Quantification of soil spatial variability across multiple scales is important in ecological modeling, environmental prediction, precision agriculture, and natural resources management. This study investigates the variability of soil map units and soil properties at multiple scales using two case studies, and demonstrates that soil spatial variability is a function of map scale, spatial location, and specific soil property. In the first case study, the variability of soil components within a map unit, termed map unit purity, was examined at three orders of soil surveys in the Backswamp Watershed in South Carolina. Soil maps of Order I (1:7920), Order II (1:24,000) and Order IV (1:250,000) were independently obtained and examined in a Geographic Information System (GIS). The results showed that the area-weighted mean purity Pm for the Order II soil map when compared to the Order I delineations was 51-99% for soil taxonomic units (soil series to order) and 65-85% for soil properties important for land management in the area (texture, structure, surface thickness, hydrologic group, and drainage class). Corresponding values of Pm for the Order IV map were 24-81% and 60-90% when compared to the Order II delineations. In the second case study, 324 soil samples were collected across the Minnesota River Basin using a nested hierarchical sampling design that allowed for variability assessment at multiple scales through a hierarchical analysis of variance. A-horizon thickness, depth to calcium carbonate, and surface soil pH values were summarized by soil regions, hillslope positions, clusters, and point scales. The majority of the variability (over 50% in most cases) for all three soil properties was at the local point scale, suggesting that careful examination of short-range soil property variability should not be overlooked. Possible causes of variability ranged from climate at the basin scale to localized effects of differential infiltration and runoff caused by the differences in landscape positions and soil characteristics. We recommend a hierarchical sampling approach similar to the one used in this study to inventory soil spatial variability at multiple scales so that an understanding could be developed not only for soil property variability across the landscape, but also for determining at what scale the variability is most likely to occur.
All Science Journal Classification (ASJC) codes
- Ecological Modeling