Characteristics Of Spatial Variability Of Nutrient In Arable Soil In Luojiang County

Geostatistics combined with GIS were used in this study for the analysis of the spatial variability and influencing factors of organic matter (OM), total nitrogen (TN), available nitrogen (AN), available phosphorus (AP) and available potassium (AK) of arable soil in Luojiang county in Sichuan province and Sampling points data pretreatment, Regression analysis and prediction, Structural factors and random factors were discussed. The conclusions were as follows:(1) The concentrations of OM, TN, AN, AP and AK in the study area were deeply obviously by the exogenous factors, and had prominent spatial correlation, and their spatial variability were affected by structural factors and randomness factors and showed C/(C+C0)>0.75, in anisotropic conditions, the spatial correlation of these five kinds of nutrient almost no improve. The spatial distribution trends of concentrations of nutrient in soil had certain similarity in picture3-2, high-value areas in the Northwest and low-value areas in the southwest, high-value areas in the Northwest and low-value areas in the Southeast, but spatial distribution trends of AK was not obvious, the high-value areas distributed under different topographic conditions.(2)Had different from random sampling methods, co-kriging method could ensure the spatial correlation of variables under the initial sampling number reduction. Prediction accuracy of collaborative kriging method was higher than ordinary Kriging (in the same sample number). Under Initial sample number were reduced respectively to50%(888) and60%(710), TN or AN only could meet the prediction precision or the improvement of related coefficient, Comparative analysis of the multiple, the optimal sampling points was728in within710and888sample points.(3)The land management, site conditions and human activities related to the process and soil nutrients had significant relationship on county scale, Prediction accuracy of the concentrations of OM, TN, AN and AP was higher, but AK was worse that the smooth obviously between the predicted value and the measured value by using the linear multivariate stepwise regression method in the use of among the site conditions, soil management and soil physical and chemical properties, Soil water and soil erosion degree was proved to be the most optimal factors and soil type and typical cropping system were the second optimal factors in regression forecast, and results showed that runoff (Especially were the rain infiltration and prompt subsurface flow)played an important role at loss of soil nutrient, The spatial distribution general trend of soil nutrient reduced significantly from north to south in Luojiang county.(4)Soil parent material led to the concentrations of OM, TN, AN, AP and AK in the study area showed certain spatial heterogeneity, and affected important factors of soil natural fertility, Terrain parts and aspects caused soil nutrients showed larger space mutation characteristics, Soil erosion played an important role in the migration of arable soil nutrients, Land use and obstacles effected strongly spatial distribution of arable soil nutrients, In the comprehensive agricultural conditions, The application of chemical fertilizer had a profound impact on the spatial variability of soil OM, TN, AN, AP and AK, whereas the spatial variability of arable soil nutrients concentration were mainly affected by The application of chemical fertilizer, agricultural population and area grain sowing.