Ant Foraging: Optimizing Self-Organization as a Solution to a Travelling Salesman Problem.

Foraging ant colonies often face the challenge that food items may appear unpredictably somewhere within their territory. This is analogous to Travelling Salesman/Salesperson problems (TSP), wherein solutions seek the least costly and most effective route to visit multiple possibly-rewarding sites. However for ants, TSP solutions are likely also constrained by cognitive limitations. Rather than envisioning entire routes, ants more likely determine their paths by individual-level responses to immediate stimuli, such as presence of other foragers or avoiding revisiting an already explored path. Thus, simple individual-level movement rules could self-organize complex group-level search patterns. Here we derive solutions through agent-based models that optimize net foraging gain for groups of eight agents with ant-like cognitive abilities in searching three different spatial networks of sites. We then compare the patterns from the evolutionary simulations to observed foraging in Argentine ants ( Linepithema humile ) in identical spatial networks. The simulations and ant data show that foraging patterns are sensitive to both network arrangement and predictability in food appearance. The modeling results are consistent in multiple ways with observed ant behavior, particularly in how network arrangements affect foraging effort, food encounters, and general searching distributions. In some distributions, however, ants are more successful at finding food than the simulated agents. This may reflect a greater premium on encountering food in ants versus in simulations increasing exploitation rate of found food. Overall, the results are encouraging that evolutionary optimization models incorporating relevant ant biology can successfully predict the expression of complex group-level behavior.