基于GIS的缓坡烟田土壤养分空间变异研究

Spatial variability characteristics of soil nutrients in tobacco fields of gentle slope based on GIS

综合运用地统计学和GIS相结合的方法,分析了缓坡地形下土壤养分的空间变异规律,并绘制养分空间分布图,为山地缓坡烟田养分分区及精准施肥决策提供理论依据。结果表明:研究区内4种养分含量均具中等程度变异。土壤有机质、碱解氮、速效磷和速效钾的变程分别为61.8、76.3、70.5 m和57 m。土壤速效钾的分形维数最高,有机质和碱解氮其次,速效磷最低。有机质和速效钾的最适模型为指数模型,块金系数分别为30.9%和31.1%;碱解氮和速效磷可用球状模型进行较好拟合,其块金系数分别为37.7%和26.4%,4种养分均有中等程度的空间相关性。各向异性和趋势性分析均显示,有机质和速效磷具有较强的各向异性,碱解氮和速效钾的各向同性范围最广。研究区4种养分的空间分布格局在海拔和坡度的影响下呈现出一定的规律性,低值均在坡度较大的中北部出现,东北部和南部较平缓区域出现高值。

Soils are highly variable spatially due to the combined effects of physical, chemical, and biological processes that operate with different intensities and at different scales. Understanding the spatial variability of soil properties is essential in determining local fertilizer needs of tobacco. The spatial variability of soil nutrients in terrain with a typical gentle slope was comprehensively analyzed by a combination of geostatistics with a Geographic InfOrmation System （GIS） platform. Scatter diagrams of the spatial distributions of soil nutrients were then quantitatively constructed. The study potentially provides theoretical bases not only for the use of management zones （MZs） of tobacco fields in gentle mountain-terrain slopes, which are field subdivisions that have relatively homogeneous attributes of landscape and soil conditions, but also for decisionmaking with respect to precision fertilization. Soil samples （O-20cm） were taken at 88 points on an approximately 25-m grid using a global positioning system to define sample locations. The resuhs showed moderate levels of spatial variability of four measures of nutrient contents in the research region. The contents of soil OM, TN, AN and AP values were low, and their means were 8.89 g/kg, 26.76 mg/kg, 1.92 mg/kg and 130.23 mg,/kg, respectively. Organic matter （OM） with a nugget/still ratio 30.9%, and available K （AK） with a nugget/still ratio 31.1%, could be modeled with an exponential model. The nugget/still ratios of available N （AN） and available P （AP） were 37.7% and 26.4%, respectively, and they could be well modeled with a spherical model. As a result, all the nutrients exhibited a moderate spatial correlation. The range of soil OM, AN, AP and AK values were 61.8 m, 76.3 m, 70.5 m and 57 m, respectively. The fractal dimension （D） of AK was the highest, followed by OM and AN, and AP was the lowest. Anisotropic analyses and trend analyses all showed that OM and AP were even strongly anisotropic. AN and AK showed the widest scope in isotropism.The spatial distributions of the four measures of soil nutrients in the study region, which were influenced by elevation and slope and represented some regularity, were characterized by the mean contents of all nutrients being low in the north central area with a relatively steep slope, while being high in the comparatively flat northeast and south regions.