Laboratory Methods for Soil Health Analysis, Volume 2

Collection of metadata (Vol. 1, Chapter 4) is an important part of site characterization and will facilitate interpretation of measurements. At a minimum, geographical, landscape, soil, vegetation, climatic, and management attributes should be recorded and maintained for future reference ( Table 2.2). When sampling as part of a long‐term study, more metadata is required to accurately characterize each sampling site (see Table 1.2in Boone et al., 1999).

Table 2.2 Suggested minimum metadata for site characterization.

Following preliminary site characterization, the evaluator must decide how to sample the site. The design used for sampling should reflect the goals of the evaluator, recognizing and reconciling tradeoffs associated with different designs ( Table 2.3). Typically, available time and resources dictate the selection of a sampling design.

Judgement sampling selects sites presumed to represent larger areas ( Table 2.3). Since site selection is based on the evaluator’s knowledge and skill, it is subject to considerable bias. As outcomes are inherent to the evaluator, application of this design for long‐term use is limited as changes in evaluator judgement are possible over time. Data collected using judgement sampling has limited statistical validity especially when isolated from other sites sampled by the same evaluator.

Despite its drawbacks, judgement sampling can be useful when resources are scarce and demands for accuracy and statistical rigor are relaxed. This design may also be applicable when samples are difficult to collect, or an evaluation involves attributes with high analytical cost.

Table 2.3 Definitions and attributes of different soil sampling designs.

The simple random sampling design assumes each given area has an equal opportunity of being selected ( Table 2.3), and this typically involves pre‐selection of sampling points to avoid evaluator bias (Dick et al., 1996). Point selection may be done using a random number table and a coordinate grid overlain on an image of the sampling area, or with a random point generator and GIS (geographic information systems) software package (e.g., ArcGIS). Use of this sampling design minimizes bias associated with point selection.

Though easy to apply, simple random sampling has notable caveats since all points in an area are assumed to have equal importance, regardless of variation in inherent features and/or spatial stratification of management attributes. Spatial distribution of selected points therefore may not adequately represent an entire site, although this concern diminishes as the number of sample points increases. This design may be appropriate when knowledge of the site is limited.

Stratified random sampling applies simple random sampling to preselected subareas of a site ( Table 2.3). This design requires knowledge of site characteristics to accurately identify the location, extent, and boundary of subareas. This sampling design is most useful at sites with clearly defined soil, landscape, management, and/or contamination zones (Dick et al., 1996).

Assessing soil health indicators within defined subareas will provide a more complete understanding of inherent and/or management‐induced variation. The drawback of this sampling design is a corollary to its strength, since subarea stratification will require more resources than simple random sampling.