具域耕地土壤养分空间变异及肥力综合评价研究

发布时间：2018-06-07 08:09

本文选题：土壤养分 + 地统计学与GIS　；参考：《昆明理工大学》2015年硕士论文

【摘要】：本研究以云南省昌宁县耕地土壤为研究对象,以研究区2012年测土配方施肥数据为基础,探讨了耕层(0-20cm)土壤pH值、有机质(OM)、全氮(TN)、碱解氮(AHN)、有效磷(AP)、速效钾(AK)、有效锌(Zn)、有效锰(Mn)、有效钼(Mo)等9种现状及空间分异情况,重点研究和分析了丰缺格局、影响土壤养分变化的因素,并进行了综合肥力进行评价；在保障精度的条件下,进一步优化采样数量。主要研究结果为：(1)统计分析表明,各个土壤养分的变异系数(CV)在13.08%-50.15%之间,PH、有机质、氮、磷、钾、锰、锌、钼的变异系数均在10%-50%之间,具有中等的空间变异性。pH的块基比C0/(Co+C)都小于25%,说明具有强烈的空间相关性,有机质、氮、磷、钾、锰、锌、钼的块基比C0/(Co+C)值在25%-75%之间,表现出中等空间相关性。PH是线性模型,有机质、有效锰是球状模型,全氮、碱解氮、速效钾、有效磷、有效铝是指数模型。(2)地形因子和其他环境变量(土壤类型、耕地制度、母质类型等)作为辅助变量,建立了县域的土壤养分回归克里格预测模型,结果表明回归克里格插值精度高于普通克里格和协同克里格插值的精度。(3)依据《云南省养分含量等级表》的分级标准,进行空间格局分析。利用主成分分析和BP神经网络算法对土壤综合肥力进行评价,两种方法都能有效的表达土壤综合肥力状况。(4)采用协同克里格方法和普通克里格法对土壤养分采样数进行优化,得出在不同样点数目下的协同克里格插值精度优于普通克里格插值精度,并确定研究区的最大采样间隔。(5)不同土地利用方式下对耕层土壤8种元素(除有效锌)差异性显著；土壤类型对养分元素均产生了显著水平；不同的成土母质对土壤养分元素含量产生了显著水平,除有效磷外；海拔对9种养分元素影响显著；阴坡和半阴坡方向比阳坡更有利于土壤养分的积累,坡向对AP、Mn、Mo影响不显著；坡度与全氮不显著。
[Abstract]:In this study, the cultivated land soil in Changning County, Yunnan Province was taken as the research object, and the pH value of the cultivated layer (0-20 cm) was discussed on the basis of the soil formula fertilization data from 2012 in Changning County, Yunnan Province. Organic matter (OMN), total nitrogen (TNN), alkali-hydrolyzed nitrogen (AHN), available phosphorus (APN), available potassium (AK), available zinc (Zn), available manganese (mn), available molybdenum (Mo) and their spatial differentiation were studied and analyzed in detail. The factors affecting soil nutrient change were studied and analyzed. The comprehensive fertility was evaluated and the sampling quantity was optimized under the condition of ensuring precision. The main results were as follows: (1) the coefficient of variation (CV) of soil nutrients was between 13.08% and 50.15%, and the coefficients of variation of organic matter, nitrogen, phosphorus, potassium, manganese, zinc and molybdenum were between 10% and 50%. The ratio of block to base (C0/(Co C) of medium spatial variability. PH is less than 25, indicating that there is a strong spatial correlation. The ratio of organic matter, nitrogen, phosphorus, potassium, manganese, zinc and molybdenum is between 25% and 75%, and the ratio of mass to base of organic matter, nitrogen, phosphorus, potassium, manganese, zinc and molybdenum is between 25% and 75%. The medium spatial correlation. Ph is linear model, organic matter, available manganese is spherical model, total nitrogen, alkali-hydrolyzed nitrogen, available potassium, available phosphorus, available aluminum is exponential model. 2) topographic factors and other environmental variables (soil type, cultivated land system, etc.) As auxiliary variables, the regression Kriging prediction model of soil nutrients in county area was established. The results showed that the precision of regression Kriging interpolation was higher than that of common Kriging and cooperative Kriging interpolation. Principal component analysis (PCA) and BP neural network (BP) algorithm were used to evaluate soil comprehensive fertility. Both methods can effectively express soil comprehensive fertility. (4) Co-Kriging method and common Kriging method were used to optimize soil nutrient sampling number. The results show that the precision of cooperative Kriging interpolation under different points is better than that of common Kriging interpolation, and the maximum sampling interval of the study area is determined. 5) the difference of 8 elements (except available zinc) in topsoil under different land use patterns is significant. Soil types had significant effects on nutrient elements, different parent materials had significant effects on soil nutrient contents, except available phosphorus, elevation had a significant effect on 9 nutrient elements. The direction of shady slope and semi-shady slope is more favorable to the accumulation of soil nutrients than that of sunny slope, and the slope direction has no significant effect on APGMnMn-Mo, and the slope and total nitrogen are not significant.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S158

【参考文献】