基于综合景观单元土壤调查样点布设密度判断机制研究

The design-based sampling strategies for regional soil survey based on synthesis landscape units is still approved by soil scientists strongly, for which follows to the conventional soil and statistics theory, and is easy to understand and operate. Among them, soil sampling density design depends on the number of the synthesis landscape unit type and its internal soil properties spatial variation in a target region, both with strong scale (grain, extent) effect. While the internal soil properties spatial variability of the synthesis landscape unit at a certain scale cannot be given correctly at present, due to the lack of its quantitative characterization and prediction mechanism, which seriously affect the scientific decision of soil sampling density design. . This project will be performed in a typical low hilly red soil region of southern China. One thousand soil samples will be collected in the synthesis landscape units at multi-scales (grains, extents), according to a sampling design by using a multi-objective stratified nested soil sampling approach. Soil properties will be analyzed at lab, such as soc, tn, tp, tk, tp, ph. Landscape indices of these multi-scales synthesis landscape units such as patch size, density, richness, diversity and shape index will be obtained by GIS technology, too. Through the secondary sampling strategy, and conventional statistics and geo-statistics analysis technology, this project aims to: reveal scale (grain, extent and coupling) effects on the internal soil properties spatial variability, and its characterization and prediction mechanism with scale index and landscape index; create a mechanism and method of soil sampling density design based on synthesis landscape units, which design satisfy the scale (grain and extent) requirement for regional soil survey; and put forward a combined optimization mechanism of soil sampling designs for multi-properties survey at same time. . This project achievement will provide new scientific basis and method for soil sampling density design based on synthesis landscape units in such region with complex geographical and environmental conditions, as low hilly red soil region of southern China.

基于综合类型景观单元的区域土壤调查抽样设计方法，遵循传统土壤及统计学理论、易于理解、操作简便，始终被土壤学家所重视。其中，土壤样点密度设计依赖于区域综合景观单元类型数目以及其内部土壤属性空间变异性，两者均具有极强的尺度（粒度、幅度）效应。但特定尺度综合景观单元内部土壤属性空间变异性还不能给定，缺乏定量表征和预测机制，严重影响样点密度的科学设计。本项目将在我国南方典型红壤低山丘陵区，针对不同尺度综合景观单元，利用多目标多层次嵌套式土壤样点设计技术进行布点、采样和分析；通过二次抽样、统计及地统计等实验分析手段，研究揭示综合景观单元土壤属性空间变异的尺度（粒度、幅度及耦合）效应及其表征和预测机制，建立基于综合景观单元、满足尺度（粒度和幅度）要求的土壤样点布设密度判断机制及方法，提出多土壤属性同时调查的样点布设方案联合优化机制，为复杂环境条件地区土壤调查样点方案设计提供新的科学依据及方法。

基于类型单元的区域土壤调查样点布设方法，增强了样点代表性和土壤属性及地理空间覆盖，但样点密度判断问题依然存在。它不仅依赖于景观单元类型数目，也依赖于景观单元内部变异性。然而，对特定幅度和粒度下各类型单元内部土壤属性空间变异，目前还缺乏定量表征及预测机制，始终是样点密度科学判断的瓶颈障碍。.项目在典型南方红壤丘陵区，针对不同尺度（幅度、粒度）类型单元，利用嵌套式样点设计方法进行布点、采样和分析，旨在研究揭示类型单元土壤属性空间变异尺度效应及其表征与预测机制，建立满足幅度和粒度要求的土壤调查样点密度判断机制及方法，提出多土壤属性同时调查的样点布设优化机制，为土壤调查样点方案设计提供理论方法。. 研究表明，随着土壤类型分类粒度加大，类型单元内部各土壤属性变异增强，13个和15个景观指数分别呈现上升和下降趋势，利用粒度指标进行可定量表征，但土壤属性变异表征存在缺陷，而利用上述景观指数可分别进行定量表征；基于显著性水平和决定系数最高原则，研究提出了各土壤属性变异的最佳表征函数及景观指数，并得到验证；TN、TP、AP、AK这4种土壤属性变异可通过幅度与粒度效应耦合进行多元线性定量表征。利用类型单元景观指数，可为不同调查粒度要求的区域样点密度判断提供土壤属性变异性特征数据。. 类型单元各土壤属性变异幅度效应特征明显，利用样点最大距离和面积指数均可进行定量表征。除pH外，各土壤属性变异系数随幅度扩增从急速增加到增加缓慢、直至平稳趋于极限，均存在不再明显增强的起始特征幅度。对超出特征幅度的调查区域，先将它划分若干特征幅度样区，确定每个特征幅度区样点数，再综合判断调查区样点密度。. 针对多种土壤属性同时调查的样点布设方案联合优化机制，提出基于最大密度布点、再按各自密度进行二次抽样的策略。. 应用表明，建立的满足给定调查幅度、粒度和精度要求的样点密度判断机制及方法切实可行，可为区域土壤调查样点设计提供方法论指导。