A regional analysis of drivers and impacts of land cover change and long-term land cover trends in the Great Basin, United States.

摘要

An improved understanding of land use/land cover change at local and regional scales is important in an increasingly human-dominated biosphere. The land surface provides resources necessary for human survival (e.g., cropland, water, raw materials) as well as providing other services such as habitat for native species, carbon storage, and nutrient cycling. A goal of land change science is to identify where land cover change is taking place, understand how land use may affect that change, and determine what the consequences of change may be. In the Great Basin Desert of the Western U.S., an important form of land cover change is invasion by non-native cheatgrass (Bromus tectorum). Cheatgrass invasion destroys native shrub ecosystems, leading to a loss of biodiversity, loss of viable rangeland and increased fire frequency. In this work, I show how remote sensing can be used to detect the regional and local extents of cheatgrass invasion. Remote sensing results are then used to assess the spatial patterns of cheatgrass invasion over time to determine how land use might have affected invasion. Further, I consider the long-term impacts of cheatgrass invasion on biodiversity and carbon storage in the Great Basin. In addition to an analysis of cheatgrass, this thesis presents a new methodology for time series modeling, which can be used to better interpret annual and inter-annual vegetation community phenology. I apply this modeling methodology to all land cover in the Great Basin to assess long-term land cover trends and localized anomalous response within the range of land cover classes present. By investigating regional land cover change I am able to provide more detailed analysis of the drivers of change for land managers while working at a scale relevant to studies of global environmental change.