Competition and Food Selection: Field Tests of a Theory

Abstract

A graphical model describing the optimal choices of two species competing for resources in two types of habitats is tested with hummingbirds in the field. In this model, both species prefer taking resources from one of the habitat types. However, one of the species (the dominant), by virtue of interference competition, can gain access to the better patch more readily than the other (the subordinate). The model begins with the result of single—species optimal foraging models: at low densities of birds, only the better patch type should be selected, but as density increases, both should be used. Interspecific competition should not lead to qualitatively different behaviors for the dominant species because the effects of the subordinate are weak. For the subordinate, however, there is a third class of behaviors; under the pressure of high densities of the dominant, the subordinate may totally avoid the better patch and use only the poorer. We tested the validity of the model's predictions using three species of hummingbirds: Blue—throated (Lampornis clemenciae), Rivoli's (Eugenes fulgens), and Black—chinned (Archilochus alexandri) coming to feeders containing 1.2 mol/L or 0.35 mol/L sucrose solution. There were no detectable effects of Rivoli's on Black—chinned or vice—versa. This allowed us to test the model's predictions about the interactions between the other pairs of species. Blue—throateds were dominant in both cases. The amount of time birds spent feeding (or in the case of Blue—throateds, feeding and defending) was analyzed, rather than actual densities. Wide ranges of feeding and defending times were obtained using species removals and natural seasonal changes in densities. These two pairs of species exhibited all the model's features. In particular, the dominant's aggressive behavior forced the subordinate to restrict itself to a patch type known to be inferior.