The Eurasian steppe is one of the largest grassland ecosystems of the world and was used for nomadic pastoralism since several thousands of years. Under the emerging communism in the 1950s followed by the implementation of a free market economy since the 1980s, the extensive, mobile grazing management turned to a sedentary, intensive livestock production in Northern China. The dramatically intensified land use and continuously increasing stocking rates of sheep resulted in severe degradation of the grasslands. In Northern China, around 30% of the grasslands (covering an area of nearly 4 million km²) are degraded that affects as many as 400 million people. The degradation of steppe vegetation due to overgrazing is associated with a decline of soil organic matter (SOM) stocks. Semi-arid grasslands are among the most important terrestrial carbon (C) sinks as they store approximately 15% of total soil organic carbon (SOC) stocks. A detailed knowledge on the amount of SOC in the steppe soils at different spatial scales and the consequences of intensive grazing is essential in the light of a rising demand for agricultural land, ongoing soil degradation, erosion and desertification and sequestration of atmospheric carbon dioxide (CO2). The Sino-German research unit MAGIM (Matter fluxes in grasslands of Inner Mongolia as influenced by stocking rate) was established in 2004 in order to investigate the impact of intensified grazing and potential benefits of grazing exclusion in a semi-arid grassland ecosystem located in the Xilin River catchment in Inner Mongolia, Northern China. The region is part of the continental semi-arid grasslands of the Central Asian steppe ecosystem, with a dry and cold mid-latitude climate. The dominant vegetation types are Leymus chinensis and Stipa grandis dominated steppe communities, which are typically used for semi-nomadic or static grazing with sheep and goats. The aim of this work was to estimate the amount of SOM in the grassland soils at different spatial scales, ranging from small scales of single grass tussocks to larger scales of experimental plots as well as the whole catchment area. In a next step, this work aimed to gain insight into the degradation processes and the loss of SOM induced by heavy grazing at the scale of soil micro- and macroaggregates. Finally, several possible short-term indicators for beginning grassland desertification were evaluated, making use of a controlled grazing experiment.weiterlesen