Biological soil crusts, known as biocrusts, are communities of microorganisms, fungi, algae, lichens, and mosses. Although occupying the uppermost soil layer, biocrusts play a crucial role in dryland ecosystems. Therefore, considerable effort has been put into biocrust mapping by remote sensing. In this regard, several spectral crust indices were developed and tested in different sites around the globe. Each of these indices relies upon various parts of the electromagnetic spectrum created by the biocrusts’ and substrates’ different physiological properties. Among these properties, it is worth mentioning the phycobilin pigments that reflect in the blue spectral region (associated with Crust Index, CI), chlorophyll absorption in the red (Normalized Different Vegetation Index, NDVI), and cellulose absorption in the shortwave infrared (Satellite Cellulose Absorption Index, SCAI).
These spectral indices were used to map the sand dunes of the Northwestern Negev (Israel) and an adjacent area in Sinai. A dense cover of biocrusts characterizes the Negev due to a strict conservation policy that lets the biocrusts develop. In contrast, the topsoil in Sinai has been trampled by anthropogenic activity, such as vehicles and herds, leaving the sand dunes exposed and mobile.
First, in situ biocrust samples were measured in the laboratory with a field spectrometer (ASD). Then, two multivariate statistical models were performed to compare multiple responses and explanatory variables: the Partial Least Squares – Discriminant Analysis (PLS-DA) and the Partial Least Squares – Regression (PLS-R). The former was utilized in order to define the significant difference between sands and biocrusts as well as between different locations. PLS-R and the Variable Importance in Projection (VIP) analyses were conducted to (1) confirm the physiological theories behind the three indices; and (2) refine the width of the critical wavelengths in which the spectra are primarily valuable for the indices. Finally, an EnMAP image was used to map the three spectral indices on both sides of the border.
