Do vegetation indices provide a reliable indication of vegetation degradation? A case study in the Mongolian pastures

Space and ground observations were applied to explore the ability of remote sensing techniques to assess the effect of grazing on vegetation degradation. The steppe biome of Mongolia was used as the study area, in which several pairs of sites were investigated - each pair comprised an ungrazed (fenced-off) area and a heavily grazed area. For each pair, the enhanced vegetation index (EVI), computed from Landsat-7 Enhanced Thematic Mapper Plus (ETM+) data, along with field-observed biophysical variables (e.g. plant density, species composite, above-ground biomass (AGB), and percentage cover) and plant spectral reflectance data were collected. As expected, plant density, AGB, and percentage cover values were significantly higher in the ungrazed areas than in the adjacent grazed ones. However, unexpectedly, the grazed areas had significantly higher EVI values than the ungrazed areas. It was found that unpalatable species had invaded into the grazed areas, substituting the native grasses. These invasive species, mostly characterized by denser leaf structure, induced higher spectral responses in the near infrared (NIR) region of the electromagnetic spectrum. EVI is the preferred vegetation index to use for detecting this phenomenon, since it is more sensitive to variations in leaf cellular structural as expressed in the NIR (rather than the red) portion of the spectrum. The current study contradicts the general assumption that the higher the vegetation index value, the better the grazing conditions.