Spatial Variability Of Coastal Saline Soil Nutrients And Salinity Of Yellow River Delta

Soil is the production base of agriculture, forestry and animal husbandry. Spatial structure of soil characteristics is the result of the interaction of variety of factors (biology, chemistry, physics), however, the influence of these factors for different areas may be different. Clearing the spatial distribution of soil characteristics is important for understanding the process of regional soil, revealing natural process and pattern of interaction, establishing the reasonable management mode. The land use types and ecological environment of Yellow River Delta exist large spatial heterogeneity. Based on the soil heterogeneity, it is significant for agricultural management of this area.Taking the northern Yellow River Delta as study area, this article analyzed the spatial distribution characteristics of soil nutrients and salinity and related affecting factors by using mathematical statistics and geostatistics methods. The study contents mainly included:spatial variability of soil nutrients and salinity based on measured data; spatial variability of soil salinity based on inversion data; spatial variability of soil nutrients and salinity based on belt transect method; the analyses of influencing factors and proposing some treatment measures.The main research conclusions are as follows:(1) From the analysis of descriptive statistics, spatial structure, spatial distribution of5kinds of soil characteristics index, the content of soil organic matter (7.66g·kg-1) and total nitrogen (0.36g·kg-1) were low, and the content of available phosphorus (12.26mg·kg-1), available potassium (216.80mg·kg-1) and salinity (5.72g·kg-1) were higher. Five kinds of soil index were better fitting with half variance function, and indicated to have obvious spatial structure characteristics. The spatial distribution of different indexes existed different trends.(2) The combine of soil spectra data and chemical analysis data can reveal spatial-temporal variability of soil, and achieve the dynamic monitoring of soil physical and chemical properties. Taking soil salinity for an example, this paper set up the mathematical relationship model between spectrum reflectance and the content of soil salinity. By testing, the regression model based on BP neural network was better than multiple linear regression model. The trend of spatial distribution based on inversion data and measured data was basically same, and provided a new and fast method to monitor soil information.(3) In north-south transects, the nugget/still of soil organic matter (SOM). total nitrogen (TN), available phosphorus (AP), and available potassium (AK) contents and soil salinity were0.38.0.40,0.50,0.32, and0.34. respectively, which demonstrated that these five parameters were moderately spatial dependence. In east-west transects, the nugget/still of five parameters were1.00,0.40,0.51,0.72and0.34, which demonstrated SOM had weak spatial correlation. The spatial distribution of two transects were also different.(4) The spatial distribution pattern of soil nutrient and salinity was closed related to soil parental material, soil texture, distance to sea, road block, land use pattern and management measure. On the basis of characteristics of different zones, more targeted measures were put forword.