Date:
Location:
Title: Distributed control of microclimate by honeybee colonies, Apis mellifera L.
Abstract: Social organisms often exhibit complex, ordered behaviors at the group level that are not observed in individuals. Such collective behaviors emerge from the iterative interactions between individuals, each of which sense and respond to local stimuli. Many group behaviors (e.g., nest construction or pheromone trails) in social insects arise not from direct interactions between individuals but from distributed interactions with a shared environment, a process known as stigmergy. In classical examples of stigmergy, an individual deposits a cue in the environment, such as a clump of pheromone-laden building material, which in turn influences the behavior of the next individual to encounter the deposition. These local manipulations lead to self-organization of large scale patterns such as complex nest architecture. In these examples, the interactions between individuals and the environment are local. However, stigmergic behaviors can also involve interactions with a dynamical physical processes in the environment such as airflow, elastodynamics or heat transfer. In such instances, non-local physical interactions play a role. In this thesis I use both experimental and computational approaches to explore a class of behaviors in honeybee colonies in which bees collectively modulate large scale physical processes through local manipulations of their shared dynamical environment. In Chapters 1 and 2, I investigate how temperature-dependent wing fanning behaviors in honeybees give rise to efficient large-scale flows during nest ventilation. In Chapter 3, I explore how honeybee swarm clusters change their collective morphology in order to maintain stability when exposed to dynamic mechanical perturbations. In Chapter 4, I explore how swarm clusters modulate their morphology to maintain thermal stability despite large fluctuations in ambient temperature. Taken together, this work suggests honeybees can coordinate large-scale physical processes through distributed local interventions. This may be a ubiquitous strategy in the evolution of complex systems.
Committee: L. Mahadevan (Advisor), Andrew Biewener, Benjamin de Bivort, George Lauder