Details

A SIMPLE GIS APPROACH TO PREDICTING RARE PLANT HABITAT: NORTH CENTRAL ROCKY MOUNTAINS, UNITED STATES FOREST SERVICE, REGION ONE

by Nock, Erin Elizabeth

Abstract (Summary)
Understanding the potential distribution of rare species is a key component in managing and regulating land-use activities. Predictive modeling of plant distributions rests on the assumption that correlations exist between the presence or absence of a species and selected climate, topographic, substrate, and land-cover variables. Using the DOMAIN algorithm along with Geographic Information Systems (GIS) techniques, a biophysical envelope model was applied to 21 rare plant species listed on the Region One Regional Foresters Sensitive Species List. Environmental variables, including annual precipitation, mean May temperature, slope, aspect, elevation, geologic material and dominant vegetation type, were used as predictors. Model output was field-verified by expert botanists who used their knowledge to assess areas predicted as potential habitat. A total of 44 new rare plant species element occurrences were located, including two new state occurrence records for Idaho. Model evaluation used a multi-layered approach: (1) the percentage of known occurrences within areas of predicted potential habitat (2) whether botanists found potential habitat within predicted areas; and (3) whether new occurrences were found within predicted areas. Model success for each species was evaluated using error matrices populated with the number of pixels correctly or incorrectly classified as habitat. Misclassification of suitable and unsuitable habitat is inevitable in any habitat modeling procedure, and sources of error may be caused by inherent problems in the modeling process or complications arising from an organisms ecology. Plant species for which habitat was not successfully modeled were often associated with microhabitats, had inappropriate environmental parameters used as input, or had unusual distribution patterns.
Bibliographical Information:

Advisor:Anna Klene; David Shively; Susan Rinehart

School:The University of Montana

School Location:USA - Montana

Source Type:Master's Thesis

Keywords:geography

ISBN:

Date of Publication:08/07/2008

© 2009 OpenThesis.org. All Rights Reserved.