Fig. 1

Bats adjust their foraging decision-making according to environmental volatility. In all panels, the boxes represent the area between the first quartile and the third quartile, with the median lines marked inside the boxes and the means marked with X. The lines from the boxes extend to the lowest/highest data points within 1.5 of the distance between the upper and lower quartiles. Circles and triangles represent females and males respectively. In panels A-C: stable: in nights 1-2: n= 14 bats, an in nights 3-4: n= 13 bats; volatile: n= 9 bats. A Both environment and night number had a significant effect on the learning rate (nights 1–2: n = 11 males, 12 females, nights 3–4: n = 11 males, 11 females). B Bats in the stable environment were more successful than bats that experienced the volatile environment. C The nightly reward effect on the stay probability for the volatile and the stable environments. D Bats that experienced the stable environment increased their success rate from the first to the second night, success rate decreased on the third night, and increased again in the fourth night. Success rate of bats that experienced the volatile environment did not change over the nights and was always close to chance (0.5). E Left—Stable environment simulations. A learning rate of 0.4 gave the optimal success rate. Right—Volatile environment simulations. A learning rate of 1 gave the optimal success rate. The estimated learning rate of the real bats is depicted by a dashed vertical line in both panels with the success rate of the bats shown as a red dot