The Spatial Resolution of Bat Biosonar Quantified with a Visual-Resolution Paradigm

Bats are navigation super-performers, flying at high speed through nocturnal forests. Numerous field observations and formal experiments have impressively shown how well bats tackle navigation in 3D with biosonar, i.e., the auditory analysis of self-generated ultrasonic emissions [1-7]. However, unlike in the visual system, where space is explicitly coded at very high resolution in the retinal fovea, the inner ear encodes frequency and time, not space. Spatial attributes of echoes are represented in the space-dependent filtering of the bats' pinnae [8, 9] and binaural computations, like interaural time and level differences [10, 11], as first proposed by Lord Rayleigh [12]. Remarkably, Rayleigh also provided a clear definition of spatial resolution: based on the shape of optical diffraction patterns arising from two closely spaced light sources, Rayleigh defined resolution as the capability to detect a trough in their joint light diffraction patterns [13, 14]. Here, we recruit Rayleigh's classical resolution paradigm to quantify how well bats can resolve multiple simultaneously presented reflectors in space. We show that biosonar spatial resolution in azimuth is no better than about 80° compared to a human visual resolution down to 0.02° [14]. We suggest that bats compensate this effective lack of spatial resolution by sequentially probing their environment in flight. Our data show that low-resolution environment perception is a viable alternative to high-resolution vision to support intelligent behavior in complex environments.