石栎-青冈常绿阔叶林土壤有机碳和全氮空间变异特征

Spatial heterogeneity of soil organic carbon and total nitrogen concentrations in a Lithocarpus glaber-Cyclobalanopsis glauca evergreen broadleaved forest

在1hm2(100 m×100 m)石栎(Lithocarpus glaber)-青冈(Cyclobalanopsis glauca)常绿阔叶林内100个10 m×10 m小样方的中心位置,按0—10 cm、10—20 cm和20—30 cm土层采集土壤样品,测定土壤有机碳(C)和全氮(N)含量。基于区域化变量理论和地质统计软件(GS+Version 9)的空间分析功能,应用地统计学的半方差函数定量研究该常绿阔叶林土壤有机C和全N的空间变异特征。结果表明:该林地土壤有机C含量平均值为18.61 g/kg,变化范围为9.53—39.40 g/kg,全N含量平均值为1.63g/kg,变化范围为0.73—3.32 g/kg。土壤有机C半方差函数的理论模型符合球状模型,全N半方差函数的理论模型符合高斯模型。土壤有机C和全N的空间异质性主要是由结构性因素引起的,且空间自相关程度均为中等程度。分形维数反映了有机C和全N空间格局差异及尺度依赖特征,有机C分形维数较大,空间格局比全N略为复杂。采用Kriging插值方法,1hm2森林内土壤有机C和全N具有相似的空间分布格局,呈现明显的条带状和斑块状的梯度变化。土壤有机C含量与海拔、凹凸度呈负相关,但相关性不显著,与林地凋落物量呈极显著正相关。土壤全N含量与海拔、凹凸度呈显著负相关,与林地凋落物量呈正相关,反映出土壤N的淋溶特性。

Subtropical evergreen broadleaved forests play an important role in regional carbon balance and sustainable development owing to their highest productivity,diverse ecosystem functions and complex habitat for abundant biological diversity in southern China. Spatial heterogeneity of soil nutrients in subtropical forests can provide useful information for understanding the spatial pattern of plants and for explaining to some extent,coexistence mechanism of diverse tree species.To investigate spatial variations in soil nutrients and the causes of the variations,soil samples at 0—10 cm,10—20 cm and20—30 cm depth were collected at the center of each 10 m×10 m quadrat within 1 hectare permanent plot of Lithocarpus glaber-Cyclobalanopsis glauca subtropical evergreen broadleaved forest. Soil organic C and total nitrogen（ N） concentrations were determined for all samples. Based on regional variable theory and spatial analysis functions of GS＋Version 9,spatialheterogeneity of soil organic C and total N concentrations was examined by using semivariogram of geostatistics. The results showed that averaged soil organic C concentration was 18. 61 g /kg,ranging from 9. 53 to 39. 40 g /kg,and the average value of total N concentration was 1. 63 g /kg with a range between 0. 73 and 3. 32 g /kg. Theoretical semivariogram model of soil organic C approached spherical model while the best semivariogram model of total N was close to Gaussian model. The spatial variability of soil nutrient primarily resulted from the structural factors and the spatial heterogeneity degree of those indices was moderate. Fractal dimensions from log-log semivariograms quantitatively described spatial pattern differences and scale dependence of the soil organic C and total N. Fractal dimension was high for soil organic C,so soil organic C spatial structure had strong scale dependence with a complex spatial pattern. Kriging was used to analyze the spatial distribution of soil nutrients. Spatial distribution patterns of soil organic C and total N concentrations similarly revealed an apparent belt-shaped and spot massive gradient change. Within the plot,soil organic C concentration was negatively correlated with topographic factors（ i. e. elevation and convexity）,but the relationship was not significant. Soil organic C showed very significantly a positive relationship with litter biomass. Total soil N concentration exhibited a significant negative relationship with topographic factors,however,positive relationship was found between total soil N and litter,indicating leaching characteristics of soil N. Spatial variations in soil organic C and total N highlight the importance of vegetation and litter protection in the hilly area of subtropical China.