HR: 09:45h
AN: B21C-08 [Abstracts]
TI: Remote sensing of Mercury-contaminated soils through plant reflection spectra
AU: * Dunagan, S C
EM: sdunagan@wesleyan.edu
AF: Dept. Earth and Environmental Sciences,
Wesleyan University, 265 Church Street, Middletown, CT 06459
United States
AU: Gilmore, M S
EM: mgilmore@wesleyan.edu
AF: Dept. Earth and Environmental Sciences,
Wesleyan University, 265 Church Street, Middletown, CT 06459
United States
AU: Varekamp, J C
EM: jvarekamp@wesleyan.edu
AF: Dept. Earth and Environmental Sciences,
Wesleyan University, 265 Church Street, Middletown, CT 06459
United States
AB:
The spatial extent of Hg contamination is often poorly known because current methods used to identify and map Hg soil
contamination on a regional scale are time consuming and expensive. Here we test whether vegetation growing in
Hg-contaminated soils has discernible characteristics in visible/near-infrared (VNIR, 350-2500nm) spectra. Previous work
indicates that Hg can cause chemical and structural changes in plant tissue, including chlorophyll substitution and cell
damage, which we predict may alter the reflectance spectra of plants in a measurable way. To test this hypothesis, Mustard
Spinach plants (n=21) were grown in mercury-spiked soils and in Hg-contaminated soils collected in the field. The plants
were grown under controlled laboratory conditions over a full growth cycle. Foliar Hg concentrations (0.174-3.993ppm) of the
Mustard Spinach plants were positively correlated with Hg concentrations of soils (0.091-39.35ppm). Leaf Hg increased
throughout the growth cycle but decreased in the plants grown in the Hg-spiked and field-contaminated soil at the end of the
growth cycle. Leaf Hg uptake appears to occur through both roots and leaves and may vary as a function of bioavailability of
Hg in the soils.
Reflectance spectra of leaves were measured under artificial light in the laboratory. The potential spectral effects of Hg
on the plants were quantified with selected vegetation indices (VIs) including Ratio Vegetation Index (RVI), Red Edge
Position (REP) and Amplitude (REA) and Normalized Difference Vegetation Index (NDVI) and compared to foliar Hg
concentrations. Correlations between VIs and foliar Hg concentrations are not statistically significant. However, RVI and
REP values of plants grown in Hg-spiked and in field-contaminated soils are lower relative to those from the control plants
during the early and middle portions of the growth cycle and decrease more rapidly than those from control plants at the end
of the growth cycle. These lower RVI and REP values may be related to lower chlorophyll abundances in the Hg-contaminated
plants. The timing of the spectral differences suggests that phenology should be a critical component for future studies of
both in situ and remote detection of foliar Hg content.
DE: 0461 Metals
DE: 0478 Pollution: urban, regional and global (0345, 4251)
DE: 0480 Remote sensing
SC: Biogeosciences [B]
MN: Fall Meeting 2005