HR: 0800h
AN: B51D-0249    [Abstracts]
TI: Investigating Plant Patterns on Alluvial Fan Deposits of the Mojave Desert Using High Resolution Imagery
AU: Phelps, G A
EM: gphelps@usgs.gov
AF: U.S. Geological Survey, 345 Middlefield Road ms 989, Menlo Park, CA 94025 United States
AU: * Robinson, S
EM: srobinson@usgs.gov
AF: U.S. Geological Survey, 345 Middlefield Road ms 989, Menlo Park, CA 94025 United States
AU: Miller, D M
EM: dmiller@usgs.gov
AF: U.S. Geological Survey, 345 Middlefield Road ms 989, Menlo Park, CA 94025 United States
AB: Maps of desert vegetation derived from low-altitude, high-resolution imagery were combined with detailed geologic mapping to address how the spatial distribution of individual plants varies with fan position and mapped geologic unit. We collected Color Infrared (CIR) imagery along two proximal to distal fan transects in the Mojave Desert, CA from a helicopter platform; because of its small pixel size (3 cm to 20 cm) and sensitivity to vegetation, this imagery is ideal for analyzing spatial patterns of individual plants. Four band ratios were calculated from the CIR imagery; a three-band composite image of these ratios provides a base map that discriminates individual plants and is sensitive to the amount of photosynthetic material in the plant. An expert classifier was used to produce an automated, georeferenced vegetation map that includes both green vegetation and non-photosynthetic vegetation (NPV), from which percent cover and location of plants were derived. By analyzing the vegetation data in concert with detailed geologic mapping (e.g. 1:20, 1:500), we can address whether mapped geologic units influence plant location and density. Two plots along the CIR transects were compared in a preliminary analysis: one representing distal fan flora and geomorphology and one representing proximal fan flora and geomorphology. The georeferenced CIR imagery was used to map the discrete location of every plant in the two plots (3500 plants over 10000 m2). Because each plant has a georeferenced location, point pattern analysis can be used to determine 1) whether the plants exhibit spatial structure, and 2) whether any spatial structure can be attributed to alluvial fan position or mapped geologic unit. Several differences between the plots were observed. First, the proximal plot has more variety (11 species vs. 4), more individual plants (2000 vs. 1500), and denser cover (27 percent cover vs. 24) than the distal plot. Second, the two major species present in both plots, Larrea tridentata (LT) and Ambrosia dumosa (AD), exhibit spatial structure. AD exhibits strong clustering in both the proximal and distal plots; LT exhibits weak clustering between 5 and 10 m in the proximal plot and weak dispersion in the distal plot between 4 to 8 m. These plot scale patterns can then be compared to the spatial structure within individual geologic units.
DE: 0476 Plant ecology (1851)
DE: 0480 Remote sensing
SC: Biogeosciences [B]
MN: Fall Meeting 2005