| The soil is the root of agriculture, and the coordination of soil’s nutrient supply and crop’s nutrient demand is the key factor in limiting high yield and land quality. With the progress and development of human society, the contradiction between population and the environment have become increasingly prominent:water pollution, heavy metal’s pollution, excessive use of industrial fertilizers, irrational utilization of soil resources and so forth, which make the soil resource problem more serious. Precision agriculture, which research and implementation depend on the precise spatial information of soil characters, gives the direction of agricultural development by modern science from macro and micro aspects. Therefore, the research on spatial variability of soil nutrient is significant to increase the utilization of soil nutrients and realise fertilizer management of precision agriculture.Zhongxiang City is located in a hilly region of Hubei, is transition zone from Jianghan Plain to the mountainous region of West Hubei, and the Han River divides the Zhongxiang City into east and west part in the middle from south to North, and the two parts have the approximately symmetric topographical pattern. The complexity and flexibility of terrain not only affect the distribution of water, fertilizer, gas heat, but also the soil-forming process. The physical and chemical properties of soil are different according to various terrain and location. With geo-statistics combining GIS spatial analysis and digital terrain analysis methods, the paper takes conventional soil nutrients and DEM as the underlying data source, discusses the effect of terrain on soil properties, analyses spatial distribution of soil nutrients in the research area, and put forward efficient and convenient method of sampling based on the road. The research contents and the main conclusions are as follows:(1)6825sampling points are divided into four parts:the western hills, western plains, eastern plains, eastern hills on the basis of distribution characters of terrain of the research area, and the study gives the descriptive statistics of Organic Matter, Available N, Available P Available K and pH value of different terrain and location separately. The results show that soil nutrient content in the hills is higher than the plain, the variability coefficient of soil nutrient in the research area ranging from11.35%to68.91%is intermediate variability; the variability in the plain is greater than the hills, and the soil nutrient content in the west is higher than the east on average.(2) Geo-statistics analysis shows that the nugget and sill ratio of Organic Matter and Available N in the four different locations ranging from0.27to0.53belong to intermediate variability, the spatial variation of Available P is diverse largely. Besides, the nugget and sill ratio in the plain and eastern hills is over0.82while the ratio of Available P in western hills is only0.054, because Available P is prone to mineralization and effect by anthropogenic interference. Spatial variation law of soil nutrients is changed by natural environment and anthropogenic interference.(3) Use DEM of the research area to extract terrain factors, including slope, aspect, plane curvature, slope curvature and topographic wetness index to process correlation analysis and autocorrelation analysis of soil nutrients and topographic factors. The results show that correlation of soil nutrients almost reach an extremely significant level (P<0.01) while the correlation between aspect and terrain factors, soil nutrient is weaker, which accounts for the influence of the aspect on soil nutrients is not obvious.(4) Process regression analysis on the soil nutrient and the terrain factors based on the correlation between soil nutrient and terrain factors, and establish regression equation using Lagrange’s multiplier P value to choose the spatial lag model. Predict distributions of soil nutrients in the research area based on quantitative relationship, and the Kriging interpolation results show clearly that terrain controls distribution of soil nutrient content in certain circumstances.(5) Given limitation of the traditional sampling, the paper proposes the method of sampling based on the road according to density and distribution of the road in the research area. The sample size is decided by soil properties, and the sampling scope of the road is based on confidence and confidence interval, and the paper determines660samples at least on either side of the road within50meters in the situation of95%confidence and5%relative error, while183samples at least on either side of the road within30meters in the situation10%relative error can represent the distribution of soil nutrients in the research area. The study shows that the method of sampling based on the road is reasonable to describe spatial distribution of soil nutrients in the research area, but to the scope of application and promotion of the method need further support of data. |
PDF Full Text Request