Abstract:The Hulunber steppe, located in the north-eastern Inner Mongolia, is one of the representative and typical steppes in northern China. The Hulunber grassland plays a key role in maintaining the ecological security in northeastern China and serves as an important north-eastern ecological protection barrier. This grassland in temperate continental semi-arid region has a remarkable transition of climate, soil, vegetation, and biogeochemical cycling from north to south and supports diverse species of plants and animals as well as the regional socio-economic development. All these features offer an unique opportunity for examining spatial heterogeneity of grasslands in relation to degeneration. Carbon-nitrogen stoichiometry is one of the important indicators of vegetation composition, ecosystem function, and nutrient status. However, to our knowledge, factors driving these soil nutrients patterns at micro-scale still remain uncertain. Given the importance of plant carbon and nitrogen chemistry in biology, physiology, and biogeochemistry, it is important to understand patterns of plant carbon and nitrogen at micro-scales and their possible links to ecosystem-scale biogeochemical cycling. Spatial heterogeneity dynamics in the mini-patches is one of the most active ingredients in the process of grassland degradation. To demonstrate how grazing affects soil carbon and nitrogen status, we analyzed large datasets including 900 observations of soil samples in temperate grasslands in northern China. Using geo-statistical analyses, we studied the spatial heterogeneity of soil total carbon, total nitrogen, and alkali-hydrolyzable nitrogen in mini-patches of degraded Stipa krylovii grassland with different grazing intensities in Hulunber, Inner Mongolia. A sampling plot of 1 m×1 m was selected within each mini-patch, and a total of 100 quadrats of 10 cm×10 cm were determined in each plot. Soil samples (0-10 cm depth) were collected from each quadrat. We hypothesized that "nutrient accumulation effect" is the primary explanation for the spatial heterogeneity of soil nutrients in temperature grasslands. The results indicated that the contents of soil total carbon, soil total nitrogen and alkali-hydrolyzable nitrogen were highest in moderate grazing plots, followed by heavy grazing and light grazing plots. The spatial heterogeneity of soil total carbon was highest in moderate grazing, followed by light grazing and heavy grazing plots, whereas the spatial heterogeneity of soil nitrogen was most significant in heavy grazing, followed by light and moderate grazing plots. The most obvious change in spatial heterogeneity of alkali-hydrolyzable nitrogen appeared in moderate grazing, followed by heavy grazing and light grazing. Changes in littering and the related soil biochemical transformation may occur in grassland communities under different grazing intensities, which, in turn, may result in reduction of nutrient pool and in increases of spatial heterogeneity. These changes in soil nutrient status may ultimately affect nutrient cycling and ecosystem productivity. These data showed that the scale-dependent variability of enrichment and migration of soil nutrition may influence the progress of grassland succession. The present study provides insight into the question of how plants have adapted to different environmental pressures, and contribute to better calibration of future soil-degradation models.