On an early June evening, physicist Raphael Sarfati breathed hard as he lumbered up a dense forest trail in Great Smoky Mountains National Park. The French-born scientist lugged more than 40 pounds of gear, including a tent, generators, butterfly nets and two GoPro 360-degree cameras vital to photographing his subject. Sarfati, a postdoctoral associate at the University of Colorado, Boulder, and his advisor, assistant professor of computer science Orit Peleg, trekked into the forest to film how synchronous fireflies conduct their impressive light display, a show that lasts for just 10-to-15 days each year and only for a few hours each night. Unlike many firefly species that flash in individualized patterns for months every summer, these special fireflies display in a specific, collective pattern that the scientists wanted to track.
With their tent and cameras set up and dusk descending, the sporadic blinking of individual fireflies harmonized into synchronous flashing. “They are everywhere around you. You can’t even count how many there are, all flashing at the same time for a few seconds and then they all stop at the same time as well. It’s dark and then it picks it up again,” Sarfati says. “It’s really astonishing.”
“How do thousands or tens of thousands of individuals all know how to flash at the same time when they can only see a fraction of the insects around them?” Peleg marvels. “There are a lot of interesting aspects of firefly communication, and we’re hoping to shed light on them.”
Now, in a study published in September in the Journal of the Royal Society Interface, Sarfati and Peleg have shown how to recreate the fireflies’ flashes and flight trajectories three-dimensionally. Their findings provide clues into how simple insects with limited cognitive functionality can accomplish complicated, synchronous tasks. By demonstrating how fireflies begin to synchronize, their research might inspire communication and coordination methods in swarm robotics technology. It will also serve as a resource for firefly conservation efforts by providing a more accurate way to monitor their populations.
Sarfati and Peleg had come to Great Smoky Mountains National Park to study Photinus carolinus. The scientists first set up their 360-degree cameras in the forest to capture the insects’ behavior in their natural, unperturbed environment. Male fireflies, thick in the air, flew around and flashed in unison to attract the relatively stationary females waiting on the ground below. Standing in the cloud of Morse-code-like intervals of light, the researchers could see a lone male flashing here or there along with his brethren. However, their cameras tracked what their naked eye could not: trajectories of exactly where individual fireflies were in three-dimensional space when they flashed. By tracking the flashes, the team was able to recreate the flight patterns of each insect caught on camera.
Sarfati and Peleg next set up the tent as their control environment and added dozens of male fireflies to the space—enough to elicit the same swarm behavior found in their natural environment. Then, with cameras rolling inside the tent, Sarfati and Peleg captured firefly behavior on an individual level within the swarm. When those data were taken back to the lab, they clearly saw that individual male fireflies in the swarm flashed for roughly three-to-five of the group’s collective 10-to-15 flashes. “The total number of flashes in a burst where you have a lot of fireflies together could be as many as 10, 12 or 15, but it’s not the same firefly flashing that many times,” Sarfati says. “It’s a relay, passing over the flash.” If firefly A flashes five times, on the third flash firefly B might pick up the rhythm and flash with firefly A for three beats. As firefly B flashes its third beat, firefly C might join and flash with firefly B for three beats, and on and on. The fireflies light up in consistent patterns with a routine six-to-eight seconds of darkness between each flashing sequence.
In the experiment’s next phase, the researchers introduced male fireflies one by one to an empty tent. By doing so, the scientists found that individual males flash in inconsistent intervals when isolated from the swarm. An isolated insect might perform three flashes, then have a two-second dark period, perform four flashes and then have a 30-second dark period. While the swarm of fireflies in the forest flashed together with consistent sequences of light bursts and dark inactivity, the solitary male did not maintain a steady rhythm of illumination. “A single firefly in isolation in the tent would flash at some point, but then it could be a few seconds or minutes later before he flashed again,” Sarfati says.
When a second male was introduced to the tent, the duration of the light and dark periods remained random with each male charting his own course. With five or ten total males in the control environment, the researchers noticed that when one began to flash, another would seem to join the flashing and then continue it like a relay, but the length of the dark period was still inconsistent—sometimes 12 seconds, other times 45 seconds. It wasn’t until there were 15 males together in the tent that the synchronous flashing found in their natural setting occurred, followed consistently by six-to-eight seconds of darkness.
The researchers were witnessing the fireflies’ transition from chaos into order. And they had it on film, where it could be analyzed, reconstructed and graphed. “They have developed a very, very powerful tool for understanding the details of flash synchrony,” says Sara Lewis, an evolutionary and behavioral ecologist at Tufts University who studies fireflies and was not part of the study. “They have also demonstrated, as we knew, that it’s an emergent property. Males can be kind of random when they’re alone, but as they get into bigger and bigger groups, then there’s this emergent property that shows the synchrony is a function of male density.”
Anders Christensen is a professor of bio-inspired robotics at the University of Southern Denmark who was not part of the study. He points out that having a clearer picture of synchrony, especially its emergent properties, through tools like Sarfati and Peleg’s will lead to a better understanding of communication and robustness, two of the self-organization principles guiding swarm robotics.
Christensen strives to design robots to perform tasks individually in the same way that a firefly has the ability to flash on its own, independently of the swarm. Yet, like fireflies, robots operating in a swarm must be able to communicate with and react to each other. The swarm should also be robust, meaning that it can continue to function even if some members break down. Christensen has accomplished communication and robustness in past robotics experiments but believes there’s more work to be done. “If we learn the connection between the microscopic rules that govern the individual firefly and the resulting global behavior of a swarm of fireflies, we can use that insight to design behaviors for robot swarms that require some form of synchronization to carry out a task,” Christensen says.
Another scientist not a part of the study, Tyler MacCready, CEO of swarm robotics data collection management company Apium Swarm Robotics, sees how the new research and the clues to synchronization it unlocks could pertain to his work—eventually. His robots are already capable of autonomous coordination between vehicles in complex, ever-changing environments like the ocean; however, they are still reliant on a human operator to provide group-level instructions. With the technology that may be possible thanks to this research, he hopes to one day be able to send a swarm of robots out to perform complex data collection tasks while just relying on each other, without communication from a human controller.
But while the swarm roboticists dream up pie-in-the sky applications, Lewis is eager to ensure the fireflies will be around long enough to serve as a muse. For her, Sarfati and Peleg’s research is useful now. Since the filming was done using a relatively simple and inexpensive method, citizen scientists associated with the community science project Firefly Watch might be able to use the cameras for research. Volunteers would gather video footage in their backyard or local park and send it back to Firefly Watch, where scientists could run a firefly-density-estimating algorithm described in the new paper to attain a more precise population count. Monitoring how the insects’ numbers fluctuate will provide clues for how to preserve them.
“Fireflies are one of our best ambassadors for Earth’s natural magic. They’re beautiful, mysterious; they inspire hope. They keep us connected to the natural world,” Lewis says. “That’s why we’re working so hard to keep the firefly magic alive for future generations.”