PARENT SESSION
Monday, August 7, 8:00-11:30 am
COS 11 - Remote sensing
L-3, Lobby Level, Cook Convention Center
Presiders: G Carter and D Rocchini
Indicators of plant species richness in airborne hyperspectral imagery of Horn Island, Mississippi.
Lucas, Kelly*,1, Carter, Gregory1, 1 University of Southern Mississippi, Ocean Springs, MS
ABSTRACT- Mississippi's barrier islands are protected from development but remain vulnerable to pollution, species invasions and the impacts of climate change. These influences may be linked with apparent decadal changes in species diversity. Owing largely to the difficulties encountered in acquiring field data, there is a pressing need to develop methods for rapidly assessing island biodiversity. This study determined the extent to which the species richness of vascular plants on Horn Island, Mississippi, might be estimated using airborne imaging spectrometry. A 126 band data cube acquired in October 2003 by the Hymap imaging spectroradiometer provided coverage of Horn island throughout the 450-2500 nm spectrum at a ground sample distance of 3 m and a spectral resolution of approximately 15 nm. Image data were calibrated to percentage reflectance using a radiative transfer model and known spectral reflectances of ground surfaces. To correspond with the image data, plant species richness was sampled for 95, 15-m transects that were established at random locations. For the 5 to 11 image spectra (pixels) that represented a given transect location, mean spectral reflectance was computed. A spectral coefficient of variation (CV) was determined also as an indicator of spatial variability in surface reflectance. When data were combined among all island habitats, simple regressions of species richness with mean reflectance or CV per spectral band yielded no significant relationships (p= 0.05). However, within-habitat richness was significantly related to band ratios of mean transect reflectance when near- and mid-infrared bands were utilized (r-squared = 0.98, 0.74, 0.4 and 0.43 for woodland, marsh, meadow and transition zones, respectively). Continuing research will compare the efficacy of this approach with that of spectral derivative analysis in estimating plant species richness on Horn Island.
Key words: remote sensing, biodiversity, barrier islands