| Based on the study of farmland soil in Changning County in Yunnan Province arid the study area in 2012 formula fertilization by soil testing data, this paper has discussed 9 nutrients of situation and the spatial variability of topsoil (0-20 cm), which include pH, OM, TN, AHN, AP, AK, Mn, Zn, Mo,mainly researched and analyzed of the factor abundance pattern, the effects of soil nutrient changes, and comprehensive evaluation of fertility; and optimized study area sampling quantity in the condition of the guarantee of accuracy. The main research results as follows:(1) Analysis showed that coefficient of variation of each soil nutrients (CV) were between 13.08% and 50.15%; the coefficient of variation PH, OM, TN, AHN, AP, AK, Mn, Zn, Mo were between 10%and 50%, that’s showing the moderate spatial correlation. The ratio of nugget to Co/(Co+C) of PH was less than 25%, showing strong spatial correlation; while the ratio of nugget to Co/(Co+C) of OM, TN, AHN, AP, AK, Mn, Zn, Mo rangend from 25%-75%, showing moderate spatial correlation. PH was fitted by line models, OM and Mn were fitted by spherical models, TN, AHN, AK, AP, Mo were fitted by exponential models.(2) Topographic factors and other environmental variables (soil type, land system, the type of parent material etc.) are able to establish soil nutrients in County Regression Kriging prediction model. The results show that the Regression Kriging precision was higher than Co-Kriging and Ordinary Kriging.(3) Each nutrient choose high precision of interpolation method in order to analysis spatial pattern on the basis of "classification standard of Yunnan province nutrient’s criterion". To evaluate the soil fertility by principal component analysis method and BP neural network algorithm, that’s showing the two methods can express the integrated soil fertility status.(4) Soil nutrient sampling number was optimized by Co-Kriging and Ordinary Kriging. The results show that Ca-Kriging prediction accuracy was higher than the ordinary Kriging with the sample gradually decreasing and embodying obvious superiority, and determined the maximum sampling spacing in the study area.(5) The different land-use types, remarkably infuenced the contents of nutrients, in addition to Zn; Soil types have generated the significant differences for elements; different soil parent material content has a significant level of soil nutrient elements, except AP; elevation with 9 kinds of nutrient elements were significantly correlation. The schattenseite is more conducive to accumulation of soil nutrients than sunny slope; Aspect has not heavy impact on AP, Mn and Mo;.Slope has not impact on TN.
|
PDF Full Text Request