摘要
Non-agricultural lands are surveyed sparsely in general.Meanwhile,soils in these areas usually exhibit strong spatial variability which requires more samples for producing acceptable estimates.Capulin Volcano National Monument,as a typical sparsely-surveyed area,was chosen to assess spatial variability of a variety of soil properties,and furthermore,to investigate its implications for sampling design.One hundred and forty one composited soil samples were collected across the Monument and the surrounding areas.Soil properties including pH,organic matter content,extractable elements such as calcium (Ca),magnesium (Mg),potassium (K),sodium (Na),phosphorus (P),sulfur (S),zinc (Zn),and copper (Cu),as well as sand,silt,and clay percentages were analyzed for each sample.Semivariograms of all properties were constructed,standardized,and compared to estimate the spatial variability of the soil properties in the area.Based on the similarity among standardized semivariograms,we found that the semivariograms could be generalized for physical and chemical properties,respectively.The generalized semivariogram for physical properties had a much greater sill value (2.635) and effective range (7 500 m) than that for chemical properties.Optimal sampling density (OSD),which is derived from the generalized semivariogram and defines the relationship between sampling density and expected error percentage,was proposed to represent,interpret,and compare soil spatial variability and to provide guidance for sample scheme design.OSDs showed that chemical properties exhibit a stronger local spatial variability than soil texture parameters,implying more samples or analysis are required to achieve a similar level of precision.
Non-agricultural lands are surveyed sparsely in general. Meanwhile, soils in these areas usually exhibit strong spatial variability which requires more samples for producing acceptable estimates. Capulin Volcano National Monument, as a typical sparsely-surveyed area, was chosen to assess spatial variability of a variety of soil properties, and furthermore, to investigate its implications for sampling design. One hundred and forty one composited soil samples were collected across the Monument and the surrounding areas. Soil properties including pH, organic matter content, extractable elements such as calcium （Ca）, magnesium （Mg）, potassium （K）, sodium （Na）, phosphorus （P）, sulfur （S）, zinc （Zn）, and copper （Cu）, as well as sand, silt, and clay percentages were analyzed for each sample. Semivariograms of all properties were constructed, standardized, and compared to estimate the spatial variability of the soil properties in the area. Based on the similarity among standardized semivariograms, we found that the semivariograms could be generalized for physical and chemical properties, respectively. The generalized semivariogram for physical properties had a much greater sill value （2.635） and effective range （7500 m） than that for chemical properties. Optimal sampling density （OSD）, which is derived from the generalized semivariogram and defines the relationship between sampling density and expected error percentage, was proposed to represent, interpret, and compare soil spatial variability and to provide guidance for sample scheme design. OSDs showed that chemical properties exhibit a stronger local spatial variability than soil texture parameters, implying more samples or analysis are required to achieve a similar level of precision.
基金
Project supported by the LSU AgCenter,USA
