Soil Trends and Variability Across Selected Landscapes in Iowa

Abstract

AbstractSoil profile features such as A horizon thickness, depth to subsoil mottles, depth to carbonates, and surface and subsoil pH were measured in sample grids on small loess and drift landscape segments. Trend analyses were made across the full range of topography, then soil variation was examined within topographic subareas that approximated detailed soil survey delineations. Comparison of standard errors and standard deviations showed that some increase in the precision of predictions was obtained for loess soils when using “trend” rather than “within delineation” analysis. The reverse was true for drift soils. The representation of data in polynomial trend equations is justified because of systematic variation observed across small landscape segments, and from a consideration of a soil landscape model that is generally applicable in Iowa. Using this model, maxima and minima of soil properties are expected along transects from upland to lower pediment slopes. Polynomials provide a statistical fit of soil landscape data and a suitable representation of maxima and minima of soil properties.