Mapping vegetation communities across home ranges of mountain goats in the North Cascades for conservation and management
Article first published online: 16 MAR 2012
© 2012 International Association for Vegetation Science
Applied Vegetation Science
Volume 15, Issue 4, pages 560–570, October 2012
How to Cite
Wells, A. G., Rachlow, J. L., Garton, E. O., Rice, C. G. (2012), Mapping vegetation communities across home ranges of mountain goats in the North Cascades for conservation and management. Applied Vegetation Science, 15: 560–570. doi: 10.1111/j.1654-109X.2012.01193.x
- Issue published online: 4 SEP 2012
- Article first published online: 16 MAR 2012
- Manuscript Accepted: 8 FEB 2012
- Manuscript Received: 4 JUL 2011
- Seattle City Light Department Wildlife Research Program
- DeVlieg Foundation
- Washington Department of Fish and Wildlife (WDFW)
- WDFW Aquatic Lands Enhancement Account (ALEA)
- Sauk-Suiattle Indian Tribe
- United States Forest Service
- United States National Park Service
- Western Washington University
- Mountaineers Foundation
- LandSat 5TM;
- non-hierarchical classification;
- Oreamnos americanus ;
- thematic classification
What is the composition of vegetation communities found across mountain goat ranges? Can we use LandSat 5TM to model those vegetation communities across a mountain range to guide conservation and management plans?
The Cascade Mountains of Washington, USA.
We surveyed vegetation across previously undocumented home ranges of mountain goats, which we determined via GPS telemetry and tracking. We used non-hierarchical cluster analysis to identify vegetation communities found therein. We linked the results of the fieldwork to a supervised classification of three LandSat 5TM (bands 2: 0.52–0.60 μm, 3: 0.63–0.69 and 4: 0.76–0.90 μm) images acquired consecutively on 29 July 2005 and to aerial imagery with 1-m resolution from the National Agriculture Inventory Program (NAIP) to create a map of vegetation communities across the Cascade Mountain range. Finally, we evaluated the success of the classification accuracy after transforming categorical land cover into percentage cover classes at a resolution that matches the positional error of the GPS telemetry collars.
Field reconnaissance and analysis identified eight vegetation communities across mountain goat home ranges: montane forest, sub-alpine forest, heather, huckleberry, east-side shrub, sparse meadow, dense meadow and cryptogram (talus). Analysis of imagery linked the vegetation communities found on mountain goat ranges to specific thematic classes and projected the classification across the entire range. We identified a scale of analysis (ca. 2 ha) that achieved balance between bias and precision resulting in an accuracy of ca. 80% based on comparison with 1-m digital aerial imagery and survey results.
Our results demonstrate that common and readily available data can be used to produce reliable maps of percentage cover type in mountainous terrain for specific applications. Our work is the first to describe the type and distribution of vegetation communities occurring within and across mountain goat ranges along the length of the Washington Cascades.