Mapping Variation in Vegetation Functioning with Imaging Spectroscopy

Abstract:

Imaging spectroscopy (otherwise known as hyperspectral remote sensing) offers the potential to characterize the spatial and temporal variation in biophysical and biochemical properties of vegetation that can be costly or logistically difficult to measure comprehensively using traditional methods. A number of recent studies have illustrated the capacity for imaging spectroscopy data, such as from NASA’s AVIRIS sensor, to empirically estimate functional traits related to foliar chemistry and physiology (Singh et al. 2015, Serbin et al. 2015). Here, we present analyses that illustrate the implications of those studies to characterize within-field or -stand variability in ecosystem functioning. In agricultural ecosystems, within-field photosynthetic capacity can vary by 30-50%, likely due to within-field variations in water availability and soil fertility. In general, the variability of foliar traits is lower in forests than agriculture, but can still be significant. Finally, we demonstrate that functional trait variability at the stand scale is strongly related to vegetation diversity. These results have two significant implications: 1) reliance on a small number of field samples to broadly estimate functional traits likely underestimates variability in those traits, and 2) if trait estimations from imaging spectroscopy are reliable, such data offer the opportunity to greatly increase the density of measurements we can use to predict ecosystem function.