We conducted movement analysis using ant experiments and simulation models. In our experiments, Japanese carpenter ants showed Lévy-like movements when they were exposed to a bowl-shaped test container filled with other foragers. We also developed a one-dimensional multi-agent random walk algorithm. In our algorithm, agents interact with each other and change their probabilities of movements by locally anticipating other agents' moving directions. Agents also modulate those probabilities on the basis of local information, thereby producing fluctuations in averaged local directional information. We were able to induce agents in our model to achieve Lévy-like movements. We also found that an agent occasionally moved or stayed together with another particular agent for a substantial duration in both experiments and simulations.
- Ant movement strategy
- Multi-agent model
ASJC Scopus subject areas
- Ecology, Evolution, Behavior and Systematics
- Ecological Modelling