A spatiotemporal analysis of aspen decline in southern Utah's Cedar Mountain, using remote sensing and geographic information systems.

摘要

Widespread mortality of quaking aspen (Populus tremuloides Michx.) has occurred over large expanses of the Western US during the 20 th century. While much of this decline was due to conifer encroachment into seral aspen, significant aspen losses also occurred in areas of persistent aspen and may have been exasperated by drought conditions. Aspen decline has been especially notable at Cedar Mountain, Utah, an area of mostly private land and extensive persistent aspen coverage.;The objectives of this study were to create a time series of live and dead aspen cover on the Cedar Mountain landscape, using remotely sensed imagery, and to test whether water stress correlated to the decline therein. To accomplish these objectives, a decision tree classifier was used to classify the Cedar Mountain area into live and dead aspen cover classes for the years 1985, 1990, 1995, 2001, 2005, and 2008. Thereafter, post-classification change analysis was performed to determine areas and time periods of elevated decline. Regression analyses were performed to ascertain correlations between climatic data and percent change in aspen cover. A topographic analysis using zonal statistics was also performed to determine landscape positions where aspen decline is more prevalent.;The time series models indicated that aspen decline followed a step-wise pattern with an overall decrease of 23.57 % in aspen cover during a 23-year period. Considerable aspen decline occurred early in the study time frame, with decreases of 1.38 and 1.36 -1 in 1990 and 1995, respectively. The middle period between 1995 and 2001 had no net change in aspen cover. However, the end of the time series showed the greatest decline with decreases of 1.56 and 1.99 % yr-1 in 2005 and 2008, respectively. There was a correlation between percent change in aspen cover and precipitation, suggesting that drought weakens aspen, making it susceptible to future decline. The topographic zonal statistics revealed that drier landscape positions had greater frequencies of dead aspen. The most significant predictor of aspen decline was elevation, which was significantly greater in the live aspen for three of the five years.