Abstract: Spatial distribution heterogeneity of Stipa grandis population at small scale in a typical steppe in Xilinhot, Inner Mongolia under two utilization ways (mowing and grazing) and exclosure (control) was studied by using semivariance analysis, fractal analysis and Kriging interpolated analysis based on field investigation data. The results showed that the highest population density of S. grandis was observed in the mowing plots, followed by the grazing plots, and the control plots had the lowest population density (P<0.05). The population structure ratio of S. grandis in the exclosure plots was 95.3%, indicating a strong spatial autocorrelation; while the population structure ratios of S. grandis in the mowing plots and grazing plots were 70.2% and 62.5% respectively, indicating a medium degree spatial autocorrelation. The spatial heterogeneity of S. grandis population was in order of mowing plots > grazing plots > exclosure plots. The optimal models of S. grandis population spatial distribution for exclosure, mowing and grazing plots were Gaussian model, spherical model and spherical model, according to fitting semivariogram. The spatial variation of S. grandis in the exclosure plots was affected by structural factors and trended to a simple distribution pattern, while the spatial variation of S. grandis in the mowing plots and grazing plots was affected by both structural factors and random factors, and showed a mosaic distribution pattern of high density patches and low density patches.

Key words: Beijing., urbanization, population spatio-temporal pattern, ecosystem quality, urban surrounding region