What's New
Research & Projects Publications
People
Data
Links
UCSB Biogeography Lab Publications Abstracts
The Predictive Modeling of Endangered Plant Species in the Santa Monica Mountains Using a Knowledge Base Approach
Noah Charles Goldstein, 2000
MA Thesis, Department of Geography, University of California, Santa Barbara. 102 pp.The National Park Service's Santa Monica Mountains National Recreation Area (SMMNRA) is a unique ecological reserve surrounded by extensive and expanding urbanization. It is home to many rare and endangered species including a number of narrowly endemic taxa. In collaboration with SMMNRA scientists, we developed an ecological knowledge base which can be tested, changed and rendered in a Geographic Information System (GIS). The knowledge base, which represents a predictive model of endangered plant species habitat, is designed to be an aid to species reconnaissance efforts, ecological research and related management decisions. In this study, the SMMNRA was divided into 27,590 Habitat Assessment Units (HAU) that represent landscape facets that were used as the unit of analysis. The test species for this study were the narrow endemics, Dudleya cymosa subspecies complex, and Pentachaeta lyonii.
The predictive model was a modified classification tree, spatially rendered in Arcview GIS. Using the software package Ecosystem Management Decision Support (EMDS), the Boolean rules of the classification tree were modified to reflect the misclassification of the species presences and to incorporate missing data. The modified rules were parameterized using methods that incorporate fuzzy logic. Both unmodified and modified classification trees were evaluated and analyzed. The results of the predictive model identified 15 out of 19 known Dudleya cymosa subspecies complex sites and identified 542 HAU's as possible sites for the Dudleya cymosa subspecies complex. The Pentachaeta model identified 26 out of 41 known sites and identified 526 possible HAU's of species habitat. The modified classification tree models identified 1,606 and 2,044 sites that belong to the "very suitable" set of solutions for the Dudleya subspecies and Pentachaeta, respectively.
The method of modeling rare species distribution by using augmented predictive mapping is examined. The benefits of this method were the inclusion of incomplete data into the modeling process and the potential to incorporate expert opinion for improved management decisions.