Krebs J.R. & Cowie R.J. (1976) Foraging strategies in birds. ARDEA 64 (3-4): 98-116

We describe some recent studies aimed at testing the predictions of optimal foraging theory models. These models are based on the premise that natural selection has designed the behaviour of predators for the goal of maximising the net rate of food intake during a bout of foraging. The two aspects of predator behaviour which we discuss are allocation of time between different areas, and choice of food items within an area. The theory of time allocation is developed from the fact that predators gradually deplete the food in an area and so have to move at some stage to anew patch. The optimal criterion for moving (the giving up threshold) predicted by the model has been tested in laboratory experiments with titmice, in which we manipulated the birds' average success either by altering the quality of patches or by changing the travel time. The experimental results approximated the predictions of the model, although in two other studies of time allocation in birds, the results were less close to the optimal prediction, perhaps because the birds in these experiments sampled the environment. We also discuss a study in which changes in the quality of patches did not result from depletion by the predator, but from external events. Again the predators (Spotted Flycatchers) seemed to approximate an optimal criterion for moving. The experiments on prey selection tested the extent to which captive Great Tits will ignore unprofitable prey at different rates of encounter with more profitable types. As predicted by the models of optimal choice, selectivity increased with increasing abundance of profitable prey, but did not depend to any extent on the encounter rate with the less profitable prey. Some field evidence on Redshanks supports this conclusion.