What if … birds are dinosaurs?

Following on from John Ostrom’s inspired work in the 1970s, the anatomical evidence for a relationship between dinosaurs and birds is now so detailed that it is possible to reconstruct the stages by which a dromaeosaurian theropod might be transformed into an early bird.

Early small-sized theropods, such as Compsognathus, have a birdlike appearance – long, spindly legs; a long neck; and fairly small head with quite large, forward-pointing eyes – though they still retain obviously dinosaurian features, such as the clawed hands, teeth in the jaws, and a long, bulky tail.

Dromaeosaurian theropods

These bird-like dinosaurs exhibit a number of interesting anatomical changes to the basic theropod body plan. Some changes are quite subtle, but others are less so.

One notable feature is the ‘thinning’ of the tail: the tail becomes very narrow and stiffened by bundles of long, thin bones, the only flexible part being close to the hips (Figure 16, top). As argued earlier, this thin, pole-like tail may well have been valuable as a dynamic stabilizer to assist with the capture of fast-moving and elusive prey. However, this type of tail dramatically changed the pose of these animals because it was no longer a heavy, muscular cantilever for the front half of the body. If no other changes had been made to its posture, such a dinosaur would have been unbalanced and constantly pitch forward on to its nose!

To compensate for the loss of the heavy tail, the bodies of these theropods were subtly altered: the pubic bone, which marks the rearmost part of the gut and normally points forward and downward from each hip socket in theropods, was rotated backwards so that it lay parallel to the ischium (the other lower hip bone). Because of this change in orientation, the gut and associated organs could be swung backwards to lie beneath the hips. This change shifted the weight of the body backwards, and compensated for the loss of the heavy counterbalancing tail. This layout of hip bones, with the pubis rotated backward, is seen in living and fossil birds as well as maniraptoran theropods.

Another equally subtle way of compensating for the loss of the counterbalancing tail would be to shorten the chest in front of the hips, and this is also seen in these bird-like theropods. The chest also shows signs of being stiffened, and this probably reflects the predatory habits of these animals. The long arms and three-clawed hands were important for catching and subduing their prey and needed to be very powerful. The chest region was no doubt strengthened to help securely anchor the arms and shoulders to withstand the large forces associated with grappling and subduing prey. Birds also have a short, and greatly stiffened, chest region to withstand the forces associated with anchoring the powerful flight muscles.

At the front of the chest, between the shoulder joints, there is a V-shaped bone (which is in fact the fused clavicles, or collar bones -Figure 17) that acts as a spring-like spacer separating the shoulders, it also helped to anchor the shoulders in place while these animals were wrestling their prey. Birds also exhibit fused collar bones; they form the elongate ‘wish bone’, or furcula, that similarly acts as a mechanical spring that separates the shoulder joints during flapping flight.

The joints between the bones of the arm and hand were also modified so that they could be swung outward and downward with considerable speed and force to strike at prey in what has been called a ‘raking’ action. When not in use, the arms could be folded neatly against the body. The leverage for this system was also of considerable advantage to these creatures, because the arm muscles that powered this mechanism were located close to the chest and operated long tendons that ran down the arm to the hand (rather than having muscles positioned further out along the arm); this remote control system kept the weight of the body closer to the hips and helped to minimize the delicate problem of balance in these theropods. The arm-striking and arm-folding mechanism is closely similar to that employed by birds when opening and closing their wings during and after flight.


The early bird-like fossil Archaeopteryx (Figure 16, bottom) exhibits many maniraptoran theropod features: the tail is a long and very thin set of vertebrae that anchored the tail feathers on either side; the hip bones are arranged with the pubis pointing backward and downward; at the front of the chest there is a boomerang-like furcula; the jaws are lined with small, spiky teeth, rather than a more typical bird-like horny beak; the arms are long, jointed so that they can be extended and folded just as in theropods, and the hands are equipped with three sharply clawed fingers that in their arrangement and proportions are identical to those seen in maniraptoran theropods.

Specimens of Archaeopteryx were preserved as fossils under exceptional circumstances that enabled an array of exquisitely delineated flight feather impressions to be seen. These are attached to the wings and along the sides of the tail and dictate that this creature is defined as a bird: feathers are regarded as unique to birds, and therefore indicate its affinity beyond any question. This is one of the reasons why Archaeopteryx is considered to be such an important fossil, and why it has been the focus of this comparison. Had chance not led to the preservation of feathers in this instance, it is tempting to wonder how this creature might have been classified. It would very probably have been redescribed in recent years as an unusually small, dromaeosaurian theropod!

34. Restoration of the living Archaeopteryx

34. Restoration of the living Archaeopteryx

Chinese wonders

During the 1990s, explorations in quarries in Liaoning Province in north-eastern China began to yield some extraordinary, and extraordinarily well preserved, fossils of Early Cretaceous age. At first, these comprised beautifully preserved early birds such as Confuciusornis, and the skeletons included impressions of feathers, beaks, and claws. Then in 1996, a complete skeleton of a small theropod dinosaur, very similar in anatomy and proportions to the well known theropod Compsognathus (Figure 14), was described by Ji Qiang and Ji Shu’an. They named the dinosaur Sinosauropteryx. This dinosaur was remarkable because there was a fringe of filamentous structures along its backbone and across its body, suggesting some sort of covering to the skin that was akin to the ‘pile’ on a roughly made carpet; there was also evidence of soft tissues in the eye socket and in the region of the gut. It was clear that some small theropods had some type of body covering. These discoveries led to concerted efforts to find more such fossils at Liaoning; they began to appear with increasing regularity and ushered in some truly breathtaking revelations.

Shortly after Sinosauropteryx was discovered, another skeleton was revealed. This animal, named Protoarchaeopteryx, was the first to show the presence of true bird-like feathers attached to its tail and along the sides of its body, and its anatomy was much more similar to that of dromaeosaurians than Sinosauropteryx. Another discovery revealed an animal that was extremely similar to Velociraptor, but this time named Sinornithosaurus (again, apparently covered in a ‘pile’ of short filaments). Newer discoveries have included Caudipteryx, a large (turkey-sized), rather short-armed creature noted for a pronounced tuft of tail feathers and shorter fringes of feathers along its arms; smaller, heavily feathered dromaeosaurians; and in the spring of 2003 a quite remarkable ‘four-winged’ dromaeosaurian, Microraptor, was unveiled to the world. This latter creature was small and classically dromaeosaur-like, with the typically long, narrow tail, bird-like pelvis, long, grasping arms, and sharp rows of teeth lining its jaws. The tail was fringed by primary feathers and its body covered in downy ones. However, what was singularly impressive was the preservation along the arms of flight feathers forming Archaeopteryx-like wings and, very unexpectedly, similar wing-like fringes of feathers attached to the lower parts of the legs – hence the name ‘four-wing’.

Such has been the avalanche of new, ever more startling discoveries from the quarries in Liaoning over such a short space of time that it is almost impossible to imagine what might be discovered next.

Birds, theropods, and the question of dinosaur physiology

The stunning new discoveries from Liaoning contribute importantly to the earlier discussion about the biology and physiology of dinosaurs; but, as ever, they do not answer quite as many questions as we would wish.

First and foremost, it is now clear that our Victorian predecessors were not correct: feathers do not, after all, make a bird. Various sorts of skin coverings appear to have been present on a wide range of theropod dinosaurs, ranging from a shaggy, filamentous type of covering, through downy, feather-like body coverings, to fully formed contour and flight feathers. The discoveries at Liaoning force us to wonder just how widespread such body coverings might have been, not only among theropods, but, perhaps, even in other dinosaur groups as well. Given the known distribution of body coverings, it is not unreasonable to ponder the probability of giants such as Tyrannosaurus rex (which was a theropod related to Sinosauropteryx) having some sort of epidermal covering – even if only as juveniles. Such tantalizing questions cannot be answered at present, and require the discovery of new geological deposits similar in quality of fossil preservation to those at Liaoning.

It is also quite obvious that a considerable diversity of feathered theropods and what we today recognize as genuine birds (ones with a well-developed flight apparatus) coexisted during Jurassic and Cretaceous times. Archaeopteryx is late Jurassic (155 Ma) in age and was clearly feathered and bird-like. However, we now know for certain that during the younger Cretaceous (c. 120 Ma) a multiplicity of these types of ‘dinobirds’, such as Microraptor and its relatives, existed alongside true birds. The sheer diversity, or biological exuberance, of these ‘dinobirds’ is rather bewildering, and to some extent obscures the evolutionary origins of the true birds that we see around today.

From a physiological perspective, however, the proof of the existence of theropod dinosaurs with some sort of insulatory covering points very conclusively toward the fact that these dinosaurs (at least) were genuine endotherms. There are two reasons for believing this:

i) Many of these feathered dinosaurs were small-bodied (20-40 centimetres long) and, as we know, small animals have a relatively large surface area and lose body heat to the environment very quickly. Therefore insulation using filaments (which mimic the fur seen on the bodies of living mammals) and downy feathers are likely to have been a necessity if these creatures generated internal body heat.

ii) Equally, the possession of an outer insulatory layer to the skin would have made basking difficult, if not impossible, because the insulatory layer would have inhibited their ability to gain heat from the sun. Basking is the ectotherm’s way of gaining body heat, so a furry or feathered lizard is a biological impossibility.

Birds from dinosaurs: an evolutionary commentary

The implications of these new discoveries are truly fascinating. It has already been argued, with logic and some force, that small theropod dinosaurs were highly active, fast-moving, and biologically ‘sophisticated’ animals. On this basis, they seemed reasonable candidates as potential endotherms; in a sense, our inferences about their way of life suggested that they had most to benefit from being endothermic. The Liaoning discoveries confirm that many of these highly active, bird-like dinosaurs were small animals. This is a crucial point, as small size puts greatest physiological stress on endotherms because a large percentage of internally generated body heat can be lost through the skin surface; so small, active endotherms would be expected to insulate their bodies to reduce heat loss. Small theropod dinosaurs, therefore, evolved insulation to prevent heat loss because they were endotherms – not because they ‘wanted’ to become birds!

Liaoning discoveries indicate that various types of insulatory covering developed, most probably by subtle modifications to the growth patterns of normal skin scales; these ranged from hair-like filaments to full-blown feathers. It may well be that genuinely bird-like flight feathers did not evolve for the purposes of flight, but had a far more prosaic origin. Several of the ‘dinobirds’ from Liaoning seem to have tufts of feathers on the end of the tail (rather like a geisha’s fan) and fringes of feathers along the arms, on the head, or running down the spine. Clearly preservational biases may also play a part in how and on which parts of the body these may be preserved. But for the present, it seems at least possible that feathers evolved as structures linked to the behaviour of these animals: providing recognition signals, perhaps, as in living birds, or being used as part of their mating rituals, long before any genuine flight function had developed.

In this context, gliding and flight, rather than being the sine qua non of avian origins, become later, ‘add-on’ benefits. Obviously, feathers have the potential for aerodynamic uses; just as with modern birds, the ability to jump and flutter may well have embellished ‘dinobird’ mating displays. For example, in the case of the small creature Microraptor, a combination of fringes of feathers along the arms, legs, and tail would have provided it with the ability to launch itself into the air from branches or equivalent vantage points. From just this sort of starting point, gliding and true flapping flight seem a comparatively short ‘step’ indeed.

Persistent problems

We should not, however, get too carried away with the scenario outlined above. Although the Liaoning discoveries are indeed incredibly important, offering, as they do, a richly detailed window on dinosaurian and avian evolution in the Cretaceous, they do not necessarily provide all the answers. One crucial point that must be remembered is that the quarries of Liaoning are Early Cretaceous in age, and their fossils are therefore considerably younger (by some 30 Ma at least) than the earliest well-preserved feathered dinosaur with highly developed and complex wings, Archaeopteryx. Whatever the path that led to the evolution of the first flying dinosaurs, and ultimately to birds, it was emphatically not via the extraordinary feathered dinosaurs from Liaoning. What we see at Liaoning is a snap shot of the evolutionary diversification of avian theropods (and some true birds), not the origin of birds: bird origins are still shrouded by sediments of Middle or possibly even Early Jurassic age – before Archaeopteryx ever fluttered to Earth. Everything that we know to date points to a very close relationship between theropod dinosaurs and early birds, but those crucial Early or Middle Jurassic theropods that were ancestral to Archaeopteryx are yet to be discovered. It is to be hoped that in future years some spectacular discoveries will be made that fill in this part of the story.

Topic 5 concluded with the view that dinosaurs lived at a time in Earth history that favoured large-bodied, highly active creatures that were able to maintain a stable, high body temperature without most of the costs of being genuinely endothermic. The ‘dinobirds’ from Liaoning suggest that this view is wrong – small, insulated theropods simply had to be endothermic and their close relationship to birds, which we know are endothermic, simply reinforces the point.

My response to this is: well, yes and no. There is now little doubt that bird-like theropod dinosaurs were endotherms in a true sense. However, I do think that the arguments suggesting that the majority of more traditional dinosaurs were inertial homeotherms (their large body size enabled stable internal temperature) still hold. There is some evidence in support of my view to be found among living endotherms. Elephants, for example, have a much lower metabolic rate than mice – for exactly these reasons. Mice are small, lose heat rapidly to the environment, and have to maintain a high metabolic rate to replenish the heat loss. Elephants are large (generally dinosaur-sized) and have a stable internal body temperature due to their size, not just because they are endothermic. Indeed, being a large endotherm is, in part at least, a physiological challenge. For example, elephants suffer problems if they move around too quickly: their postural and leg muscles create a great deal of extra chemical heat, and they need to use their large, ‘flappy’ ears to help them to radiate heat rapidly to prevent fatal overheating.

Dinosaurs were on the whole super-large and their bodies would have been capable of maintaining a constant internal temperature; extrapolating from the elephant, it would not have been in dinosaurs’ interests to be genuine endotherms, in a world that was in any case very warm. Having evolved physiologically as mass-homeotherms (having a stable internal body temperature that was made possible by large body size), the only group of dinosaurs that bucked the general dinosaurian trend toward large size and evolved into a small-bodied group were the dromaeosaurian theropods.

It is clear, from their anatomy alone, that dromaeosaurians were highly active and would have benefited from homeothermy, and their relatively large brains would have demanded a constant supply of oxygen and nutrients. Paradoxically, homeothermy cannot be maintained at small body size without an insulatory covering because of the unsustainable heat loss through the skin. The choice was stark and simple: small theropods had to either abandon their high-activity lifestyle and become conventionally reptilian, or boost internal heat production and become properly endothermic, avoiding heat loss by developing skin insulation. So, I propose that it is not a case of ‘all or nothing’; most dinosaurs were basically mass-homeotherms that were able to sustain high activity levels without the full costs of mammalian or avian styles of endothermy; however, the small, and in particular the dromaeosaurian, theropods (and their descendants, the true birds) were obliged to develop full-blown endothermy and the associated insulatory covering.

Next post:

Previous post: