Rate currencies and the foraging starling: the fallacy of the averages revisited

In classical optimal foraging models long-term rate of energy intake (the ratio of expected amount of food over expected time) is assumed to be die maximized currency, because doing this is consistent with minimizing die loss of alternative opportunities. Here this possibility and various alternatives are examined quantitatively using European starlings (Sturnus vulgaris) in the laboratory. The birds chose between two cues. One signaled an option Uiat led to either a fixed delay to food ("single standard," Experiment 1) or to one of two equally probable delays to food ("double standard," Experiments 2 and 3). The other cue signaled an "adjusting option" consisting of a single delay to food. This option adjusted according to the previous choices made by the birds, improving when the standard had been preferred and worsening when the adjusting option had been preferred. Adjustments were made either by changing the delay to food or the amount of food. The rationale underlying this procedure was that the parameter values at which the adjusting option stabilizes should reflect the subjective value of the standard. This was validated in Experiment 1. In Experiments 2 and 3 die adjusting option fluctuated around parameter values that are interpreted as yielding subjective equivalents of the double standard. The results contradict the predictions based on minimizing the lost opportunity. First, the birds did not include all the time intervals in their assignment of value to the two options, and second, die birds used die expected ratio of amount over time rather dian the ratio of expected amount over expected time as their rate currency.