Spatial variation of urease activity measured in soil monitoring

Since time and money are usually limited, researchers need to optimize and select sampling scales that reflect the spatial variation of the properties under consideration. This paper addresses the question how sampling designs can be evaluated with respect to the selection of sampling scales when monitoring soil enzyme activity, The proposed methodology is illustrated by studying the spatial variation of urease activity and organic C content at three sites that have different types of land use (pasture, arable land, and forest) as an example. At each site, an area of 0.75 hectares was sampled using a hierarchical multistage sampling scheme called nested sampling. Large differences in both the statistical and spatial distributions were observed between the three sites. For the arable land, a considerable part of the total variance of the two variables, urease activity and organic C content, could be statistically explained by stratifying the samples according to soil color, thereby reflecting differences in the origin of the organic matter. No correlation between the two variables was found within the forest and the pasture site, as well as within each of the two strata distinguished by soil color on the arable land site. Spatial autocorrelation of urease activity was found only for sample spacings of < 1 m for the pasture site, while autocorrelation extended up to 15 m for the other two sites. To represent the full site-specific range of spatial variation, the sample spacings must encompass these distances. Because of its efficiency in identifying spatial scales of variation, nested sampling is especially well suited for application to pilot surveys by providing a basis for the design of more intensive sampling campaigns, including long-term soil monitoring programs.