Uphill Flight

A partridge's ability to climb overhanging slopes might explain how dinosaurs took to the skies.

By Adam Summers ~ Illustration by Roberto Osti

Chukar partridge does not use its wings when on level ground But when it climbs a steep slope, it flaps its wings from roughly its head to its tail, generating a force (purple arrow) perpendicular to the plane in which the wings move. That force "holds" the animal to the ground, giving extra traction to the bird's feet as it climbs. When the bird climbs a vertical surface (3), however, its wings beat in a more back-to-front fashion, and the force they generate has both a horizontal (blue arrow) and a vertical (red arrow) component. Although the vertical component is not necessary for climbing a tree trunk-the bird generates enough force for that with its legs alone-the component shows that the bird (or, equally, perhaps, a protobird or a feathered dinosaur) can redirect the wing-flapping force merely by altering the plane in which the wings are flapped. Such an ability would have been crucial to the origins of flight, as wings were co-opted to provide thrust instead of traction

The debate over the origin of birds has raged through the paleontological community for more than a century. Fitting species into evolutionary family trees is painstaking and often contentious work, but truly amazing discoveries of feathered fossils in Liaoning Province in northeastern China have enabled paleontologists to identify the group of dinosaurs that gave rise to Tweety and brethren. The fossils, unearthed in the past decade, even give a peek at the origin of feathers. But paleontologists still debate one point: How did bipedal but terrestrial archosaurs (the "old lizards," which include dinosaurs, birds, and crocodilians) learn to flap their arms and fly? Not surprisingly (given the title of this column), biomechanics has come to the rescue. One of the most compelling hypotheses for the evolution of avian flight has recently been well fortified by observing the habits of some of today's poorest fliers.

Two main camps have dominated the debate about the origin of flight. According to the "trees-down" camp, arboreal dinosaurs first evolved the ability to glide off their perch in a tree, much the way colugos-the so-called "flying lemurs"-and some frogs, lizards, snakes, and squirrels do today. Later, the gliders evolved the ability to flap from tree to tree.

Proponents of the trees-down scenario maintain that wings and feathers would have been useful for gliding, even if they preceded such adaptations as the shoulder girdle, the huge pectoral muscles, and the peculiar wrist and hand structures that make possible the powered, flapping flight of birds. Yet, as detractors of the hypothesis point out, none of the extant gliding animals perform even rudimentary flapping. They are all strictly gliders, and there is no reason to suppose they will ever be otherwise. Even worse, the dinosaurs most closely related to birds, the unfeathered dromaeosaurs, which include such terrors as Deinonychus and the better known Velociraptor, were clearly terrestrial. So even though a change from gliding to flapping might be an easy idea to swallow, neither the several independently evolved gliders nor the fossil record lend it any support.

Partisans from the second camp, in contrast, favor a "ground-up" hypothesis. In their view, terrestrial, bipedal dinosaurs flapped their "arms" first and later evolved into fliers. But the ground-up hypothesis has faced an even tougher challenge than the trees-down view. Although the fossil record clearly demonstrates that pre-avian dinosaurs were fond of terra firma, explanations that require the transition from bald, sprinting dinosaur to feathered, flapping bird seem a bit far-fetched. Feathers might have, for example, evolved as insulation, which would further imply that dromaeosaurs were endothermic, or warm-blooded. Or maybe feathered arms were useful as a net to catch flying insects, or as a horizontal stabilizer-like a tightrope-walker's pole-for swiftly running, predatory bipeds.

One biologist has come up with a ground-up proposal that, on the face of it, might seem even more off the wall. Kenneth P. Dial studies the biomechanics of flight at the University of Montana in Missoula. He suggested recently that flight arose from arm movements intended to push a bird (or a feathered dinosaur) into the ground rather than lift it up. The genesis of that odd idea was his observation that, when running up a slope, a chukar partridge (Alectoris chukar) flaps its wings quite differently than a bird does when it tries to get off the ground.

Partridges, chickens, and quail are known as galliform birds (the name comes from the Latin word gallus, meaning "rooster," and the Galliformes are all chickenlike). Typically, they have broad, stubby wings; easily fatigued flight muscles; and chicks that are ready to run, though not to fly, when they hatch. When a predator such as a fox or a weasel threatens a young chukar partridge, the bird escapes by fleeing up a steep slope. As it runs uphill, the chukar flaps its wings madly. The behavior has long been regarded as a failed attempt at flying, pointless because the young chukar's flight feathers (called remiges) are not yet fully developed.

Dial first established that though the remiges are not long enough to enable takeoff, they do improve traction enough for the young chukars to climb. After trimming or removing the remiges of chukars of various ages, Dial discovered that without the help of feathers, the birds could not run up slopes steeper than sixty degrees. Fully feathered animals, however, could scamper and flap their way up vertical and even slightly overhanging slopes.

Dial then turned his attention to the birds' legs. To measure their contribution to the climb, he constructed two kinds of ramp, smooth and textured, which gave quite different traction to scrabbling claws. No matter how well feathered they were, adult birds and young birds alike couldn't scale smooth ramps steeper than fifty degrees.

The data could be explained in two ways. It might initially seem obvious that the flight feathers, though short on the younger birds, nonetheless provide enough vertical lift to make the chicks light on their feet, boosting them up the steeper slopes. Alternatively, the flapping wings could be generating force in the direction of the ramp, increasing the hind-limb traction of the fleeing chicks. This hypothesis also fits with another observation: the stroke of every chukar's (whether young or old) wing beat while running is quite different than that of its wing beat while flying. Rather than flapping the wings from back to belly, as other birds do, the partridges flap from head to tail.

To test the two hypotheses, Dial and his student Matthew W. Bundle attached a small accelerometer to the back of a bird (the instrument measures the acceleration of the bird's center of mass at any point in time) and filmed the animal running up a ramp. They confirmed that from late in the downstroke through the middle of the upstroke, much of the force generated by the flapping wings helps a chukar's feet get traction.

This research implies a plausible model for the selective advantage of both the flapping motion and a poorly feathered wing. Lightly feathered dromaeosaurs might have relied on wings for help in climbing steep slopes and even entering trees, just as extant galliform birds do. The peculiar flapping style that helps ground the bird could then easily be co-opted into the wing stroke now present in flighted birds. The chukars vary the angle of their wings depending on the slope of the substrate they're climbing, and the angle becomes increasingly similar to that of a flying bird as a chukar climbs slopes of ninety degrees or steeper.

It's not conclusive evidence for the evolution of flight-and since behavior doesn't fossilize, one can never be certain. For the first time, however, the ground-up proponents have a model that's not so much "off the wall" as up it.

Adam Summers is an assistant professor of ecology and evolutionary biology at the University of California, Irvine (asummers@uci.edu).