Semester

Fall

Date of Graduation

2010

Document Type

Thesis

Degree Type

MA

College

Eberly College of Arts and Sciences

Department

Geology and Geography

Committee Chair

Timothy Warner.

Committee Co-Chair

James B. McGraw

Committee Member

Jamison Conley

Committee Member

Rick Landenberger

Abstract

This thesis investigated the potential for integration of remotely sensed and GIS data into an agent-based modeling environment in order to model seed dispersal and subsequent establishment of windborne seeds. In order to explore the applicability of agent-based modeling to predicting seed dispersal, a case study was carried out using the representative example species Ailanthus altissima, an invasive tree found throughout North America's temperate regions. Seed movement was modeled in two stages, primary and secondary dispersal; primary dispersal was calibrated using existing field data, while secondary dispersal was calibrated only qualitatively. Establishment potential was accounted for probabilistically, based on landuse type. Environmental controls on seed movement and establishment were accounted for with several remotely sensed datasets. The general model characteristics and structure are representative of a potential class of predictive models that incorporate raster data and vector-based seed movement. Agent-based modeling provides a link between raster and vector data and processing methods, and is therefore a potential tool for projects involving both raster and vector data types as well as vector processing. Because seed dispersal and establishment modeling benefits from incorporating both of these data types, it was found that the agent-based approach provided an appropriate framework for modeling the phenomenon, while further research is necessary to fully parameterize and field-validate the model.

Share

COinS