Dinosaurs Rediscovered

by Michael J Benton

  Dinosaurs may not have been especially adapted for swimming, as crocodiles or seals are. For example, they generally did not have deep, narrow tails for beating in the water, or fused fingers and toes, and paddle-like limbs. Still, as already noted, nearly all animals today can swim, even if they show no special adaptations for it – think of horses and cattle swimming strongly enough to cross fast-flowing rivers, just by beating their spindly legs underwater.

  Could dinosaurs fly?

  Birds can fly, and so too could the pterosaurs, the flying reptiles that were close relatives of dinosaurs. Until the 1990s, most palaeontologists would have said that dinosaurs proper did not fly. However, the discoveries of feathered dinosaurs from China changed everything, as we saw in Chapter 4. In fact, John Ostrom had enough evidence for dinosaur flight back in 1969, when he described Deinonychus, but he could not say so. He wondered at the long arms of this dinosaur, and interpreted them as perhaps necessary for predatory behaviours, such as grabbing or wrestling with prey. He and Bob Bakker even speculated whether Deinonychus might have had feathers along its strong arms, but feathers were not preserved, and other palaeontologists derided such speculations.

  A diversity of Maniraptora, or a ‘maniraptoran montage’.

  In 1986, Jacques Gauthier named this key group of theropods – Deinonychus, birds, and their close relatives – as Maniraptora, a wonderful choice of name that means ‘hand hunters’. He noted that Deinonychus and its kin shared with birds their long arms and other characters, and that in this regard they entirely bucked the general trend in theropod evolution, which saw the arms reducing in size through time – reaching extremes of uselessness, as we have seen, in T. rex and Carnotaurus.

  The discovery of ever more specimens of feathered dinosaurs from China confirmed earlier suspicions, that the elongate arms of Maniraptora were indeed lined with large, complex feathers, equivalent to the primaries and secondaries seen in bird wings today. These dinosaurs could fly. But what is flight? It’s important to realize that flight includes all kinds of ways to move through the air apart from leaping and plummeting to the ground. Today, there are many flying tetrapods, not just birds and bats, but also frogs with membranes between their toes, snakes that can spread their bodies out sideways, lizards with membranes down their sides and supported by elongate ribs, and a whole range of mammals with membranes extending between their arms and legs. These animals can all fly, but it is not the flapping flight of a bird or bat. However, they extend the distance they can travel when they leap from tree to tree, and that’s a form of flight. It is commonly called parachuting if they are mainly slowing down their rate of descent, or gliding if they are mainly moving horizontally, but dropping as they go.

  The feathered maniraptoran dinosaurs such as Anchiornis and Microraptor, then, could fly, in the sense of parachuting or gliding. There have been experiments on how they flew and how well, often using models in wind tunnels, and sometimes digital models. For example, Colin Palmer, an engineer who runs businesses selling sailed surfboards, carried out experiments on a model of Microraptor. He built a full-scale model from structural foam and coated it with resin to create a smooth finish, and lined the arms and legs with feathers plucked from modern birds, but arranged exactly as in the fossils. The model was tested in the 2 × 1.5-metre (7 × 5-foot) wind tunnel at Southampton University, normally used for testing the aerodynamics of cars and aircraft parts. Palmer mounted his model, and varied the wind speed, orientation of flow, and model posture, allowing the limbs either to dangle down or to stick out sideways. This did affect the glide, and the legs-down posture allowed the Microraptor model to glide further. In more detail, the best policy for a gliding Microraptor was to jump off its perch with its legs sprawled out to the sides, and then to let them drop down to the dangling position as it entered the glide phase.

  Distances travelled by Colin Palmer’s model of Microraptor according to the position of the legs – dangling down or tucked up.

  Why didn’t Microraptor engage in powered flight? Palmer and colleagues concluded that for moving around in the trees, at heights of 20 or 30 metres (66 or 100 feet), there was no advantage in evolving powered flight. With its feathered fore and hindlimbs and, one assumes, decent eyesight and coordination so it could land efficiently without crashing into trees, gliding was enough. The next step in flight, as seen in early birds such as Archaeopteryx, was to expend considerable energy in pumping the wings up and down, requiring massive pectoral muscles and high volumes of nutritious food. The Late Jurassic and Early Cretaceous woods were probably home to dozens of species of small theropods, sailing through the air, legs up or down, snatching insect prey from among the branches, or even on the wing, and escaping predators by their bold leaps.

  Did bird flight originate from the ground up or the trees down?

  In Chapter 4, we saw how feathers evolved, and gliding dinosaurs like Microraptor had all the specialized flight feathers seen in a modern bird. These Chinese fossils have settled a controversy about aerodynamics that raged when I was a student. Some experts declared that it was impossible for a glider to become a flapping flyer – they said the muscles and specialized bone joints were not there, and they pointed to modern gliders, such as lizards with extended ribs and skin, and flying foxes with skin flaps stretched between their arms and legs. Surely these gliders could not switch one aerodynamic structure for another, the arm-supported wing of a bird or bat? Those researchers also argued that birds evolved their flight from the ground up – leaping as they ran, and beating their wings to extend their leaps. I’d always thought this was a crazy idea.

  Well, Microraptor and its relatives had proper wings, indeed four of them, lined with complex flight feathers. The step to Archaeopteryx, still accepted as the first bird, as we saw in Chapter 4, seems slight – increase the area of the wing a little so it can support the body weight, and strengthen the main muscle of the breast region, the pectoral muscle, so it can power the driving down-beat of the wing. Archaeopteryx might not have had the massive pectoral muscle of some modern birds, but its powers of flight were probably sufficient to lift it above the trees for short journeys of a kilometre or so. That would have been enough to escape from predators, or to hop to a new copse to exploit the crawling and flying insects in the tree tops.

  Ground up or trees down, or something in between? I think the evidence of the Chinese fossils is hugely in support of the traditional ‘trees down’ model for the origin of flight. As any physicist will tell you, ‘Use gravity, stupid!’ Why would running dinosaurs fight against gravity by leaping up from the ground to snatch insects, and then somehow evolve ever more advanced abilities to fly? The maniraptorans most likely became small and extended their arms and powerful hands as direct adaptations for climbing and hanging on to trees. While their larger cousins, such as Allosaurus and ancestors of T. rex, were barging around on the ground in pursuit of large plant-eating dinosaurs, Microraptor and its kin were creeping around in the trees snatching insects and spiders, as well as lizards, frogs, and early mammals. The rich faunas of Jurassic and Cretaceous tree-dwellers have been revealed by recent collecting in the Burmese amber.

  The ‘trees down’ theory for the origin of flight is that small, feathered theropods of the Jurassic pranced and displayed, using their bright feather colours to win mates, and maybe partly for camouflage in the dappled sunlight. Grasping the tree boughs with their powerful hands, they leapt from branch to branch, and those with slightly elongate feathers along their arms had an advantage by being able to take bold leaps across several metres of open space to another tree. Evolution favoured the jumpers, as they got more food or escaped more readily from predators. Longer feathers and longer, stiffer arms made the wing more effective and extended the leap. However efficient the gliding wing, though, gravity means the animal sinks towards the ground, so any gliding dinosaur that could pump its wing, even ever so slightly, could counter the downwards pull and extend its flight.

  This migh
t all be a bit simplistic, of course, and many biomechanics experts now focus on how flight performance might relate to growth, and how juvenile dinosaurs might have operated. For example, Ashley Heers and Ken Dial have explored how the flight performance of ground-dwelling birds today changes with age. In a series of experiments on the Chukar partridge, they found that very young specimens charge around and can use their undeveloped, short wings to aid them in jumping and taking runs at tree trunks.

  These investigators have focused on a specialized mode of locomotion seen in these birds, called wing-assisted incline running, which is the best evidence to rescue some aspects of the old ‘ground up’ theory. Even juvenile Chukar partridges, which cannot yet fly, can enhance their speed and manoeuvrability by scampering and flapping their stumpy little wings at the same time. These authors then transfer the Chukar model to the early theropods, and make a case that these little dinosaurs charged around the Jurassic forests, running at trees to catch insects by propelling themselves, cat-like, halfway up the trunk, and then falling back. They argue that over time, this led to the evolution of full, flapping flight – and true birds.

  Dinosaurs in films – do they get it right?

  Back in 1995, I was invited by the BBC to act as one of six consultants for their new documentary series Walking with Dinosaurs. I was amazed to discover the close convergence between animators and physicists. In fact, this was a practical example of the overall shift of palaeobiology from speculation to science. If the film-makers had been working ten years earlier, we would have had little to bring to the table other than general observations of how large modern animals, such as elephants and ostriches, run and walk. Now we had biomechanics and digital models.

  My PhD student, Don Henderson, now Curator of Dinosaurs at the Royal Tyrrell Museum in Alberta, Canada, was applying his background in pure physics to calculate the leg motions of dinosaurs from first principles. He decided to treat each element of the leg as an independent pendulum, hanging down from the hip, knee, and ankle joints. He then solved a set of three equations that described how far forwards and backwards the thigh, shin, and elongate ankle bones could swing, but constrained by the need to touch the ground neatly at the end of each stride. Most of his calculations ended with the foot below or above the ground, but that was clearly impossible. He accepted only those cases where the foot, correctly, touched the ground at the end of each step.

  Henderson then went 3D, and visualized the walking hip with two legs from below. As the biped walked, the body rocked from side to side to keep balance, and the points of contact with the ground formed a triangle, from right to left to right. The secret was to keep the centre of mass within the advancing triangle. The centre of mass is the 3D pivot point that lay just in front of the hips in the lower belly region – balance a model on this pivot, and it would not tip forwards or backwards, or to left or right. If the centre of mass moved outside the moving stride triangle, the animal would fall over.

  Having seen Don Henderson’s stick-dinosaur animations, we went to the studios of Framestore, the animation company, located in the centre of London. In those days, the animation software they used to create live-action dinosaurs was enormously cumbersome, and required banks of the biggest Macintosh computers on the market to process the data. The animators used a four-step process to make the films, and they still do. First, they make a storyboard and live film of the background scenes. Where a dinosaur is to brush against a tree, a technician pushes the tree. Where the dinosaur is to splash through a river, a helper dons great boots and splashes through the water. The person is then edited out. Second, the dinosaur motions are sketched as stick models moving through the background. Third, grey cylinders are hung on the stick models like a crazy suit of loose-fitting armour, to represent the flesh of the limbs and body – they are hung in such a way that they stay put as the stick model moves about. Finally, the skin is constructed like an old-fashioned lion rug, with the skin opened out and painted in detail with all the scales, feathers, and colours required. This ‘rug’ is then draped over the grey-armour-suited dinosaurs, and the animators hope it sticks. In the early days, the stick-dinosaur sometimes vacated his coat and ran off, leaving it trailing a few frames behind.

  The animated skin of Allosaurus, designed to be slung over the stick model.

  That’s not the point, though. The animators at Framestore had not carried out the careful first-principles feasibility study that Don had, and yet they got it right. As Mike Milne, the chief of the operation, explained, ‘We’ve all spent time watching modern animals move, and so we know what’s plausible. The animal has to be light on its feet, balanced, and moving in proportion to its intended body mass.’ It’s like the intuition a ball player has in aiming or catching a ball – no need to calculate speeds and trajectories; you just know which way it’s going.

  When Walking with Dinosaurs was shown by the BBC in 1999, people were critical of all sorts of things, but nobody said the locomotion was implausible. This was to the credit of the animators, and their good sense in checking with biomechanics experts such as Don Henderson. Any deficits were down to money and computing – high-quality dinosaur animation costs a great deal, and the BBC did not have the budget of a Hollywood film producer.

  The animators found that making the animals look heavy as they marched around the Jurassic landscapes was hard to achieve. Look at any film with animated dinosaurs – they often look as if their feet are hovering a few inches above the ground, not really thundering along and making deep footprints. The animators add deep, pounding sound effects and music, or puffs of dust, or they hide the feet behind a discreet curtain of plants. Still, the movements are astonishingly well rendered – and I haven’t even mentioned how they get the sway of the body right. Sway right as you lift the left foot, left as you lift the right foot; they are making sure the animal obeys the centre of mass-stride triangle rule that Don Henderson had imposed, but they are doing that instinctively. The tail also sways in time with the undulating body, setting up rippling waves of balancing adjustments from front to tip of the tail. The head bobs from side to side and up and down as the animal walks – look at how a pigeon or pheasant nods its head as it walks. The muscles bulge in all the right places.

  The impact of Walking with Dinosaurs can be appreciated from this publicity poster.

  As for the quibbles and niggles by palaeontologists every time an animated dinosaur feature is produced – well, all I can say is, ‘relax, they got it 99 per cent right, and they have mastered an astonishingly subtle set of commands to produce something plausible’. We would know for sure, and without a training in palaeontology or biomechanics, if they got it seriously wrong. Think of those old dinosaur movies with plasticene dinosaurs animated by stop-motion photography, or worse, the films with lizards sporting cardboard crests and spikes to make them look like some dystopian kind of yet-to-be-discovered dinosaur!

  Pinning down how dinosaurs moved is important. We have seen how the earlier work by scientists and artists was compromised by changing views on their correct posture. A proper biomechanical approach shows that the dinosaurs must be correctly balanced, and that their bodies must have swayed from side to side and up and down as they walked. The fundamental point is that many earlier endeavours reconstructed impossible modes of locomotion, in which the animal would have pitched onto its nose or fallen sideways.

  McNeill Alexander led the way, and Hutchinson took things into the modern world of computational digital modelling. These approaches have shown which postures and modes of walking and running are plausible, and which are not. The methods also allow researchers to calculate speeds and gaits – could a particular dinosaur run or gallop, or was its fastest speed achieved as a determined walk? Other questions about locomotion, such as the function of the arms of T. rex, and how dinosaurs could swim and fly, are moving forward, but much more remains to be discovered.

  All of this brings palaeontologists directly in touch with film-makers, and most
ly the film-makers get it right. The latest installment in the Jurassic Park franchise, Jurassic World: Fallen Kingdom (2018), continues the good work, with agile, heavy, plausible dinosaurs running across the screen – it’s just a pity that the feathered ones are still shown without feathers, but that was a deliberate, stylistic decision by the producers.

  John Hutchinson reflects on his recent work: ‘We do not yet know whether the Mesozoic was played out in dinosaurian slow motion, or at modern speeds.’ But he is convinced that what he does is testable science:

  Indeed what we do in evolutionary biomechanics is science. We span from basic (but vital) descriptive research (‘what is that?’) to general questions (‘how does that work?’) to specific hypothesis-testing (‘is that most likely right or wrong?’). And we emphasize that science is a continual process where we investigate our and others’ past work with improved methods and evidence, where we can. We admit we’re wrong when we must, and we admit uncertainty or variation where it exists, but also embrace subjectivity as a crutch that we must sometimes lean on to make progress. But the ground we walk on is that of science itself: clear, reproducible data and tools, a spirit of sharing and professionalism, and open-mindedness.

  1v = 0.25 g0.5 × SL1.67 × h−1.17, where ‘v’ is velocity (speed), ‘h’ is the height of the hip, ‘SL’ is stride length, and ‘g’ is gravity (10 metres per second per second).