MicroAnts 2.5 is a general-purpose agent-based model designed as a flexible workhorse for simulating ecological and evolutionary dynamics in artificial populations, as well as, potentially, the emergence of political institutions and economic regimes. It builds on and extends Stephen Wright’s original MicroAnts 2.0 by introducing configurable predators, inequality tracking, and other options.
Ant agents are of two tyes/casts and controlled by 16-bit chromosomes encoding traits such as vision, movement, mating thresholds, sensing, and combat strength. Predators (anteaters) operate in static, random, or targeted predatory modes. Ants reproduce, mutate, cooperate, fight, and die based on their traits and interactions. Environmental pressures (poison and predators) and social dynamics (sharing, mating, combat) drive emergent behavior across red and black ant populations.
The model supports insertion of custom agents at runtime, configurable mutation/inversion rates, and exports detailed statistics, including inequality metrics (e.g., Gini coefficients), trait frequencies, predator kills, and lineage data. Intended for rapid testing and educational experimentation, MicroAnts 2.5 serves as a modular base for more complex ecological and social simulations.

This is a NetLogo version of Buhl et al.’s (2005) model of self-organised digging activity in ant colonies. It was built for a master’s course on self-organisation and its intended use is still educational. The ants’ behavior can easily be changed by toggling switches on the interface, or, for more advanced students, there is R code included allowing the model to be run and analysed through RNetLogo.

Positive feedback can lead to “trapping” in local optima. Adding a simple negative feedback effect, based on ant behaviour, prevents this trapping

The model explores how two types of information - social (in the form of pheromone trails) and private (in the form of route memories) affect ant colony level foraging in a variable enviroment.

Simulates impacts of ants killing colony mates when in conflict with another nest. The murder rate is adjustable, and the environmental change is variable. The colonies employ social learning so knowledge diffusion proceeds if interactions occur.

The mode implements a variant of Ant Colony Optimization to explore routing on infrastructures through a landscape with forbidden zones, connecting multiple sinks to one source.

Ants in the genus Temnothorax use tandem runs (rather than pheromone trails) to recruit to food sources. This model explores the collective consequences of this linear recruitment (as opposed to highly nonlinear pheromone trails).