Spatial priors affect sensory weighting in navigation and landing in Egyptian fruit bats

Within the framework of Bayesian navigation models, spatial priors weigh the probability of encountering the locations of obstacles, landmarks and goals. Past work has evaluated the role of spatial priors in human and animal navigation; however, the contribution of moment-to-moment spatial attention in navigation performance in different sensory environments has not been considered. Here, we leveraged the directional aim of the Egyptian fruit bat's sonar clicks to monitor its spatial attention during navigation in light and dark environments. Priors were established by training bats to fly to a perch at a fixed location. Expectations were violated on intermittent trials by displacing the perch 15 or 30 cm. Bat flight/echolocation behavior was quantified with high-speed video and microphone array recordings. In the light, with access to vision and echolocation, bats successfully landed on the perch after it was moved 15 cm but often failed after the perch was displaced 30 cm. In the dark, using echolocation alone, bats often failed to find the perch after it was moved only 15 cm, indicating that spatial priors were weighted more heavily when vision was excluded. When bats failed to land in perch-displaced trials in the light and dark, priors guided them to direct their sonar attention to the region of the room where they expected the perch. Bat navigation failures share characteristics with inattentional blindness in humans, where priors interfere with stimulus processing. Our research demonstrates the effect of priors on spatial attention, which differentially affects sensory weighting in light and dark environments.