Simple cellular automata to mimic foraging ants submitted to abduction

TL;DR

This paper develops cellular automata models to simulate ant foraging behavior under localized threats, exploring how danger information transmission affects their movement, with parameters based on real experiments, and suggests ants may not transmit danger info.

Contribution

It introduces simple cellular automata models for ant foraging under threat and tests different danger information transmission protocols based on experimental data.

Findings

Ant behavior best explained by no danger information transmission

Models are parameterized with real experimental data

Framework allows testing danger communication hypotheses in future experiments

Abstract

Many species of ants forage by building up two files: an outbound one moving from the nest to the foraging area, and a nestbound one, returning from it to the nest. Those files are eventually submitted to different threats. If the danger is concentrated at one point of the file, one might expect that ants returning to the nest will pass danger information to their nestmates moving in the opposite direction towards the danger area. In this paper, we construct simple cellular automata models for foraging ants submitted to localized abduction, were danger information is transmitted using different protocols, including the possibility of no transmission. The parameters we have used in the simulations have been estimated from actual experiments under natural conditions. So, it would be easy to test our information-transmission hypothese in real experiments. Preliminary experimental results published elsewhere suggest that the behavior of foraging ants of the species Atta insularis is best described using the hypothesis of no transmission of danger information.