Document Type

Dissertation

Degree

Doctor of Philosophy

Major

Physics

Date of Defense

4-3-2018

Graduate Advisor

Sonya Bahar

Co-Advisor

Thomas Vojta

Committee

Ricardo Flores

Paul Parris

Zuleyma Tang-Martinez

Abstract

This work examines the possibility of evolving the phenotypic specialization associated with division of labor in an agent-based model without task-switching costs. The model examines two groups competing for vital resources, where members of one group are capable of sharing resources with other agents in their group. Agents attempt to collect resources which allow them to reproduce, with more resources leading to a greater number of offspring by asexual reproduction. Four variants of the model are examined, with combinations of one or two resources and the presence of a foraging risk. The presence of the foraging risk can lead to agents in the sharing group specializing in each trait but, by looking at the fraction of foragers per generation, this event appears to be a transient state. Division of labor is quantified by calculating the normalized mutual entropy, and is shown to be higher when a population contains agents which specialize on different tasks.

Share

COinS