基于地统计学的农田尺度旱地红壤养分空间变异性研究  

A Study On Spatial Variance of Soil Nutrients on the Farmland Scale in Dryland of Red Soil Based on Geostatistics

作　　者：刘祖香[1] 陈效民[1] 靖彦[1] 张佳宝[2] 黄欠如 李秋霞[1] 

机构地区：[1]南京农业大学资源与环境科学学院,江苏南京210095 [2]中国科学院南京土壤研究所,南京210008 [3]江西红壤研究所,江西进贤331717

出　　处：《土壤通报》2013年第2期392-397,共6页Chinese Journal of Soil Science

基　　金：国家973计划项目"典型地区农田地力定向培育理论与技术对策"子课题"典型红壤区农田地力定向培育与提升技术研究"(NO:2011CB100506);土壤与农业可持续发展国家重点实验室项目(0812201208)资助

摘　　要：运用地统计学与经典统计学相结合的方法研究了农田尺度下旱地红壤有机质、全氮、铵态氮、硝态氮、有效磷和CEC的空间变异状况，研究结果表明：（1）土壤表层（0～15em）和亚表层（15～30em）的养分变异系数各不相同，有效磷、铵态氮和硝态氮为中等变异，有机质、全氮和CEC为弱变异。（2）经半方差检验结果表明，在0—15erfl土层中有机质和CEC均在变程为55m处表现较弱的空间相关性；全氮、硝态氮和有效磷分别在变程为55m、44m和8m处表现出中等空间相关性；铵态氮在变程为15m处表现出强烈的空间相关性；在15—30em的土层中硝态氮和CEC在变程为21m和59m处表现弱的空间相关性；有机质、全氮和有效磷分别在变程为8in、20m和55m处表现出中等的空间相关性；铵态氮在变程为15m处表现出强的空间相关性；除了表层土壤有机质的理论模型为高斯模型和CEC的理论模型为球状模型外，其余指标均为指数模型。（3）利用克立格插值方法绘制旱地红壤的有机质、全氮、硝态氮和CEC空间分布图，得出有机质和CEC为不规则的斑块状分布，全氮和硝态氮呈条带状和块状分布。研究结果为了解旱地红壤养分状况、预测空间养分分布状况和指导农业上的精确与合理施肥提供了科学依据。This study was conducted on the basis of Geostatistical Analysis and classical statistics to evaluate the spatial variance and changes in soil organic matter （SOM）, total nitrogen （TN）, ammonium nitrogen （AN） and nitrate nitrogen （NN）, available phosphorus （AP） and CEC contents in dryland of red soil. The results were as follows： （1） The coefficients of variance of soil surface （0 - 15 cm） nutrients were different from subsurface （15 - 30 cm）. The variance coefficients of AP, AN and NN were at moderate level, and the others were low. （2） In 0 - 15 cm soil layer, the semi-variance analysis results showed that the SOM and CEC had a weak spatial correlation at major range of 55 m. There were moderate differentiations of TN, NN and AP at major range of 55 m, 44m and 8m, respectively. The AN was a strong spatial correlation at major range of 15 m. In 15 - 30 cm soil layer, the NN and CEC showed a weak spatial correlation at major range of 21 m and 59 m. The SOM, TN and AP had a moderate spatial correlation at major range of 8 m, 20 m and 55 m, respectively. AN had a strong spatial correlation at major range of 15 m. Surface SOM and CEC were described by Gaussian and spherical models, and the others were index models. （3） From the SOM, TN, NN and CEC spatial distribution maps drawn by Kriging interpolation method, the distributions of SOM and CEC showed irregular patchy characteristics, and then TN and NN showed band and block characteristics. The above results provided scientific data for understanding soil nutrients, predicting spatial nutrients distribution and guiding precise and rational fertilization in red soil area.

关 键 词：土壤养分 地统计学 变异系数 半方差分析 空间分布 

分 类 号：S152.7[农业科学—土壤学] S158.3[农业科学—农业基础科学]