喀斯特地区土壤空间异质性及采样方法研究

发布时间：2018-10-26 11:50

【摘要】：土壤采样是土壤研究的根本,也是造成土壤研究误差的主要来源,尽管已出台大量采样规程,但在土壤科学向更精细的方向发展的大趋势下,土壤采样的精度提高是土壤科学向前发展的最大障碍。土壤采样方法的研究和更新已成为土壤科学向前发展必须攻克的难关。本次研究以喀斯特石漠化退耕坡地、喀斯特森林样地和白云岩草地三种类型共5个样地为研究对象,以土壤有机碳作为典型指标,通过对样地空间异质性研究为基础对常用采样方法在喀斯特样地2000次重复模拟比较,得出以下几点主要结果和结论:1.喀斯特地区石灰岩样地各小生境间土壤碳差异显著,按照小生境对喀斯特石灰岩样地土壤进行分类是比较科学的。2.相对与采样方法,采样点数量对样品准确率和准确度的影响要大得多。3.采样的准确率和准确度与采样点数量之间线性回归结果如下:采样方法采样准确率采样准确度回归模型决定系数回归模型决定系数“S”型法y=0.4653e4.9053x R2=0.8573 y=0.8058e4.5397x R2=0.7957梅花法y=0.5433e4.7505x R2=0.8508 y=0.666e4.9282x R2=0.8952随机法y=0.5438e4.9426x R2=0.9498 y=0.7058e5.0536x R2=0.8902网格法y=0.4016e5.3666x R2=0.9101 y=0.6443e5.1035x R2=0.8728小生境法y=0.5104e4.5613x R2=0.6714 y=0.8731e4.6995x R2=0.90394.从样品的代表性和重现性角度,小生境法和“S”型法是需要采样点数量最少的两种采样方法。5.综合样品的代表性、重现性及采样方法的可操作性,在喀斯特地区进行采样时,对于面积较小且小生境分布集中明显的样地,小生境法是最适方法;面积较大、小生境分布复杂或无小生境的样地,“S”型法是最适方法。6.要达到80%的采样准确率和70%的采样准确度,采用小生境法在喀斯特400m2样地进行采样时,至少采集23个采样点;采用“S”型法则至少需要24个采样点。
[Abstract]:Soil sampling is the basis of soil research and the main source of errors in soil research. Although a large number of sampling procedures have been introduced, but in the general trend of soil science developing in a more precise direction, The improvement of soil sampling precision is the biggest obstacle to the development of soil science. The research and renewal of soil sampling methods has become a difficult problem for the development of soil science. In this study, five plots of karst rocky desertification, three types of karst forest and dolomite grassland were taken as the research objects, and soil organic carbon was taken as the typical index. Based on the study of spatial heterogeneity of sample plots, the following main results and conclusions are obtained: 1. 2000 repetitions of sampling methods in karst samples are obtained. There are significant differences in soil carbon among different microhabitats in limestone sample plots in karst area. It is more scientific to classify karst limestone soil according to niche. 2. Compared with the sampling method, the number of sampling points has much greater influence on the accuracy and accuracy of the sample. The results of linear regression between the accuracy and accuracy of sampling and the number of sampling points are as follows: sampling accuracy regression model determination coefficient "S" method y=0.4653e4.9053x R2O 0.8573 Y=0.8058e4.5397x R2O0.7957 y=0.5433e4.7505x R2O0.8508 y=0.666e4.9282x R2O0.8952 Random y=0.5438e4.9426x R2O0.9498 y=0.7058e5.0536x R2O0.8902 mesh method y=0.4016e5.3666x R20.9101 y=0.6443e5.1035x R2O0.8728 niche method Y0 5104e4.5613x R2P 0.6714 y=0.8731e4.6995x R2P 0.90394. From the point of view of representativeness and reproducibility of samples, niche method and "S" method are two sampling methods that require the least number of sampling points. The representative, reproducibility and maneuverability of the sampling method are integrated. When sampling in karst area, the niche method is the best method for the sample with small area and obvious niche distribution. The "S" type method is the most suitable method for the sample plots with large area and complex distribution of microhabitats or no microhabitats. In order to achieve the sampling accuracy of 80% and 70%, at least 23 sampling points are collected when the niche method is used in the karst 400m2 sampling, and at least 24 sampling points are needed to adopt the "S" type rule.
【学位授予单位】：贵州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S151.9

【参考文献】