Gain curves in depletable food patches: A test of five models with European starlings

A forager’s gain curve, the cumulative number of prey harvested from a patch as a function of time spent in the patch, influences optimal patch departure rules and interpretations of patch use data. We describe models of five different search strategies that yield different gain curves. Hence they would influence a forager’s decision for patch departure differently and, consequently, how researchers should interpret patch residence times and giving-up densities. However, the models are virtually impossible to separate based on data of the gain curves per se. Therefore, we develop a series of diagnostic tests that can be used to discriminate among models. These tests consider how the instantaneous harvest rate within patches depends on initial (IPD) and current prey density (CPD) and search time. We applied these tests to data collected from European starlings (Sturnus vulgaris) foraging in experimental food patches of known initial prey density. The starlings’ harvest rate increased with CPD, an indication of diminishing returns. However, a given CPD yielded a lower instantaneous intake rate the higher the IPD. Thus, the two models most commonly assumed in foraging studies, systematic and random search, can be unequivocally rejected. Instead, we found support for a new model, negative stirring, in which the starlings spoil their own future foraging returns by aggregating the remaining prey items as they search.