Foraging strategy of a non‐omniscient predator in a changing environment (II) model with two data windows and a relative comparison criterion

Abstract

Summary The model with two data windows and the relative comparison criterion (SRC) is proposed based on a Dynamic Programming problem. The model deals with predators whith remember separately the profitability of each patch and visit patches repeatedly. The model predator can remember foraging results that have occurred within the duration of memory (Dm). But when the number of foraging results in one patch exceeds the capacity of the memory chamber (Cm), older results are deleted from memory. Cm is the short data window for patches searched recently and Dm is the long one for patches searched in the past, but not recently. The model predator always searches at the patch with the highest estimation of profitability. The validity of SRC was checked by the experiment with great tits ofSmith andSweatman (1974). SRC could describe satisfactorily weil the experimental result that great tits at first searched almost uniformly and then concentrated their foraging efforts to good patches but they could not detect the new good patch after changes in food distribution. These analyses showed that great tits did not forage in an optimal manner in the experimental conditions ofSmith andSweatman (1974). Differences of SRC from the exploration and exploitation model ofKrebs et al. (1978) and from the model with one data window and the absolute criterion (SAC) of Inoue (1983) were discussed to show characteristics of the two extreme strategies for sampling the patchy environment.