or, to put the question more operationally, shall we classify dinosaurs and birds in the same group, with birds as the only living representatives? Paleontologists R. T. Bakker and P. M. Galton advocated this course when they proposed the new vertebrate class Dinosauria to accommodate both birds and dinosaurs.
The Telltale Wishbone. With permission from Natural History, November, 1977. © American Museum of Natural History, 1977
A decision on this question involves a basic issue in taxonomic philosophy. (Sorry to be so technical about such a hot subject, but severe misunderstandings can arise when we fail to sort formal questions in taxonomy from biological claims about structure and physiology.) Some taxonomists argue that we should group organisms only by patterns of branching: if two groups branch from each other and have no descendants (like dinosaurs and birds), they must be united in formal classification before either group joins another (like dinosaurs with other reptiles). In this so-called cladistic (or branching) system of taxonomy, dinosaurs cannot be reptiles unless birds are as well. And if birds are not reptiles, then according to the rules, dinosaurs and birds must form a single, new class.
Other taxonomists argue that branching points are not the only criterion of classification. They weigh the degree of adaptive divergence in structure as well. In the cladistic system, cows and lungfishes have a closer affinity than lungfishes and salmon because the ancestors of terrestrial vertebrates branched from the sarcopterygian fishes (a group including lungfishes) after the sarcopts had already branched from the actinopterygian fishes (standard bony fishes, including salmon). In the traditional system, we consider biological structure as well as branching pattern, and we may continue to classify lungfishes and salmon together as fish because they share so many common features of aquatic vertebrates. The ancestors of cows experienced an enormous evolutionary transformation, from amphibian to reptile to mammal; lungfish stagnated and look pretty much as they did 250 million years ago. Fish is fish, as an eminent philosopher once said.
The traditional system recognizes unequal evolutionary rates after branching as a proper criterion of classification. A group may win separate status by virtue of its profound divergence. Thus, in the traditional system, mammals can be a separate group and lungfishes can be kept with other fish. Humans can be a separate group and chimps can be kept with orangutans (even though humans and chimps share a more recent branching point than chimps and orangs). Similarly, birds can be a separate group and dinosaurs kept with reptiles, even though birds branched from dinosaurs. If birds developed the structural basis of their great success after they branched from dinosaurs, and if dinosaurs never diverged far from a basic reptilian plan, then birds should be grouped separately and dinosaurs kept with reptiles, despite their genealogical history of branching.
Thus, we finally arrive at our central question and at the union of this technical issue in taxonomy with the theme of warmblooded dinosaurs. Did birds inherit their primary features directly from dinosaurs? If they did, Bakker and Galton’s class Dinosauria should probably be accepted, despite the adherence of most modern birds to a mode of life (flight and small size) not wonderfully close to that of most dinosaurs. After all, bats, whales, and armadillos are all mammals.
Consider the two cardinal features that provided an adaptive basis for flight in birds—feathers for lift and propulsion and warmbloodedness for maintaining the consistently high levels of metabolism required by so strenuous an activity as flight. Could Archaeopteryx have inherited both these features from dinosaur ancestors?
R. T. Bakker has presented the most elegant brief for warmblooded dinosaurs. He rests his controversial case on four major arguments:
1. The structure of bone. Coldblooded animals cannot keep their body temperature at a constant level: it fluctuates in sympathy with temperatures in the outside environment. Consequently, coldblooded animals living in regions with intense seasonality (cold winters and hot summers) develop growth rings in outer layers of compact bone—alternating layers of rapid summer and slower winter growth. (Tree rings, of course, record the same pattern.) Warmblooded animals do not develop rings because their internal temperature is constant in all seasons. Dinosaurs from regions of intense seasonality do not have growth rings in their bones.
2. Geographic distribution. Large coldblooded animals do not live at high latitudes (far from the equator) because they cannot warm up enough during short winter days and are too large to find safe places for hibernation. Some large dinosaurs lived so far north that they had to endure long periods entirely devoid of sunlight during the winter.
3. Fossil ecology. Warmblooded carnivores must eat much more than coldblooded carnivores of the same size in order to maintain their constant body temperatures. Consequently, when predators and prey are about the same size, a community of coldblooded animals will include relatively more predators (since each one needs to eat so much less) than a community of warmblooded animals. The ratio of predators to prey may reach 40 percent in coldblooded communities; it does not exceed 3 percent in warmblooded communities. Predators are rare in din
osaur faunas; their relative abundance matches our expectation for modern communities of warmblooded animals.
4. Dinosaur anatomy. Dinosaurs are usually depicted as slow, lumbering beasts, but newer reconstructions (see essay 25) indicate that many large dinosaurs resembled modern running mammals in locomotor anatomy and the proportions of their limbs.
But how can we view feathers as an inheritance from dinosaurs; surely no Brontosaurus was ever invested like a peacock. For what did Archaeopteryx use its feathers? If for flight, then feathers may belong to birds alone; no one has ever postulated an airborne dinosaur (flying pterosaurs belong to a separate group). But Ostrom’s anatomical reconstruction strongly suggests that Archaeopteryx could not fly; its feathered forearms are joined to its shoulder girdle in a manner quite inappropriate for flapping a wing. Ostrom suggests a dual function for feathers: insulation to protect a small warmblooded creature from heat loss and as a sort of basket trap to catch flying insects and other small prey in a fully enclosed embrace.
Archaeopteryx was a small animal. It weighed less than a pound, and stood a full foot shorter than the smallest dinosaur. Small creatures have a very high ratio of surface area to volume (see essays 29 and 30). Heat is generated over a body’s volume and radiated out through its surface. Small warmblooded creatures have special problems in maintaining a constant body temperature since heat dissipates so quickly from their relatively enormous surface. Shrews, although insulated by a coat of hair, must eat nearly all the time to keep their internal fires burning. The ratio of surface to volume was so low in large dinosaurs that they could maintain constant temperatures without insulation. But as soon as any dinosaur or its descendant became very small, it would need insulation to remain warmblooded. Feathers may have served as a primary adaptation for constant temperatures in small dinosaurs. Bakker suggests that many small coelurosaurs may have been feathered as well. (Very few fossils preserve any feathers; Archaeopteryx is a great rarity of exquisite preservation.)
Feathers, evolved primarily for insulation, were soon exploited for another purpose in flight. Indeed, it is hard to imagine how feathers could have evolved if they never had a use apart from flight. The ancestors of birds were surely flightless, and feathers did not arise all at once and fully formed. How could natural selection build an adaptation through several intermediate stages in ancestors that had no use for it? By postulating a primary function for insulation, we may view feathers as a device for giving warmblooded dinosaurs an access to the ecological advantages of small size.
Ostrom’s arguments for a descent of birds from coelurosaurian dinosaurs do not depend upon the warmbloodedness of dinosaurs or the primary utility of feathers as insulation. They are based instead upon the classical methods of comparative anatomy—detailed part-by-part similarity between bones, and a contention that such striking resemblance must reflect common descent, not convergence. I believe that Ostrom’s arguments will stand no matter how the hot debate about warmblooded dinosaurs eventually resolves itself.
But the descent of birds from dinosaurs wins its fascination in the public eye only if birds inherited their primary adaptations of feathers and warmbloodedness directly from dinosaurs. If birds developed these adaptations after they branched, then dinosaurs are perfectly good reptiles in their physiology; they should be kept with turtles, lizards, and their kin in the class Reptilia. (I tend to be a traditionalist rather than a cladist in my taxonomic philosophy.) But if dinosaurs really were warmblooded, and if feathers were their way of remaining warmblooded at small sizes, then birds inherited the basis of their success from dinosaurs. And if dinosaurs were closer to birds than to other reptiles in their physiology, then we have a classical structural argument—not just a genealogical claim—for the formal alliance of birds and dinosaurs in a new class, Dinosauria.
Bakker and Galton write: “The avian radiation is an aerial exploitation of basic dinosaur physiology and structure, much as the bat radiation is an aerial exploitation of basic, primitive mammal physiology. Bats are not separated into an independent class merely because they fly. We believe that neither flight nor the species diversity of birds merits separation from dinosaurs on a class level.” Think of Tyrannosaurus, and thank the old terror as a representative of his group, when you split the wishbone later this month.*
27 | Nature’s Odd Couples
From Nature’s chain whatever link you strike,
Tenth, or ten thousandth, breaks the chain alike.
Alexander Pope,
An Essay on Man (1733)
POPE’S COUPLET EXPRESSES a common, if exaggerated, concept of connection among organisms in an ecosystem. But ecosystems are not so precariously balanced that the extirpation of one species must act like the first domino in that colorful metaphor of the cold war. Indeed, it could not be, for extinction is the common fate of all species—and they cannot all take their ecosystems with them. Species often have as much dependence upon each other as Longfellow’s “Ships that pass in the night.” New York City might even survive without its dogs (I’m not so sure about the cockroaches, but I’d chance it).
Shorter chains of dependence are more common. Odd couplings between dissimilar organisms form a stock in trade for popularizers of natural history. An alga and a fungus make lichen; photosynthetic microorganisms live in the tissue of reef-building corals. Natural selection is opportunistic; it fashions organisms for their current environments and cannot anticipate the future. One species often evolves an unbreakable dependency upon another species; in an inconstant world, this fruitful tie may seal its fate.
I wrote my doctoral dissertation on the fossil land snails of Bermuda. Along the shores, I would often encounter large hermit crabs incongruously stuffed—big claw protruding—into the small shell of a neritid snail (nerites include the familiar “bleeding tooth”). Why, I wondered, didn’t these crabs trade their cramped quarters for more commodious lodgings? After all, hermit crabs are exceeded only by modern executives in their frequency of entry into the real estate market. Then, one day, I saw a hermit crab with proper accommodations—a shell of the “whelk” Cittarium pica, a large snail and major food item throughout the West Indies. But the Cittarium shell was a fossil, washed out of an ancient sand dune to which it had been carried 120,000 years before by an ancestor of its current occupant. I watched carefully during the ensuing months. Most hermits had squeezed into nerites, but a few inhabited whelk shells and the shells were always fossils.
I began to put the story together, only to find that I had been scooped in 1907 by Addison E. Verrill, master taxonomist, Yale professor, protégé of Louis Agassiz, and diligent recorder of Bermuda’s natural history. Verrill searched the records of Bermudian history for references to living whelks and found that they had been abundant during the first years of human habitation. Captain John Smith, for example, recorded the fate of one crew member during the great famine of 1614–15: “One amongst the rest hid himself in the woods, and lived only on Wilkes and Land Crabs, fat and lusty, many months.” Another crew member stated that they made cement for the seams of their vessels by mixing lime from burned whelk shells with turtle oil. Verrill’s last record of living Cittarium came from kitchen middens of British soldiers stationed on Bermuda during the war of 1812. None, he reported, had been seen in recent times, “nor could I learn that any had been taken within the memory of the oldest inhabitants.” No observations during the past seventy years have revised Verrill’s conclusion that Cittarium is extinct in Bermuda.
As I read Verrill’s account, the plight of Cenobita diogenes (proper name of the large hermit crab) struck me with that anthropocentric twinge of pain often invested, perhaps improperly, in other creatures. For I realized that nature had condemned Cenobita to slow elimination on Bermuda. The neritid shells are too small; only juvenile and very young adult crabs fit inside them—and very badly at that. No other modern snail seems to suit them and a successful adult life requires the discovery and possession (often through conquest) of a
most precious and dwindling commodity—a Cittarium shell. But Cittarium, to borrow the jargon of recent years, has become a “nonrenewable resource” on Bermuda, and crabs are still recycling the shells of previous centuries. These shells are thick and strong, but they cannot resist the waves and rocks forever—and the supply constantly diminishes. A few “new” shells tumble down from the fossil dunes each year—a precious legacy from ancestral crabs that carried them up the hills ages ago—but these cannot meet the demand. Cenobita seems destined to fulfill the pessimistic vision of many futuristic films and scenarios: depleted survivors fighting to the death for a last morsel. The scientist who named this large hermit chose well. Diogenes the Cynic lit his lantern and searched the streets of Athens for an honest man; none could he find. C. diogenes will perish looking for a decent shell.
This poignant story of Cenobita emerged from deep storage in my mind when I heard a strikingly similar tale recently. Crabs and snails forged an evolutionary interdependence in the first story. A more unlikely combination—seeds and dodos—provides the second, but this one has a happy ending.
William Buckland, a leading catastrophist among nineteenth-century geologists, summarized the history of life on a large chart, folded several times to fit in the pages of his popular work Geology and Mineralogy Considered With Reference to Natural Theology. The chart depicts victims of mass extinctions grouped by the time of their extirpation. The great animals are crowded together: ichthyosaurs, dinosaurs, ammonites, and pterosaurs in one cluster; mammoths, woolly rhinos, and giant cave bears in another. At the far right, representing modern animals, the dodo stands alone, the first recorded extinction of our era. The dodo, a giant flightless pigeon (twenty-five pounds or more in weight), lived in fair abundance on the island of Mauritius. Within 200 years of its discovery in the fifteenth century, it had been wiped out—by men who prized its tasty eggs and by the hogs that early sailors had transported to Mauritius. No living dodos have been seen since 1681.