Spatial Distirbution And Its Influencing Factors Of Soil Nutrients In Matiyu Small Watershed Based On Regression-kriging Model

As the dynamic nature with loose state existing in land surface and possessingfertility, soil was an important research content in small watershed comprehensivemanagement, and there had important significance to guide ecological environmentmanagement, regional soil resource use and management and sustainable managementof agricultural production by studying the spatial variation and spatial distributioncharacteristics of soil nutrients, especially the researching of soil nutrients’quantification process.By analyzing the representativeness and operability of land use and sample pointsin general, samples from97sites （0-20cm） were collected during October2010inMatiyu small watershed in hilly area of middle southern Shandong Province. Byemploying regression-kriging methods which combine geostatistics and ‘3S’technology to analyze the spatial distribution characteristics of soil organic matter,total nitrogen, available nitrogen, total phosphorus, available phosphorus and discussthe influence to spatial distribution of surface soil nutrients contents in smallwatershed under different land use patterns, terrain factors, vegetation coveragefactors and so on. The main results were as follows:（1） Except available phosphorus, the spatial distribution of soil nutrients werecomparatively uniform, and the coefficient of variation were all belongs to moderatevariability. Among those soil nutrients, the mean value and medium value of soilorganic matter content were20.45g/kg and19.96g/kg respectively, the mean valueand medium value of soil total nitrogen content were1.18g/kg and1.15g/kgrespectively, the mean value and medium value of soil available nitrogen content were90.27mg/kg and88.95mg/kg respectively, the mean value and medium value of soiltotal phosphorus content were1.00g/kg and0.95g/kg respectively, the mean valueand medium value of soil available phosphorus content were8.06g/kg and7.42g/kgrespectively.（2）Regression-kriging models of the soil nutrients in the Matiyu small watershedwere established and verified, by comparing the C₀/（C₀+C） and R2from different models, spatial autocorrelation of residual error of the regression-kriging predictionmodel from soil organic matter, total nitrogen, available nitrogen, availablephosphorus were all medium（>25%）, the C₀/（C₀+C） were33.93%、25.58%、31.63%and29.14%respectively, however, spatial autocorrelation of residual error of theregression-kriging prediction model from total phosphorus was intensity, and theC₀/（C₀+C） was24.46%（<25%）. Semivariogram test showed that structural factors liketerrain, vegetation and climate were the main reason to arouse spatial variation of soilnutrients, while random factors had little influence. From the predication accuracy ofsoil nutrients’ spatial distribution, the result of regression-kriging was comparativelyideal, its’ predication value was quite close to the measured value, and its’ predicationaccuracy was better than ordinary kriging.（3） The distribution map of regression-kriging showed that soil organic mattercontent decreased from the southeast to the northwest, which was similar to the trendof DEM in Matiyu small watershed. The higher value was distributed in southwestand southeast around the watershed boundary, which was between23.01g/kg and31.09g/kg, the lower value was distributed in watershed outlet, which was between13.85g/kg and18.21g/kg. The research showed that both land use patterns andelevation had much more remarkable influence than aspect factor to the distribution ofsurface soil organic matter in small watershed.（4）The distribution map of regression-kriging showed that both soil total nitrogenand available nitrogen contents decreased from the southeast to the northwest, whichwere similar to the trend of DEM in Matiyu small watershed, especially the availablenitrogen which was distributed successively in an hierarchy state around thewatershed. The higher value of total nitrogen was distributed in extreme southern partin small watershed, which was between1.56g/kg and1.89g/kg, the lower valueappeared in the watershed outlet, which was between0.89g/kg and1.13g/kg. Whilethe higher value of available nitrogen was distributed around the watershed boundarypart in small watershed, which was between111.61mg/kg and134.73mg/kg, thelower value appeared in the watershed outlet, which was between52.82mg/kg and70.81mg/kg. The research showed that land use patterns, elevation and aspect those facors all had remarkable influence to the distribution of surface soil total nitrogenand available nitrogen in small watershed.（5）The distribution map of regression-kriging showed that both soil totalphosphorus and available phosphorus contents increased from the southeast to thenorthwest, which were opposite to the trend of DEM in Matiyu small watershed. Inthe distribution map of soil total phosphorus and available phosphorus, both thehigher value were distributed in northwest part in samll watershed, which wererespectively between1.20g/kg and1.58g/kg,11.18mg/kg and14.99mg/kg, thelower value was distributed in southern part in small watershed, which wererespectively between0.49g/kg and0.66g/kg,1.30mg/kg and4.15mg/kg. Theresearch showed that land use patterns, elevation and NDVI those facors all had muchmore remarkable influence than sediment transport index （STI） to the distribution ofsurface soil total nitrogen and available nitrogen in small watershed.