Remote detection of non-native invasive plant species using geospatial imagery may significantly improve monitoring, planning and management practices by eliminating shortfalls, such as observer bias and accessibility involved in ground-based surveys. The use of remote sensing for accurate mapping invasion extent and pattern offers several advantages, including repeatability, large area coverage, complete instead of sub-sampled assessments and greater cost-effectiveness over ground-based methods. It is critical for locating, early mapping and controlling small infestations before they reach economically prohibitive or ecologically significant levels over larger land areas. This study was designed to explore the ability of hyperspectral imagery for mapping infestation of musk thistle (Carduus nutans) on a native grassland during the preflowering stage in mid-April and during the peak flowering stage in mid-June using the support vector machine classifier and to assess and compare the resulting mapping accuracy for these two distinctive phenological stages. Accuracy assessment revealed that the overall accuracies were 79% and 91% for the classified images at preflowering and peak flowering stages, respectively. These results indicate that repeated detection of the infestation extent, as well as infestation severity or intensity, of this noxious weed in a spatial and temporal context is possible using hyperspectral remote sensing imagery.
Woody plant encroachment into grasslands and rangelands is a world-wide phenomenon but detailed descriptions of changes in geographical distribution and infilling rates have not been well documented at large land scales. Remote sensing with either aerial or satellite images may provide a rapid means for accomplishing this task. Our objective was to compare the accuracy and utility of two types of images with contrasting spatial resolutions (1-m aerial and 30-m satellite) for classifying woody and herbaceous canopy cover and determining woody infilling rates in a large area of rangeland (800 km2) in north Texas that has been invaded by honey mesquite (Prosopis glandulosa). Accuracy assessment revealed that the overall accuracies for the classification of four land cover types (mesquite, grass, bare ground and other) were 94 and 87% with kappa coefficients of 0.89 and 0.77 for the 1-m and 30-m images, respectively. Over the entire area, the 30-m image over-estimated mesquite canopy cover by 9 percentage units (10 vs. 19%) and underestimated grass canopy cover by the same amount when compared to the 1-m image. The 30-m resolution image typically overestimated mesquite canopy cover within 225 4-ha sub-cells that contained a range of mesquite covers (1–70%) when compared to the 1-m image classification and was not suitable for quantifying infilling rates of this native invasive species. Documenting woody and non-woody canopy cover on large land areas is important for developing integrated, regional-scale management strategies for rangeland and grassland regions that have been invaded by woody plants.