桂西北喀斯特洼地土壤有机碳和速效磷的空间变异

空间变异性的研究有助于认识土壤养分的空间分布特征与生态过程之间的关系。采用网格取样(20m×20m),利用地统计学方法分析了典型喀斯特洼地土壤有机碳(SOC)和速效磷(AP)的空间变异特征。结果表明SOC和AP的半变异函数分别用球状模型和指数模型拟合的效果较好。块金效应对SOC和AP总变异的贡献分别为37.9%和49.8%,说明SOC主要受结构性因素控制,而AP受结构性因素和随机因素双重控制。SOC的分布呈明显的带状各向异性特征,在洼地的长轴方向具有较大的变程和较小的基台值,其全向半变异函数的相关距离为135.5m,与洼地不同种植类型地块的覆盖尺度基本相当。AP的分布特征与SOC存在明显差异,具有较大的漂移趋势,其相关距离为413.4m;但在分离漂移趋势后,其相关距离为167.4m,反映了不同地块间的施肥差异对AP的影响较大。喀斯特洼地SOC的空间分布主要受地形和土地利用结构控制,而AP的分布的规律性不明显,可能与施肥等因素的随机性较强有关。

Spatial variability of soil organic carbon and available phosphorus in a typical Karst depression, northwest of Guangxi

Spatial variability of soil nutrients provides useful information for improving agricultural practices and ecological management. Peak-cluster depression forms a typical Karst landscape with agriculture mainly based in depressions, such as in the Karst region in south-west China. In this study, the depression area in a typical landscape unit selected from the Karst region of Guangxi province was investigated for the spatial variability of soil organic carbon (SOC) and available phosphorus (AP) in the surface soil (0-15 cm). Grid sampling method (20 m×20 m) and geological statistics were applied to analyze the spatial distribution of SOC and AP contents. Sample variograms of SOC and AP were fitted well by spherical and exponential models, respectively. Nugget effects due to random errors or variability within the distance of the sampling interval (20 m) contributed to 37.9% of the overall variation for SOC, and 49.8% for AP. This suggests that the spatial variation in SOC was mainly controlled by structural factors, which mainly depended on topographic conditions and land use types in the study area, but AP rather more controlled by random factors mainly due to the variability of fertilizer application which differed widely in the depression. Correlation length of SOC was about 135.5 m, which had similar scope to the average range of difference in land use types. Based on directional semi-variogram analysis, zonal anisotropy, and lower sill and longer range in the long axis direction were found in the spatial distribution of SOC in the depression. AP also had anisotropy, but the pattern in its spatial distribution was quite different from that of SOC, as AP had much longer correlation length (413.4 m) and showed clearly a drift trend which enhanced variability in the long axis direction, presumably due to the random application of P-fertilizers. If the effects of drift trend were subtracted, the correlation length of AP decreased to 167.4 m. This study suggests that the pattern in the spatial distribution of SOC in the depression area in Karst regions was mainly controlled by structure factors (topographic conditions and land use types), but for AP rather more by random factors such as fertilizer application.