Posted on Feb 17, 2009

The amazing or phenomenal is often found in the seemingly simplest thing – like a dragonfly. Common insects though they are, dragonflies possess an exceptional ability that has fascinated one Union biologist for years.

A dragonfly takes a break on Prof. Robert Olberg's neck in the indoor flight arena at the Janelia Farm Research Campus in Virginia.

“Dragonflies have the best eyes in the insect world and probably the best flight – and they’re 97-percent accurate in catching flying insects like mosquitoes,” professor Robert Olberg said. “With a visual response time of only 30 milliseconds, they typically capture their prey midflight in about 150 milliseconds.”

“Our own visual systems are much slower. By the time we realize they’ve taken off, they’ve already caught their prey,” he continued. “I want to understand how they do this.”

In his quest to discover why the dragonfly is such a remarkable aerial assassin, Olberg has learned the insect owes much of its enviable skill to a specialized set of neurons.

“We have identified neurons directly involved with this behavior,” Olberg said. “And when those neurons fire, they make the dragonfly turn its wings just so or move its head a certain way.”

Olberg knows this because he and his colleagues have actually placed tiny electrodes on the insects’ neurons. These microelectrodes, whose tips are smaller than a wavelength of light and require specialized equipment to insert, measure the response of neurons to stimuli in Olberg’s lab.

“We show a stationary dragonfly moving images on a computer screen and see how they respond,” he explained. “We’ve found the specific function for each neuron this way. One might say, ‘Go left.’ Another might say, ‘Go right.’”

A prototype version of the telemetry chip, developed to record and transmit neuron and muscle activity from a freely flying dragonfly.

Olberg and his student-researchers also conducted experiments outside with a small bead masquerading as tasty prey. In filming the dragonflies’ pursuit of the bead and playing it back in slow motion, they saw something else that makes these arthropods awesomely efficient predators.

“We’ve learned they have a very sophisticated prey-capture strategy – they predict the location prey will be at and intercept it there,” Olberg said. “Most animals aim at where the prey is, but dragonflies aim at where it will be.”

Having accumulated all this knowledge, Olberg is preceding with the next phase of his investigation into the neurological workings of the dragonfly. He knows how the insects’ neurons respond to images on a computer screen, but he doesn’t know how those same neurons respond when the animals are actually free to behave normally.

In an indoor flight arena full of dragonflies, their prey, a built-in pond and artificial sunlight, Olberg is working with Dr. Anthony Leonardo at Janelia Farm, a research campus of the Howard Hughes Medical Institute in Ashburn, Va. Here, he explained, the dragonflies can react to food-items like fruit flies more naturally.

“The big dream is to outfit the arena dragonflies with very small telemetry chips containing accelerometers, amplifiers and transmitters,” Olberg said. “We have a colleague, Dr. Reid Harrison at the University of Utah, who is working on this.”

“We are now testing the first version, a 600-milligram chip, and development is underway for an even smaller, lighter one,” he added. “We’re not there yet – this is a many-year project – but it’s where we’re going.”

The Air Force is interested in where Olberg’s efforts are taking him.

Union Prof. Robert Olberg, and Anthony Leonardo of Howard Hughes Medical Institute, attach an RFID tag to a dragonfly. The RFID tags are being tested as a possible power source for an implanted telemetry chip.

“My lab at Union gets helpful support from the Air Force Office of Scientific Research, whose mission is to fund pure research in areas it believes have value,” Olberg said. “There are possible applications for what we’re studying in biomimetics. We can take what we learn from insects and apply that knowledge to technologies like robotics and machine vision.”

For his part, though, Olberg doesn’t think of his research in terms of applications.

“I just want to understand how it works,” he said.

To learn more about Olberg's work, click here.