Ecological Assessment of Spatio-temporal Changes of Vegetation in Response to Piosphere Effects in Semi-arid Rangelands of Uzbekistan

Since livestock production in Uzbekistan in arid and semi-arid zones is based on rangeland vegetation, contemporary studies have focused on improvement of low productive grazing lands and rehabilitation of degraded rangelands. Less interest has been given to the driving factors of land degradation and their interaction with other ecosystem components, e.g. response of vegetation to grazing induced disturbances. In recent years, due to poor localized grazing management, rangeland degradation has increased. An understanding of the current trend of vegetation changes due to ecological and anthropogenic factors gives the potential of sustainable management of natural rangelands. The aim of this study was to assess the spatial and temporal vegetation succession of Karnabchul semidesert as a function of piospheric effects in two different range sites. Seasonal measurements of vegetation parameters (biomass, density, vegetative cover) along the apparent grazing gradient were conducted during 2005–2007. A conceptual framework of successive vegetation changes along the grazing gradient was developed by applying State and Transition (S&T) models. Vegetation data were analyzed by applying Non Metric Multidimensional Scaling (NMS) ordination. Seasonal dynamics of the Normalized Difference Vegetation Index (NDVI) derived from Landsat imagery were examined to detect the vegetation changes caused by grazing. The vegetation succession was demonstrated as a presence/absence of unpalatable/palatable species in plant composition as a function of piosphere effects. NMS analysis indicated the evident different processes of vegetation cover of two study sites due to distinct grazing regimes. NDVI analysis showed that Landsat imagery provides relevant information about vegetation changes along the grazing gradient in the Karnabchul semidesert. Such results of multidisciplinary approach of vegetation assessment with combination of ecological models open the way in conception of realistic methods for sustainable rangeland management.