Researchers have developed a robotic model that offers a mechanistic explanation for how a tropical bat species locates silent insects resting on leaves using echolocation, addressing a longstanding question about bat foraging in acoustically complex environments. The work provides experimental support for the so-called acoustic mirroring effect, in which leaves function as reflective surfaces that help bats identify prey without directly sensing the insect itself.
The study, conducted by scientists from the Smithsonian Tropical Research Institute, the University of Cincinnati, and the University of Antwerp, examined the hunting behavior of the common big-eared bat, Micronycteris microtis. This species forages in the dense understory of tropical forests, where echoes from vegetation create significant acoustic clutter and many prey insects remain motionless and silent.
Previous behavioral observations suggested that these bats emit echolocation calls toward leaves from an oblique angle, similar to viewing a mirror from the side. When a leaf is unoccupied, the sound reflects away from the bat, producing little or no returning echo. If an insect is present, part of the sound reflects off the three-dimensional body of the prey and returns to the bat, indicating a potential food source. A central challenge to this hypothesis was explaining how bats could select appropriate approach angles without first determining the precise orientation and position of each leaf, a process that would require additional time and energy.
To test whether such a strategy could function without explicit measurements of leaf geometry, the researchers built a robot equipped with ultrasonic sensors. The robot was programmed to emit sound pulses and to move in response to returning echoes from cardboard leaves arranged in a test environment. One leaf included a dragonfly model, while others were left empty. The robot did not assess leaf size or orientation, instead relying solely on changes in echo strength as it moved.
Using this approach, the robot correctly identified the leaf with the dragonfly model in most trials, while producing fewer false detections on empty leaves. Data collected during the experiments showed that echoes from smooth, unoccupied leaves rose and fell sharply as the robot changed its angle of approach, whereas echoes from leaves with the insect model remained comparatively stable across angles. This stability allowed the robot to focus on prey-bearing leaves without prior knowledge of their orientation.
The findings suggest that bats may rely on consistent echo patterns from prey rather than detailed spatial mapping of leaves, enabling efficient foraging in cluttered habitats. According to the researchers, the results support the idea that the interaction between an animal’s movement, its sensory emissions, and environmental structures can simplify complex perceptual tasks.
Beyond advancing understanding of bat behavior, the work may have implications for the design of bio-inspired sonar systems. Similar principles could potentially be applied to technologies intended to detect objects such as fruit or pests within vegetation, where traditional sensing methods face limitations.
Photo credit: Steven Paton, Smithsonian Tropical Research Institute
