Item

Abstract

Biological soil crusts (biocrusts) form a regular and relevant feature in drylands, as they stabilize the soil, fix nutrients, and influence water cycling. However, biocrust forming organisms have been shown to be dramatically vulnerable to climate and land use change occurring in these regions.

In this study, we used Normalized Difference Vegetation Index (NDVI) data of biocrust-dominated pixels (NDVIbiocrust) obtained from hyperspectral and LANDSAT-7 data to analyse biocrust development over time and to forecast future NDVIbiocrust development under different climate change and livestock density scenarios in southern Africa. We validated these results by analysing the occurrence and composition of biocrusts along a mesoclimatic gradient within the study region.

Our results show that NDVIbiocrust, which reached maximum values of 0.2 and 0.4 in drier and wetter years, respectively, mainly depended on water availability. A predicted decrease in rainfall events according to all future climate scenarios combined with increased temperatures suggested a pronounced decrease in NDVIbiocrust by the end of the 21st century caused by reduced biocrust coverage. Livestock trampling had similar effects and exacerbated the negative impacts of climate change on biocrust coverage and composition. Data assessed in the field concurred with these results, as reduced biocrust cover and a shift from well-developed to early stages of biocrust development were observed along a gradient of decreasing precipitation and increasing temperatures and livestock density.

Our study demonstrates the suitability of multi-temporal series of historical satellite images combined with high-resolution mapping data and Earth system models to identify climate change patterns and their effects on biocrust and vegetation patterns at regional scales.