Individual-based models provide a framework for understanding individual and collective behavior of a complex system, such as social insects. We applied an individual-based simulation model of foraging of the western harvester ant ( Pogonomyrmex occidentalis) to two environments, shrub-steppe and shortgrass-steppe, in which the movement characteristics of individual ants differed according to vegetation structure. Using regular, clumped, and fractal seed dispersions, we simulated the influence of environment and foraging speed on rates of seed harvest by a colony of ants. The relative importance of individual and co-operative foraging behavior, where ants use spatial memory or chemical recruitment, was also explored in conjunction with environment and seed dispersion. Our replicated modeling "experiment" permitted us to study the three-way interaction among environment, seed dispersion, and behavior in the rates of seed harvest by ants. When foraging individually or co-operatively at observed movement rates, ants had similar harvest rates of seeds in a regular dispersion. In contrast, ants rapidly removed seeds from a fractal dispersion of seeds when foraging co-operatively. This suggests that differences in ant behavior were less important to the rate of food intake when foraging in a homogeneous resource dispersion, whereas co-operative foraging was most responsive to food resources that were spatially correlated over a range of scales. A slight difference was found between observed and modeled distributions of ant foraging distances in the shrub-steppe environment, while observed and modeled distributions were not significantly different in the shortgrass-steppe. Thus, our simulation model captured many of the foraging properties of individual ants, although it did not always accurately describe the spatial distribution of foraging ants. Our application to P. occidentalis shows that processes operating in a social unit, such as the rate of food intake by an ant colony, are strongly influenced by the interplay of the environment, resource dispersion, and the behavioral responses of individuals to food resources. This interaction arises because of the spatial properties of both foraging ants and seed resources, which suggest that spatially explicit modeling approaches may yield different insights into foraging than those based on energetic criteria.
Read full abstract