小尺度下烟田土壤养分空间变异性研究

Spatial Variability of SoiI Nutrients in Tobacco Fields Based on Geostatisics with Small Scale

为探讨烟田精准施肥，应用地统计学，结合地理信息系统（GIS）技术分析了小尺度下烟田土壤碱解氮、速效磷、速效钾、有机质和pH等5种养分含量的空间分布特征。结果表明，土壤养分的变异系数为4.43%～33.45%；半方差函数分析结果显示，除有机质具有强空间相关和碱解氮不具有空间相关性外，其余土壤养分均具有中等空间相关性。碱解氮为线性模型，pH为球形模型，其余为指数模型。利用克里格最优插值，绘制了土壤养分碱解氮、速效磷、速效钾、有机质的空间分布图，并分析了其含量分布状态，可为烟田的精准施肥提供依据。

Spatial distribution of soil available nitrogen, available phosphorus, available potassium, organic matter, and pH in tobacco fields was studied to provide theoretical basis for precision agriculture by using geostatistics and GIS technique on small scale. The results indicated that soil available nutrients had medium variability with coefficient of variation from 4.43%to 33.45%. Semivariogram analysis showed that soil nutrients were moderately spatially dependent in a given spatial range, except that soil organic matter was strongly spatially dependent and available nitrogen was not spatially dependent. The semivariograms of soil nutrients were best described by exponential model, except for that of available N, which was best fitted by linear model and pH was best fitted by spherical model. The distributing maps of soil nutrients were drawn by using ordinary kriging, and the distributing results of soil nutrient were analyzed. The study provides an important basis for precision fertilization in tobacco production.