"August 2003."

Includes bibliographical references.

We used three basic indicators of watershed integrity: hydrologic function, water quality, and upland wildlife habitat to evaluate potential land cover changes within the upper Yellowstone River watershed. A satellite-based land cover classification was completed for 2,474,141 acres within the Yellowstone River basin (10070001-Yellowstone Headwaters and 10070002-Upper Yellowstone 4th code subbasins) using Landsat satellite imagery dated July 13, 1999, and July 12, 1985. Differences in spectral attributes between 1999 Enhanced Thematic Mapper (ETM+) and 1985 Thematic Mapper (TM) scenes, in addition to excessive cloud cover on the 1985 scenes prevented accurate comparison of land cover change over time. The land cover assessment was performed solely on the 1999 classification. Post-stratification accuracy was 72.2 percent. A Geographic Information System (GIS) analyzed the distribution and intersection of key resource theme attributes (soil, climate, ownership, topography, census, and important wildlife habitat) with the 1999 land cover classification. Results indicated that the very diverse landscape was largely composed of federally managed, coniferous forest, and shrub/grasslands. Urban or Developed and Agricultural Land-Irrigated land cover together accounted for less than 2 percent of the watershed area. Broadleaf Riparian represented the next to least in extent of the 15 cover classifications identified. Differences in land cover characteristics were measured between 5th code hydrologic units (HUCs). Low/Moderate Cover Grasslands, Agricultural Lands/Irrigated, Urban or Developed, and Broadleaf Riparian cover categories increased in relative composition in a downstream direction and in proximity to the river corridor. Low/Moderate Cover Grasslands surprisingly were the most prevalent land cover category within the 1/2 mile wide corridor bisected by the river. Evaluations of land cover related to hydrologic function, water quality characteristics, and upland wildlife habitat are also presented and discussed. Although land cover composition at the watershed scale appears to be relatively uninfluenced by human activity at present, we recommend periodic reassessment of land cover at the watershed and stream corridor scale in conjunction with monitoring common biotic indicators to track and evaluate the effect of land cover trends over time on stream and watershed function.

Mode of access: WorldWideWeb, PDF reader.