DragonflyPano_edit.jpg

How do insects fly? How are the physical aspects of flying connected to the insect's neural circuitries? How do scientists use this information?
Beatrice Jin
Beatrice Jin

20180222-153652C_edit.jpg

After building a framework for analyzing the physics of insect flight, Jane Wang now studies the internal structure—including neural circuitry—that directs flight, pondering questions such as how a dragonfly rights itself in air.
Dave Burbank
Dave Burbank

20180222-161334_edit.jpg

"Once a dragonfly senses that it's falling, it quickly goes through a set of neurocomputations to instruct its muscles. The muscles contract and modulate the wing motion. The wings interact with the air, modifying the aerodynamic forces."
Dave Burbank
Dave Burbank

20180222-160909A_edit.jpg

Explaining to a colleague, "When flying insects hover, it is as if they stand in air. If they don't have an active sensory feedback circuitry, they will eventually fall."
Dave Burbank
Dave Burbank

JaneWangCollage_edit.jpg

The Wang lab also studies fruit flies, combining computer simulations with lab experiments as they fit together the puzzle pieces of how flying insects move through the air.
Beatrice Jin; Dave Burbank
Beatrice Jin; Dave Burbank

Insect Flight—Still a Mystery

by Jackie Swift

Approximately 350 million years ago, long before birds took to the skies, insects solved the riddle of flight. Some of those early insect aviators, like the four-winged dragonfly, are still with us today, virtually unchanged. Others, like the fly, have evolved into a two-winged marvel of aerodynamic maneuverability. The flight capabilities of insects are nature’s solution to locomotion in air, according to Z. Jane Wang, Physics, and there are general principles of locomotion and evolution we can learn from them.

Wang was drawn into studying insect flight as a postdoctorate when she became intrigued by the puzzle of the bumble bee. “The bumble bee myth says conventional aerodynamics don’t explain how it can fly,” Wang says. “Obviously that can’t be right, and I felt it would be good to get to the bottom of this, so I did.” Wang worked out the mathematical calculations of the governing equation for the aerodynamic force upon the insect’s wings. After that she was hooked. “Insect flight is beautiful,” she says. “It’s also mysterious.”

Dragonflies, Connecting Neurons and Muscles for Flight

Over the years, Wang has built a theoretical framework, grounded in physical laws, to analyze fundamental mechanisms underlying insect flight. Now she has turned her attention to the internal structure that orchestrates insects’ flight. In her current study, examining how dragonflies right themselves in air, she releases the insects upside down from a magnetic tether. The dragonflies free fall for about 100 milliseconds before they roll over and right themselves. “Once a dragonfly senses that it’s falling, it quickly goes through a set of neurocomputations to instruct its muscles,” Wang explains. “The muscles contract and modulate the wing motion. The wings interact with the air, modifying the aerodynamic forces, and the resulting torque rotates the dragonfly’s body 180 degrees.”

Dragonflies flap their wings about 40 times a second. To track the wing motion, Wang uses high-speed video cameras, filming at three different angles. She constructs computer simulations to examine the consequence of wing motions on the insect’s body movement. “3D tracking tells us about the changes in a dragonfly’s wing motion, but relating these changes to the body rotation is a complex dynamical problem,” Wang says. “We take advantage of the fact that the governing laws of flight can be described by equations, and this allows us to make specific predictions. These predictions can be tested in experiments and can be further related back to the other pieces in the puzzle involving neural responses.”

“3D tracking tells us about the changes in a dragonfly’s wing motion, but relating these changes to the body rotation is a complex dynamical problem.”

Wang has also found that dragonflies rely on their compound eyes and the three additional eyes called ocelli on the top of their heads to sense their orientation during free fall. Once the dragonflies realize they are falling, they angle their left and right wings in an asymmetrical pitch. “It’s only for two or three strokes,” Wang says, “but it’s enough to create the aerodynamic forces that rotate their body 180 degrees.”

Insects could right themselves in many ways, but they invariably use the asymmetrical wing angle to initiate the rollover. This implies it is an innate recovery strategy. “Dragonflies are hardwired to do a stereotypical correction,” Wang says, “to right themselves against unpredictable circumstances. Humans have these innate reflexes, too, that allow us to stand and walk in a seemingly effortless way.”

One application of Wang’s research is to use the understanding of dragonfly flight to probe hardwired reflexes in other living systems. “This research has a broad impact on connecting the reflex involving neurons and muscles to the function, in this case flying,” she says. “With flight simulation, we can make quantitative predictions about the neurofeedback circuitry for flight. In some cases, we can give mechanistic explanations of the observed behavior.”

Fruit Flies, Standing in Air

In another project investigating insect flight, Wang uses computer simulations to understand how fruit flies keep their upright posture when hovering in air. “Insects are intrinsically unstable, just as we are when we stand,” Wang explains. “Going from four feet, which is a stable configuration, to standing on two was a big evolutionary problem that bipedal animals had to solve. A lot of neuro-rewiring had to evolve in humans to make that happen. When flying insects hover, it is as if they stand in air. If they don’t have an active sensory feedback circuitry, they will eventually fall.”

A fruit fly beats its wings 250 times a second to stay aloft, but without active feedback control, the flight is unstable. “We can model a fly with and without control on a computer,” Wang says. “If we let the model fly beat its wings the same way every wing beats, it will tumble and fall, succumbing to flight instability. So we asked, ‘How often do they have to sense their body orientation and how fast do they have to adjust the wing motion in order to stay stable?’”

Wang conjectured that the fly adjusts its wing motion every wing beat, or about every four milliseconds, to stay upright. To do that, it relies on the vestigial stubs of its hind wings, called halteres, which oscillate like wings but have now become a gyroscope, sensing body rotation. The haltere sends neural signals to a subset of a group of 17 steering muscles. Earlier researchers had noticed that one of those muscles, the first basalar muscle, has the unusual property of firing every wing beat.

“An aha moment came when I realized that perhaps one of the functions of the first basalar muscle is to maintain flight stability,” Wang says. “All other steering muscles fire when the fly turns, but this one fires all the time.” Wang is testing her theory with fruit flies that have been genetically altered so the motor neurons to their first basalar muscles are silenced. The genetically altered flies tumble and fall, which is consistent with Wang’s predictions.

Wang’s work on insect flight is part of current scientific initiatives to understand the brain. “There has been a great effort to examine the architecture of neural circuitries. For many years to come, a big puzzle will be not only how a neural circuitry is wired and what it does, but why. The fly study is an example of this. We asked, ‘Why does a specific motor neuron fire every wing beat?’ Our model provides an explanation that connects the physics of flight to the design of the circuitry. It offers new evidence of how the physical laws at macroscopic scales shape the function of neural circuitry.”