芦芽山亚高山草甸、云杉林土壤有机碳、全氮含量的小尺度空间异质性

分析比较了山西芦芽山不同海拔处分布的亚高山草甸（样地A，海拔2756.3 m；样地B，海拔2542.3 m）和云杉林（样地C，海拔2656.8 m；样地D，海拔2387.2 m）土壤有机碳和全氮的小尺度空间异质性特征。结果表明：相同植被类型下海拔较高的样地有机碳含量较高（A：49.84 g/kg，B：38.33 g/kg，C：47.06 g/kg，D：40.67 g/kg），而较低海拔的样地土壤有机碳含量的异质性较高；除样地A以外的其他3个样地均表现为高度空间依赖性。亚高山草甸土壤全氮含量的异质性远远高于云杉纯林，四个样地中均表现出强的空间自相关性。亚高山草甸样地土壤有机碳和全氮含量均在较大尺度上空间自相关，云杉纯林样地则表现为较小尺度的空间自相关变异。

Spatial heterogeneity of soil organic carbon and total nitrogen at small scale in subalpine meadow and Picea meyeri forest in Luya Mountain

The patterns and controls of soil organic carbon(SOC) and total nitrogen are critical for our understanding of recycling of nutrients. Spatial heterogeneity causes uneven soil resource distribution. Quantification of the spatial variability is essential for evaluating attributes at unsampled locations. This study was conducted in Luya Mountain, Shanxi Province, China. Four sample plots (30m×30m) were placed, two subalpine meadow plots (Plot A, 2756.3 m; Plot B, 2542.3 m) and two Picea meyeri forest plots (Plot C, 2656.8 m;Plot D, 2387.2 m), The spatial heterogeneity of soil organic carbon(SOC) and total nitrogen(TN) was analyzed based on theory and methodology of spatial pattern analysis in geostatistics.. Soil samples (n=119) were collected from each of the plots in the summer of 2010. We calculated the isotropic semivariograms of SOC and TN. Then spherical models were used to test the semi-variances of SOC and TN for spatial dependence. The experiment results indicated that the SOC of four plots were 49.84 g/kg (A), 38.33 g/kg (B), 47.06 g/kg (C), and 40.67 g/kg (D), respectively. The SOC contents appeared to be higher in the plots with higher elevation. Rather, the spatially variance showed more intensively in the plots with lower altitude. In contrast to the random variation of SOC in plot A, there were high spatial dependences in the spatial distribution of SOC in the others. The spatial heterogeneity of TN in Picea meyeri forest plots were higher than those in subalpine meadow plots,and spatial autocorrelation were high in all the plots. The ranges of spatial autocorrelation variation (the distances within which parameters are spatially dependent) for SOC and TN were larger for subalpine meadow plots, but smaller for Picea meyeri forest plots.