水保所知识产出(1956---)知识库

采样设计是土壤有机碳研究中面临的首要问题。论文对黄土高原不同地貌类型区农田
土壤有机碳进行抽样样本代表性评价，对比分析了不同采样方法的样点布设效率，结果表明：
1）高塬区合理采样布点方法应为网格布点法，采样点布设格网间隔4 km，网格布点法较随机布
点法和联合单元布点法效率分别提高64.3%和31.8%；2）平原区合理采样布点方法应为网格布
点法，采样点布设格网间隔2 km，网格布点较随机布点和联合单元布点效率分别提高64.8%和
128.8%；3）丘陵区合理采样布点方法应为联合单元布点法，样点布设密度为1个/1 314 hm 2 ，联
合单元布点较随机布点和网格布点效率分别提高205.8%和294.2%。

Sampling design is a critical issue in soil organic carbon research. In this paper, the
purpose was to find out the most appropriate sampling method among various sampling
designs in farmlands with different landform types. We used the sample representative
evaluation in classical statistics to compare the efficiency of most commonly used sampling
methods, including random distribution, grid sampling design and joint unit method. The
representative samples of soil organic carbon in farmlands with different landform types on the
Loess Plateau were evaluated. The results demonstrated that: 1) A reasonable sampling method
for Ning County in the upland area of Loess Plateau was grid sampling design with a suggested
grid interval of 4 km. The efficiency of grid sampling design was 64.3% and 31.8% higher than
the efficiency of random distribution method and joint unit method, respectively. 2) A
reasonable sampling method for Wugong County in the plain area of the Loess Plateau was also
grid sampling design with a suggested sampling interval of 2 km. The efficiency of grid
sampling method was 64.8% and 128.8% higher than the efficiency of random distribution
method and joint unit method, respectively. 3) The reasonable sampling method was joint unit
layout method in the hilly and gully region of the Loess Plateau, such as in Zhuanglang County.
The suggested density of the joint unit layout method in this area was 1 unit/1 314 hm 2 , and the
efficiency of joint unit distribution was 205.8% and 294.2% higher than the efficiency of
random distribution and grid distribution, respectively.