To get to where they need to go without crashing, fruit flies(Drosophila melanogaster) turn away from so-called visual expansion,caused by the image of a looming object expanding on the retina. But the fly's world is complex, and `any insect flying forwards is going to experience a certain degree of visual expansion', says Seth Budick of the California Institute of Technology. If flies always turned away from this visual expansion, then they would never get anywhere. The question is, what keeps flies flying in a straight line?

Budick and his colleagues Michael Reiser and Michael Dickinson used a simple magnetic tether to find out what keeps flies on the straight and narrow(p. 4092). Each fly had a steel pin attached to its back, and the pin's end sat in a small depression in a minute sapphire block attached to a magnet fixed directly above the fly. This set-up allowed the flies to swivel on the spot and orientate themselves in different directions.

The team knew that some flying insects tend to orientate themselves so that the wind is blowing in their faces, so first they investigated how flies orientated at different wind speeds. By blowing winds of 0–1 m s–1 over the flies, they found that at all wind velocities,the flies turned themselves around so that they faced into the wind. However it was not clear how much the flies were blown about by the wind, and how much they were actively orientating. To find out, the team tethered both dead and live flies and repeated their experiment, finding that although dead flies were blown to face into the wind, the live flies still orientated more accurately into the wind, especially at lower wind speeds.

Having shown that flies respond to wind stimuli and turn into the wind, the team wondered which of the flies' sense organs were sensing the wind and helping them fly straight. Results from other insects indicate that the Johnston's organs, sense organs near the base of the antennae that respond to antennal movements, could play a role. The team glued segments on one or both of the antennae together, which removes feedback to the Johnston's organs. The team found that flies needed both antennae intact to orientate properly into the wind, especially at lower wind speeds. This indicates that feedback from the Johnston's organs, caused by wind movements, might help flies to fly in a straight line.

But what about visual stimuli? In the final part of the experiment, the team exposed the flies to different wind speeds and to a pattern representing the visual expansion that a fly would experience during flight. They did this using a cylinder lined with light emitting diodes surrounding the tether, with gaps in front of and behind the fly to let the wind pass through. When the team presented the visually expanding stimulus from the downwind direction,the flies flew very reliably into the wind, turning away from the `looming'object. When they presented the expanding stimulus from the upwind direction,however, the flies still generally orientated into the wind, and didn't turn away from it. They mostly turned away only when the pattern expanded very quickly. This shows that there is a trade-off between flying into the wind,and avoiding expanding visual stimuli. By using feedback from the antennae,this trade-off makes sure that flies don't always turn away from visual expansion, and reach their destination while still avoiding crashes.

Budick, S. A., Reiser, M. B. and Dickinson, M. H.(
). The role of visual and mechanosensory cues in structuring forward flight in Drosophila melanogaster.
J. Exp. Biol.