Assessing the use of remotely sensed measurements for characterizing rangeland condition.

摘要

There are over 233 million hectares (ha) of nonfederal grazing lands in the United States. Conventional field observation and sampling techniques are insufficient methods to monitor such large areas frequently enough to confidently quantify the biophysical state and assess rangeland condition over large geographic areas. In an attempt to enhance rangeland resource managers' abilities to monitor and assess these factors, remote sensing scientists and land resource managers have worked together to determine whether remotely sensed measurements can improve the ability to measure rangeland response to land management practices. The relationship between spectral reflectance patterns and plant species composition was investigated on six south-central Kansas ranches. Airborne multispectral color infrared images for 2002 through 2004 were collected at multiple times in the growing season over the study area. Concurrent with the image acquisition periods, ground cover estimates of plant species composition and biomass by growth form were collected. Correlation analysis was used to examine relationships among spectral and biophysical field measurements. Results indicate that heavily grazed sites exhibited the highest spectral vegetation index values. This was attributed to increases in low forage quality broadleaf forbs such as annual ragweed (Ambrosia artemisiifolia L.). Although higher vegetation index values have a positive correlation with overall above ground primary productivity, species composition may be the best indicator of healthy rangeland condition. A Weediness Index, which was found to be correlated with range condition, was also strongly linked to spectral reflectance patterns recorded in the airborne imagery.