Item

Abstract

Roughly one hundred million years ago, solitary insect species evolved social interactions that enabled the formation of colonies. A main reason for this advance was their ability to feed each other with previously ingested food. Among other things, this allowed them to develop the well-known division of labor: groups or castes of individuals specializing in certain tasks. This social organization reached its climax in the evolution of non-reproductive castes, sacrificing their own reproduction to the benefit of the colony. The mutual feeding technique that supported this social evolution is called ‘trophallaxis’. This thesis is based on the question how ant colonies use trophallaxis to supply their members with food. The main goal of this thesis is to understand the collective properties of the food distribution in ant colonies with the simplest possible computational and analytical models. To this end, we construct a series of biophysically motivated simulation models and analytical descriptions of trophallaxis that include all its essential features. Our models are the first complete theoretical description of the physical mechanisms behind the self-organized food distribution in ant colonies. Despite our reductionist approach, the models exhibit a number of interesting properties that reproduce some of the behaviors seen in real ant colonies. We are confident that our models can serve as benchmarks for the behavior of real ant colonies or more biologically detailed models. As statistical null models, they can be used to assess to what extent an observed behavior is due to non-random strategies or due to the collective properties of a stochastic system. We find and analytically predict the characteristic time scales of trophallaxis for both well-mixed colonies and colonies with small spatial fidelity zones. We even successfully cover the range between these two limits with semi-analytic predictions. These newly discovered relationships between individual behavior and global food distribution dynamics provide microscopic explanations of experimental observations and phenomenological theories that were unknown so far.