Birds do it, bees do it, but unfortunately we cannot. I'm alluding to the capacity for self-powered aerial flight, an activity that we humans can only appreciate from a foreign perspective. Nonetheless, because we have designed our own flying machines from scratch, we recognize all too well the difficulties inherent in getting something aloft and keeping it there. Hence it makes perfect sense that questions relating to the origins of flight in various animal groups perennially pique our interest. In a newly published study, James Marden and Michael Thomas draw upon numerous sources (paleontology, behavior, developmental genetics and morphology) to make the case that flying insect's wings evolved from the gills of aquatic or amphibious ancestors.

As with the origins of bird flight, myriad hypotheses have been proposed regarding the evolution of flight in insects. Two leading hypotheses differ markedly in their approach to understanding this remarkable event. One proposes that insects are a sister-group of myriapods (e.g. millipedes and centipedes) with wings that evolved from lateral protrusions of the thorax. Implicit in this hypothesis is the notion that wings would have arisen in a terrestrial environment, after progressing through a sequence of functional stages like parachuting, gliding and ultimately through to flapping flight. The other hypothesis, and the scenario that Marden and Thomas favor, suggests that insects are a sister-group of crustaceans (e.g. crabs, brine shrimp and copepods) whose wings were derived from gills, presumably in an aquatic or amphibious environment. How might a gill evolve into a wing? Marden and Thomas address this by developing a `wings from gills' model based on their study of Diamphipnopsis samali, a species of stonefly that possesses wings on its thorax and gills on its abdomen, a condition considered primitive for flying insects.

Marden and Thomas first collected a number of live specimens from southern Chile and performed tests of their locomotor capacity in air and water. Their ability for flight is only marginal, but on the water surface they create lift- and drag-based forces for propulsion using their hindwings and forewings respectively (see video sequences at http://www.bio.psu.edu/People/Faculty/Marden/movies/rowing.mov). The drag-based rowing locomotion of the forewings is intriguing because it may serve as an analog for an intermediate step in the evolution of insect flight.

As I understand it, their `wings from gills' scenario goes something like this: (1) ancestors of flying insects were aquatic, bore moveable thoracic and abdominal gills and relied on a blood-based gas exchange system rather than the tracheal system seen in insects today; (2) to take advantage of high aerial oxygen content, pterygote ancestors began to exploit the water surface where their gills could be exposed to air as well as water; (3) once at the air–water interface the tracheal system began to evolve, reducing the need for external gills in gas exchange and allowing them to become specialized for other tasks; (4) structurally sound and mobile thoracic gills began to be used as locomotory structures involved in various forms of water-surface travel (e.g. skimming, rowing, sailing) and eventually were co-opted for more sophisticated 3-dimensional aerodynamic function.

Like any evolutionary scenario, this one requires several major assumptions, not least of which is that insects at one time used a blood-gas exchange system. Regardless, by blending varied modern analyses and a bit of imagination, Marden and Thomas have opened a new window into how insects might have evolved not only flight but their wings as well.