Task Selection in Honeybees - Experiments Using Multi-Agent Simulation

One of the most interesting questions about social insects is how task allocation is controlled. Often specialization of individual animals occurs; there exist different hypotheses to explain this phenomenon. By using multi-agent simulation we want to gain evidence for the hypothesis that in honeybees (Apis mellifera) the fitness of a colony increases with increased variation of individual response thresholds, which would result in specialization. Based on a rather complete model of activities performed in a bee hive we conducted simulation experiments using different kinds of variation in task selection parameters. According to global measures of colony performance we showed that merely by variation of individual disposi-tions to engage in certain tasks the effectiveness of the hive increases .