Abstract Detail

Ecological Section

Lucas, Kelly L. [1], Carter, Gregory A. [2].

The Use of Hyperspectral Remote Sensing to Evaluate Plant Species Richness on Horn Island, Mississippi.

Mississippi barrier islands are protected from development but remain vulnerable to pollution, species invasions and potential influences of climate change on the marine environment.  These influences may alter barrier island landscapes, water quality and ecosystem processes, and may substantially impact biodiversity.  At local to global scales, remote sensing is increasingly important in providing rapid assessments of biodiversity owing in part to difficulties involved in ground sampling.  The goal of this study is to evaluate potential relationships of hyperspectral remote sensing data with vascular plant species richness, defined as the number of unique species in a given area, on Horn Island, Mississippi.  A 126 band data set acquired in October 2003 by a Hymap imaging spectroradiometer provided coverage of the island throughout the 450-2500 nm spectrum at a ground sample distance (pixel size) of 3 m.  Image data were calibrated to percentage reflectance using a radiative transfer model and known spectral reflectances of ground features.  To correspond with the image data, plant species richness was sampled for 100, 15-m line transects that were located randomly from the georeferenced imagery.  Linear regressions were used to evaluate the relationships of terrain spectral properties with species richness per transect.  The hypothesis that significant relationships of terrain spectral features with richness would be explained by degree of bare soil exposure was tested, exclusive of beach areas.  This was not supported (p= 0.05) by combined data representing all island habitats including dunes, marshes, meadows, woodlands, and transition areas.  However, the hypothesis was supported when relic dunes and meadows were considered separately.  In these cases, band ratios within the near infrared spectrum correlated strongly with richness and soil exposure (r-square = 0.43 and 0.78, respectively).  Continuing research will evaluate additional procedures in assessing species richness via remote sensing and include a comparison of results among islands.

1 - United States Department of Agriculture, Forest Service, 1700 Bypass Road, Winchester, Kentucky, 40391, USA
2 - University of Southern Mississippi, Gulf Coast Geospatial Center and Department of Coastal Sciences, 703 E. Beach Drive, P.O. Box 7000, Ocean Springs, Mississippi, 39564, USA

Keywords:
Hyperspectral
Remote Sensing
islands
biodiversity.

Presentation Type: Oral Paper
Session: 47-3
Location: Salon F - Austin Grand Ballroom/Hilton
Date: Wednesday, August 17th, 2005
Time: 10:45 AM
Abstract ID:166