Listening to Mark Frye talk, you might think he's passionate about jet flight. But listen closer. Frye's not discussing the latest developments in high performance aeronautics, he's describing a much more diminutive aviator;the fruit fly. Fascinated by the neural coordination of complex behaviours,Frye, Michael Dickinson and Lance Tammero have focused on the fly because its extreme performance makes it easier to relate complex neural controls to the behaviours they regulate. Tammero and Frye have turned their attention on the visual reflexes that protect flies from crashing into objects as they cruise around. By watching flies' responses in a flight arena as they flew freely,Tammero realised that the flies turned to avoid images that expanded, in much the same way as objects loom during an approach. But was that the whole story?Tammero began interfering with the insects' visual world to untangle how expanding images trigger the flies' about face().
First Tammero needed to be able to control the flies' visual world, so he tethered the insects in Dickinson's high-tech fly-flight arena, where he could control moving LED patterns inside the cylinder to simulate the world's movements around the insects as they 'flew'. He also recorded the stationary flies' reactions to their environment by tracking their wing beats. 'It's a bit like a fly video game' says Frye.
Secured inside the arena, Tammero convinced the tethered flies that the world was spinning around them, by making the LED patterns circulate towards the left; the insects made leisurely turning wing beats. Then he played the flies a highspeed translation sequence that moved from right to left; the flies beat their wings to dodge out of the way and avoid the expanding image appearing on one side. But how could the flies tell both movements apart when both movements' front views moved leftwards? Tammero divided the flies visual world in two; half infront and half behind, so that he could control each view independently to see how the insects reacted to the separate visual hemispheres.
Tammero was in for a shock. The tethered flies seemed to be taking their visual cues from the view behind! This was completely unexpected. Why would flies depend more on the backward view to tell them which way to turn? Frye explains that it's the rear view that tells the speeding insects whether they are on a collision course with an expanding image or simply turning around. If the rear image moves in the same direction as the front view, then flies know they are heading for a crash and need to turn quickly. But if the front and rear views move in opposite directions, then they are taking a leisurely spin,and there's no collision to avoid.
But what happens to the flies when they are allowed to direct their own visual world? Frye simulated the insects moving sideways. He expanded the view to the left and contracted the view to the right, and then allowed the flies to direct the sequence by analysing the insects' wing beats as they tried to manoeuvre. Surprisingly, the flies turned their view of the world to fly towards the contracting image. Frye was astonished, it was as if the insects were flying backwards to avoid a collision! 'Expansion avoidance is so strong'says Frye, 'that it may override other visual behaviours'.
So why do flies chose to behave in such a bizarre fashion? 'It could be a mechanism for flying up wind' explains Frye. When a fly is caught in a strong gust, it might be carried backwards as it struggles against the flow, so all its visual cues tell it that it's going backwards, even when it's making headway.
