| Description |
Chapter 1 introduces a classic question from optimal foraging theory regarding space-use strategies of a forager, and gives context for addressing similar questions in groups of foraging ants. Chapter 2 generalizes the marginal value theorem (MVT) model by describing a rate-maximizing forager searching for pointwise resources with a specific searching distribution around previous resource finds, and giving-up value (GUV) strategy at resources. The model shows that the optimal ARS breadth increases, and the optimal GUV decreases, with increased dispersion of the resource distribution. Chapter 3 builds an agent-based model (ABM) and corresponding PDE model derived from an isotropic diffusion limit. The model links individual movement biases in the presence of pheromone to the colony-wide searching distribution. Parameterized with movement data obtained from Tetramorium caespitum (the pavement ant), the model predicts bistability in pheromonal recruitment at resource distances of 3 - 6 m; the onset-distance of bistability increases with colony size. Data collected from the field are used to estimate parameters of the PDE model for T. caespitum in Chapter 4. The ability of T. caespitum to find autocorrelated resources during recruitment is analyzed using a Cox proportional hazards model, the results of which are compared to those predicted by the PDE model developed in Chapter 3. Finally, Chapter 5 develops a simulation to assess the effect of individual trail fidelity on the ability of a colony to capitalize on autocorrelated resources in different resource scenarios; the results suggest that T. caespitum is tuned to exploit large, nonautocorrelated resource distributions. |
