Steering by echolocation: a paradigm of ecological acoustics

1. Flights of three big brown bats (Eptesicus fuscus) landing on a hand and catching a suspended mealworm were video analysed.
2. Results were consistent with the bats using the same basic control procedure in the quite different approach tasks — namely keeping τ(r) = k _rand τ(a)τ(r) = k _αr. Here r is the current distance to the destination; α is the angle between the current direction of the destination and the goal direction of final approach (β _min); τ(r) = r/r, $\tau (\alpha ) = \alpha /\dot \alpha $ ; and k _r, k _αr are constants.
3. The bats were each quite consistent on a particular task (hand or mealworm) in the values they used for the control parameters k _r, k _αrand β _min. However, different values were used in the two tasks, which reflected the different behaviour required at the destination. Flights to hand required twisting and landing upside down and approach angle β _min was closer to vertical and k _rwas smaller and corresponded to decelerating nearly to a stop. In contrast, the mealworms were caught in mid flight and approach angle β _min was shallower and speed of approach was about constant.
4. τ(r) might be registered acoustically by τ(echo-delay) or by τ(echo-intensity). τ(α) might be registered by the bat's directional hearing and gravity sense.
5. The bats learned the tasks easily, suggesting that the control procedure they used in the experiments was part and parcel of the natural skills they had developed in the wild.