| Abstract: | Land use practices and vegetation cover distribution are considered to be the most important dynamic factors that influence the land degradation or the soil erosion of a region. In this study, a Soil Protection Index (SPI) is defined as a function of land use practices and intensity of vegetation cover. This index is used to map the relative degree of protection of topsoil from being eroded by external effects such as rainfall and overland flow. A fuzzy rule‐based model integrated within ArcGIS® has been set‐up and tested with the aim to develop SPI maps. The amount of vegetation cover distribution, that is, Normalized Difference Vegetation Index as proxy parameter and Land Use–Land Cover map are chosen as fuzzy input parameters for the SPI as the desired system output. The approach was tested in the Upper Awash basin in Ethiopia. The output SPI map was qualitatively evaluated against the expert‐defined land degradation risk class, and it was found that locations that are mapped with ‘low and very low’ SPI classes at different time periods of the year have a high potential land degradation risk. Furthermore, socio‐economic data (‘population and livestock densities’) and environmental parameters (‘altitude and soil erodibility’) for the region are used to correlate with the SPI map as an indirect method of evaluation. It is found that population and livestock density explained 68 per cent of the spatial distribution pattern of predicted SPI and an adjusted R‐squared value of 0·681 (p < 0·05) was obtained. It was also found that the SPI distribution over the region for two different time periods, that is, January and July 2001, correlated positively (R2 = 0·41 and R2 = 0·51) with the soil erodibility of the region. The transferability and applicability of the model for different environmental settings or landscapes were tested by mapping the SPI of Italy. This SPI map of Italy was compared with the soil erosion map of Italy produced by the European Soil Bureau. It can be concluded that the SPI map reflects the potential land degradation risk distribution of the case‐study region. Results show that a fuzzy rule‐based model can provide useful preliminary information even without detailed and precise data information for developing appropriate strategies for land degradation assessment vital for sustainable land use management. Copyright © 2012 John Wiley & Sons, Ltd. |
