Bats use echolocation to hunt insects, many of which fly in swarms. In this process, bats emit a sound signal that bounces off the target object, revealing its location. Smaller insects like mosquitos are individually hard to detect through echolocation, but a new Tel Aviv University study reveals that they become perceptible when they gather in large swarms. The findings could provide new insights into the evolution of bat echolocation and explain why tiny insects are found in the diets of bats that seem to use sound frequencies that are too high to effectively detect them. The new research was conducted by Dr. Arjan Boonman and Prof. Yossi Yovel at TAU’s Department of Zoology and colleagues at Canada’s Western University. It was published in PLOS Computational Biology on December 12.

Modeling bat vision

Few studies have addressed what swarms of insects — as opposed to single insects — “look” like to bats. To find out, Dr. Boonman and colleagues combined three-dimensional computer simulations of insect swarms with real-world measurements of bat echolocation signals to examine how bats sense swarms that vary in size and density. They found that small insects that are undetectable on their own, such as mosquitos, suddenly become “visible” to bats when they gather in large swarms. They also discovered that the fact that bats use signals with multiple frequencies is well suited to the task of detecting insect swarms. These signals appear to be ideal for detecting an object if more than one target falls inside the echolocation signal beam at once. “Using simulations, we investigated something that could never have been measured in reality,” Dr. Boonman says. “Modeling enabled us to have full control over any aspect of an insect swarm, even the full elimination of the shape of each insect within the swarm.”

From insects to drones

The insect model the researchers used has a tiny mesh (skeleton) and minuscule legs and wings. “We are still adding new features, such as the bat’s acoustic beam or ears, which were not in the original model,” says Prof. Yovel. “We also developed a faster version of the algorithm. All of this will open a new world for us in which we can get echoes even from entire landscapes, so we can learn what a bat or sonar-robot would ‘see’ much more quickly.” The study could also affect technology being developed to improve defense systems. “The algorithms developed for this study could potentially be applied to radar echoes of drone swarms in order to lower the probability of detection by enemy radar,” Dr. Boonman explains. “Since drones are playing an ever more prominent role in warfare, our biological study could spawn new ideas for the defense industry.”