中文摘要：

应用经典统计学和地统计学方法,分析了八达岭地区土壤全氮（TN）在不同层次（A,B,C）的空间变异特征。同时结合地理信息系统（GIS）,分析了该地区植被类型和土壤TN之间的关系。应用分类回归树模型（classification and regression trees,CART）分析了土壤TN和海拔与植被分布格局的关系。得到以下结论：（1）TN在A、B、C层平均值分别为2.94、1.30、0.63g/kg,变异系数（CV）分别为33%、33%、45%,都表现为中等变异。（2）TN在不同土层的变异函数理论模型符合球状模型,TN在A层为弱空间相关,在B、C层为中等空间相关。（3）泛可里格插值表明,TN在不同层次都表现出了明显的空间分布趋势。不同植被类型所对应土壤全氮的空间分布则各不相同。（4）CART研究结果表明,该区植被类型分布格局可大致划分为四大部分。可初步确定海拔725m,TN含量4.23 g/kg和5.69 g/kg为影响该区植被分布格局的重要参考值。

英文摘要：

Spatial variability causes uneven soil resource distribution and controls species distribution and recruitment in terrestrial ecosystems.Quantification of the spatial variability is essential for understanding the relationship between soil properties and environmental factors and to estimate attributes at unsampled locations.Spatial variability of soil nutrient can provide guidance for the proper management of forest ecosystem.According to forest resource inventory subplot maps of China,121 subplots were established.Soil samples for the experiment were collected from soil profiles in the central portion of each plot.Three replicate samples of each horizon at every plot were mixed with roots and stones removed by hand.Soil total nitrogen（TN） was measured using sulfate-perchlorate acid heating digestion-azotometer distillation titration method.All statistical analysis was performed using the open source software R（version 2.7.0）.TN spatial distribution predicted by kriging was exported to ArcGIS 9.2 to produce maps.Spatial variability of soil total nitrogen under different layers was examined using classical statistics and geostatistics in Badaling.At the same time,geostatistics combined with geographic information system（GIS） were applied to analyze the relationship of vegetation type and soil total nitrogen.Relationship between elevation,TN of soil and vegetation distribution pattern was evaluated by classification and regression trees（CART）.The results showed that：（1） The means of TN were 2.94 g/kg,1.30 g/kg,0.63 g/kg in three different layers,respectively.Coefficient of Variation（CV） of TN were 33%,33%,45%,respectively.So they showed medium variability.（2） Optimal theoretical models of TN were spherical model in different layers.Spatial correlation distances of TN-A,TN-B and TN-C were 804m,1038m and 1400m,respectively.The nugget/sill C0/（C0＋C） ratio for TN-B and TN-C were 55% and 63%,respectively,suggesting moderate spatial correlation.TN-A has a weak spatial correlation with