Importance of biocrusts in dryland monitoring using spectral indices

Multi-temporal remote sensing information and spectral indices have been extensively used in studies to monitor ecosystem functioning and surface-energy budgets. However, most of these indices did not show good results in areas covered by sparse vegetation, like most of the Drylands. In these ecosystems, open spaces between plants are often covered by biological soil crusts (biocrusts), i.e. communities of cyanobacteria, algae, microfungi, lichens, mosses and other microorganisms growing in the uppermost millimeters of the soil. Due to their mostly dark color, biocrusts influence the spectral response of dryland surfaces, making it necessary to assess the sensibility of widely used spectral indices to variations in biocrusts cover. In this study we used spectra of biocrusts, bare soil and vegetation to analyze the effect of biocrust cover on the spectral response of heterogeneous areas. In a second approach we investigated the impact of biocrust water status on spectral characteristics. Based on spectral mixture analysis, we calculated the response of a wide range of vegetation/biocrust/bare soil landscape compositions, obtaining a total of 702 spectra. These were used to calculate the Normalized Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), the Water Index (WI) and surface albedo, and the effect of biocrust cover and water status on these indices was analyzed. Biocrusts exerted a considerable effect on vegetation indices and surface albedo, whereas WI was mostly affected by vegetation type and cover. As biocrust cover increased, the value of NDVI and EVI also increased, whereas albedo decreased, and these effects were more important under low vegetation cover. Moreover, as biocrusts almost immediately turned dark after water pulses, the effect of biocrust cover on spectral indices increased already 30 min after wetting. Although these results varied depending on vegetation type, they demonstrate, that biocrusts largely affect the spectral response of dryland surfaces, and they illustrate how this effect is reinforced by water. Thus, biocrusts need to be considered in studies analyzing dryland phenology, productivity and water status. Moreover, in order to increase the accuracy of hydrological and climate forecast predictions, biocrust effects on surface albedo, both in a dry and wet stage, need to be included. (C) 2015 Elsevier Inc. All rights reserved