Abstract
1. Animals are often required to make decisions about their use of current resources while minimising travel costs and risks due to uncertainty about the forthcoming resources. Passive soaring birds utilise warm rising-air columns (thermals) to climb up and obtain potential energy for flying across large areas. However, the utilisation of such inconsistent natural resources may be challenging for soaring-gliding birds and involve a set of decisions to maintain efficient flight. 2. To assess which temporal scales of previous experience with environmental inputs best predicted thermal-climbing departure decisions of soaring birds, we used movement data from Eurasian griffon vultures (Gyps fulvus) tracked by GPS transmitters. We applied Cox proportional hazard regression and a model selection approach to identify thermal-climbing departure decisions and to compare a range of temporal scales. 3. Our findings support the use of current and recent (short-term; last 20 min) experiences, compared to longer term, past experiences, in predicting the time until departure from thermals. The models supported decision rules that integrated information originating from different temporal scales, implying a tendency to depart from a thermal later when the current climb rate was higher than experienced recently and vice versa. In addition, climb rates in thermals revealed significant autocorrelation over short time-scales (shorter than 30 min). 4. The correspondence between thermals' characteristics and the factors that best predicted thermal-climbing departure decisions presumably reflects optimal decisions individuals make to handle their dynamic environment and to reduce movement-related costs of such a basic activity for soaring-gliding birds.
