Predicting shrub ecophysiology in the Great Basin Desert using spectral indices

Arid and semi-arid ecosystems are likely to be particularly sensitive to changes in precipitation. Reflectance spectra can provide valuable information and scaling opportunities regarding plant physiological responses to precipitation. However, their effectiveness in and and semi-arid ecosystems with low vegetation cover remains equivocal. We compared stand-level spectral indices to leaf-level photosynthetic gas exchange, stem water potential, and leaf area index (LAI) of a Great Basin Desert shrubland ecosystem exposed to multiple, long-term snow depth manipulations. At the seasonal scale, several spectral indices were able to predict gas exchange variables, LAI to a lesser degree, but never stem water potential. Whereas spectral indices exhibited a seasonal response, they did not differ across snow depth treatments, in agreement with a lack of physiological responses to the snow depth treatments. This may be in part the result of the 2004-2005 ENSO event, which delivered heavy snowfall that may have overwhelmed responses to snowfall manipulations. The Normalized Difference Vegetation Index (NDVI) and the Red Edge Maximum Slope (REM) best predicted gas exchange variables in this ecosystem, although some relationships were species-specific. However, this research suggests that indices previously found to correlate well with ecophysiological variables may not serve as effective proxies in all ecosystems. (C) 2009 Elsevier Ltd. All rights reserved.