Dragonflies' Incredible Hunting Acrobatics Captured In Slow-Mo Footage
If "Top Gun" were recast with insects, the roles of Tom Cruise, Val Kilmer and fighter jets would all go to dragonflies. These charismatic bugs execute complex aerial maneuvers as they swoop beneath prey undetected. Neuroscientists at the Janelia Research Campus in Ashburn, Va., recently filmed hunting dragonflies in slow-motion, and it's one of the first times predictive movements - meaning that the bug can tell in advance where his prey will go- have been seen in insects:
By attaching minuscule beads to the dragonflies' backs, the scientists tracked the animals' flight paths. It took years to create a system that could film the beads, which are sort of like the motion-capture technology used to animate actors. Neuroscientist Anthony Leonardo, Ph.D., in an email to The Dodo, said that he picked dragonflies because "they are large insects, capable of carrying instrumentation without encumbrance." (Previously, Leonardo had outfitted the bugs with tiny backpacks, no heavier than a few grains of rice.)
(Igor Siwanowicz, Leonardo Lab, HHMI/Janelia Research Campus)
Not only do dragonflies twist in aerial acrobatics that would make even the most seasoned jet jockey jealous, but they also keep a compound eye on their prey with swift head movements. Every 30 milliseconds, the dragonfly's head circles left, then right, "all of which serve to hold the prey steady, dead-centered," in what's believed to be the best part of the dragonfly's vision, Leonardo said.
A dragonfly swoops up beneath his prey, forming a basket with his six legs.
Unlike catching a football - in which a receiver runs toward the ball's destination - the dragonfly's flight plan isn't so reflexive. Rather, the animal can predict his prey's motion before it happens with deadly accuracy, Leonardo and his colleagues reported in the journal Nature on Wednesday. By preemptively steering with head and body, Leonardo said, "this lets the dragonfly respond with greater speed and accuracy than it could based on reactions alone.
"Properly orchestrating that ballet of motion requires considerable planning and prediction - reactions just aren't good enough," said Leonardo.
Leonardo has a few ideas about how this research can help humans, too. "The most obvious, and what I have pondered the most, is the control of small autonomous drones," he said. "If we would like to engineer little flying devices that do useful things for people (e.g., deliver packages, aid rescue operations), there is a lot we can learn from insect flight and navigation."