Can a bat catch prey on a mirror? A bat’s expert foraging skills revealed using a robot

Scientists built a robot to help explain how a tropical bat spots insects perched on leaves using echolocation, a highly sophisticated behavior that requires precise, split-second decision making on the part of the hunting bat. In a study published today in the Journal of Experimental Biology, a bat researcher at the Smithsonian Tropical Research Institute (STRI) teamed up with two robotics engineers from the University of Cincinnati and the University of Antwerp, respectively, to provide the first plausible explanation for how the common big-eared bat (Micronycteris microtis) can efficiently determine whether a leaf is occupied by a silent insect amidst the dense, cluttered understory of the tropical rainforest using only sound.

Co-author Inga Geipel, a research associate at STRI, previously suggested that M. microtis detects silent prey by scanning leaves with sound from an angle, akin to a human viewing a mirror from the side, and listening for an echo that indicates prey is there. If the leaf is unoccupied, then the bat’s sounds will bounce off the leaf and away from the bat, and the bat will not hear an echo. If the leaf is occupied by prey, then most of the echo will reflect away from the bat, but some will bounce off the prey and back to the bat, signaling it has found a meal.

But Geipel’s theory suggests that for the bat to know the angle at which it should approach a leaf to be able to detect prey, it first needs to know the orientation and position of each leaf. Taking those measurements, however, would cost the bat significant time and energy, so the researchers proposed that the bats can zoom in on the interesting leaves merely by taking advantage of the steadiness of the echoes originating from their prey perched upon them.

“I always have been amazed how these small animals can not only navigate the complex entanglement of a forest understory in complete darkness, but also how they find and catch tiny insects with an incredible accuracy, using their own sounds,” Geipel said. “Still, we only have a limited understanding of how bats hunt in this crowded environment, and our study helps explain how they are able to accomplish this challenging task.”

Using a Robot To Model Foraging Behavior

To model the bats’ hypothesized foraging technique, the research team built a robot and programmed it to emit ultrasonic signals and follow the echoes from carboard leaves—without measuring the size or orientation of the leaves themselves. The robot randomly explored a set of these model leaves, one of which featured a fake dragonfly, until it detected an echo, and it moved in the direction of the echo. If the echo got too weak—which in nature, would indicate the leaf is unoccupied—the robot would move on.

“Behavioral experiments had already suggested how these bats might solve the problem of finding prey-occupied leaves, but we wanted to know whether that explanation was actually sufficient to make the behavior work,” said Dieter Vanderelst, the paper’s lead author and an associate professor of biology, mechanical engineering and electrical engineering at the University of Cincinnati. “By building the bat’s hypothesized foraging strategy into a robot and testing it in the physical world, we could ask whether a simple, elegant solution can succeed under complex acoustic conditions.”

Using the described algorithm, the robot was able to distinguish between leaves occupied by the dragonfly model and vacant leaves. It successfully detected the dragonfly 98% of the time and wrongly signaled the presence of prey on unoccupied leaves only 18% of the time. As it swept across the leaves, the robot collected data that demonstrated how it made the distinction: If a leaf was unoccupied, the echo volume peaked and dropped quickly as the robot approached the smooth leaf from different angles, while an occupied leaf produced stable echoes regardless of the angle, since the three-dimensional insect reflected the echo in many different directions. The robot was most accurate when approaching the leaf from the angles at which these bats typically approach leaves in nature. This model demonstrates that bats could detect prey using a leaf as a mirror without having to determine the leaf’s position and orientation first.

“What fascinates me most is that by flying at a well-chosen height, the bats can use leaves as mirrors to automatically restrict their attention to particular leaves of interest,” said Herbert Peremans, the study’s senior author and a professor of robotics at the University of Antwerp. “This example nicely shows that nature doesn’t evolve components; it evolves systems. It is the dynamic interaction between the bat and the environment—through SONAR—that makes this hunting behavior work.”

These data expand the understanding of how M. microtis and other bats that catch prey from surfaces, can forage so efficiently. The method by which they scan leaves for prey could potentially help inform the design of new SONAR systems for use in agriculture, such as for the detection of fruit in trees or pests on crops—designs that translate foraging efficiency in bats into efficiency in food production for humans and other animals.

About the Smithsonian Tropical Research Institute   

Headquartered in Panama City, Panama, STRI is a unit of the Smithsonian Institution. Our mission is to understand tropical biodiversity and its importance to human welfare, to train students to conduct research in the tropics, and to promote conservation by increasing public awareness of the beauty and importance of tropical ecosystems. Watch our video, and visit our website, Facebook, X and Instagram for updates.