Fly wings are technological marvels, flapping at frequencies over 100 Hz to produce enough lift to keep the tiny insects aloft. One of the innovations driving this extreme performance is the resonant oscillation of the thorax that allows the wings to reach frequencies beyond those attainable by direct nervous stimulation. Fly hindwings have also been modified over the course of evolution into halteres: gyroscopic sensors that oscillate in exact opposition to the wings, providing continuous fast feedback to the fly's flight apparatus. The tight association between forewing and haltere movement seems critical for fly flight. However, there are still peculiarities of fly flight that remain unexplained. For instance, how are the movements of the wings and halters so tightly coupled at frequencies much faster than those that can be achieved by neural systems?
Tanvi Deora and her colleagues at the Tata Institute of Fundamental Research in Bangalore, India, devised a beautiful series of experiments to test the roles of neural control, passive mechanical control and sensory feedback in wing–wing and wing–haltere coordination.
If the nervous system were to control wing–haltere coupling, flapping the wing of a dead fly would not elicit the synchronous motion of the wings and halteres typical during flight. And yet, upon actuating a single wing of a dead fly, the researchers observed that the opposite wing still moved in synch with the wing that they were moving, and the halteres moved perfectly in opposition to the wings. Given the synchronization of wings and halteres in the absence of a living nervous system, this suggested a passive mechanical means of precise coordination.
Figuring that a mechanical linkage must exist between wings and halteres, the researchers set about finding where that linkage might be. In live flies, the authors cut one of two sections of the fly thorax: either the large ‘scutum’ near the head or the smaller ‘scutellum’ toward the rear. Flies with a snipped scutum still flapped their wings synchronously, but flies with a cut scutellum could not. Thus, the scutellum was the site of wing–wing linkage.
However, even when the scutellum was cut, the wing and haltere on each side continued moving in perfect opposition – suggesting the presence of another mechanical linkage. Deora and her colleagues noticed a region of thickened cuticle on each side of the fly, the ‘subepimeral ridge’, linking the base of the wing with the base of the haltere, which seemed to be a good candidate for a second linkage. And when they cut the subepimeral ridge, the coordination between the haltere's movement with its wing was lost. Thus, the subepimeral ridge was another mechanical connection driving high-frequency coordination. Even after artificially increasing the frequency of the wingbeat, the wings and halteres stayed in synch. And the tight coupling between the wings and halteres must be important because when both subepimeral ridges were cut the insects could not fly at all.
However, despite both of these mechanical linkages working to keep the wings and sensors in synch, flies can still activate one wing independently of the other, suggesting that there must be a ‘clutch’-like mechanism allowing the flies to disengage the wings from the thorax to decouple their motion. Taken as a whole, these elegant experiments show that thoracic anatomy acts as both transmission and clutch to overcome some of the seemingly unsurmountable physiological challenges of fly flight.
