Department of Zoology, University of Oxford
Anim Behav. 1997 Jun;53(6):1129-42. doi: 10.1006/anbe.1996.0388.
Risk-sensitive foraging theory is based on the premise that unpredictable runs of good or bad luck can cause a variable food source to differ in fitness value from a fixed food source yielding the same average rate of gain but no unpredictability. Thus, risk-sensitive predictions are dependent on the food intake from variable sources being not only variable but also unpredictable or 'risky' in outcome. This study tested whether unpredictability is a component of the value that foraging starlings, Sturnus vulgarisattribute to food sources that are variable in the delay to obtain food. Two groups of birds chose between a fixed and a variable delay option; the variable option was unpredictable in the risky group and predictable in the risk-free group in the overall rate of intake it yielded. In both groups the fixed option was adjusted by titration to quantify the magnitude of preference for predictable and unpredictable variance. On negative energy budgets both groups were significantly risk-prone, with the risky group being significantly more risk-prone than the risk-free group. Switching the birds to positive budgets by doubling the size of each food reward had no significant effect on preference, and similar trends to those found with negative budgets were observed. These results are not readily explained by risk-sensitive foraging theory, but may be explained by the algorithm used by the birds to attribute value to average expected rewards.
风险敏感觅食理论基于这样一个前提,即不可预测的好运或坏运会导致可变食物源的适应性价值与固定食物源不同,后者虽然平均收益相同,但没有不可预测性。因此,风险敏感预测取决于可变食物源的摄入量不仅是可变的,而且是不可预测的或“有风险的”。本研究测试了不可预测性是否是觅食椋鸟(Sturnus vulgaris)对食物来源的价值的一个组成部分,这些食物来源在获得食物的延迟方面是可变的。两组椋鸟在固定延迟和可变延迟选项之间进行选择;在整体摄入量方面,可变选项在风险组中是不可预测的,而在无风险组中是可预测的。在两组中,通过滴定调整固定选项,以量化对可预测和不可预测方差的偏好程度。在负能预算下,两组都有明显的风险倾向,风险组比无风险组更有风险倾向。通过将每个食物奖励的大小增加一倍,将鸟类切换到正能预算对偏好没有显著影响,并且观察到与负能预算相似的趋势。这些结果不易用风险敏感觅食理论解释,但可以通过鸟类用来赋予平均预期奖励价值的算法来解释。