Packed with tall trees and vine-like lianas, the tropical forests of South and Southeast Asia present a maze for vertebrate flyers. In these forests, a wealth of species – frogs, lizards, snakes, squirrels and colugos – flit about on makeshift wings, trading potential energy of height for the benefit of gliding flight. But without the muscle-powered flapping ability of birds and bats, gliders possess far fewer means for control, making navigating a cluttered forest a potentially perilous endeavor. How do gliders avoid arboreal obstacles while gliding and landing safely? Pranav Khandelwal and Ty Hedrick of the University of North Carolina, USA, studied the flying lizard Draco in an Indian rainforest to see how visual information is integrated with mechanics in real-world gliding.
Most quantitative studies of gliding have been done in experimental conditions, but natural gliding may be different when an animal chooses to glide for its own reasons. Khandelwal and his field assistant traveled to the Agumbe Rainforest Research Station in the Western Ghats of India, where they arranged a rig of GoPro cameras to record a population of flying lizards going about the business of gliding in the wild. The lizards glided most often in the mornings, motivated by territoriality and mate pursuit, and the highly portable setup enabled the team to move swiftly to record the glides of multiple flying lizards.
To maneuver around obstacles in a cluttered natural environment, a flying animal must create directional forces, or it will plod along its forward path, subservient to Newton's third law. Birds, for example, turn by creating asymmetric forces on the wings by altering their flapping kinematics. It is unclear how flying lizards maneuver, but they must create forces with a different set of tools; mainly, a pair of non-flappable wings made of long ribs embedded in patagial skin and a fixed energy budget set by their initial takeoff height. Khandelwal and Hedrick hypothesized that lizards use these tools guided by their visual view of the world. For example, a lizard might plan its path from the start, choosing a route that requires minimal maneuvering when faced with a forest of obstacles. Alternatively, when gliding they could react immediately to an obstacle as it looms into view, although that could be energetically costly. Finally, the lizards might use some form of vision-based planning for braking when landing, lest they barrel headfirst into their landing site, receiving a deathly smack.
To test these ideas, the researchers used stereo recordings to extract the 3D paths, velocities and accelerations of the lizards and the exact locations of the trees in their visual environment. By mapping the locations of all trees in the area, they could calculate all possible combinations of takeoff and landing for comparison with the trees that the lizards actually chose. The lizards appeared to use their knowledge of the lay of the land prior to taking off as they navigated each flight. They chose to jump in directions with less surrounding clutter, producing glides with less maneuvering in the air and, in turn, they wasted less energy. Surprisingly, lizards did not take off directly toward their target tree, leaping instead 10–41 deg off the straight-line path. In the air, flying lizards minimized maneuvering with respect to both the obstacle and the target tree, evidence that they employ a vision-based steering model.
The data also provide insight into the lizard's flight biomechanics: modeling revealed that they maneuvered around trees by rolling a maximum of 21 deg, a side tilt that provides lateral force but reduces support for body weight by ∼7%. When landing, the lizards used a visual strategy known as ‘tau-dot’, where they gradually decelerate as they approach a target, a way to reduce impact forces as they land while maintaining enough lift to stay aloft. Overall, flying lizards appear to use visual input to guide all aspects of flight, from takeoff to landing, helping to avoid costly aerial collisions and surviving to climb another tree.