Computer Science ETDs

Publication Date

Spring 4-25-2017

Abstract

Communication is an important factor in the foraging performance of social insects, such as ants. During foraging, ants keep track of food sources by using memory (site fidelity) or by communicating through pheromones. Previous field experiments showed that the rate of seed collection depends on the distribution of food in the environment. If food is spatially clustered, then it is beneficial for ants recruit nest mates to collect seeds from large clusters. However, we do not know when the recruitment occurs in natural ant population. To explore this question, we used a power law distribution to arrange seeds in piles of different sizes. We observe that simulated ants use pheromone more when the food sources are clumped, and they re discover the pile more frequently if the food source is large. Simulated ants don’t use the pheromone to recruit from the food source that is scattered in the environment. We use simulations to determine how to correlate change points with recruitment events, and then use that relationship to infer recruitment event in in field data. We also observed that ants may repeatedly lose track of found piles and then re-find them. Using change point analysis on seed intake time series, we were able to trace the discovery of piles by detecting changes in the foraging rate.

Language

English

Keywords

Ant, SWARM, Computational Biology

Document Type

Thesis

Degree Name

Computer Science

Level of Degree

Masters

Department Name

Department of Computer Science

First Committee Member (Chair)

Melanie E. Moses

Second Committee Member

Abdullah A. Mueen

Third Committee Member

Tatiana P. Flanagan

Share

COinS