One assumption made in pedology is that humans have little impact on the soil over geologically short timespans. I examined catenas of north-central Iowa in order to determine whether the catena concept still applies to soils under intensive land use. The catena model is used worldwide to predict soil distribution across hillslopes and is therefore central to soil mapping. However, I found soil distribution is hillslope-by-land use dependent, especially in areas of intensive cropping, tile drainage, fertilization, tillage and erosion. Soil pH, thickness of mollic colors, organic carbon, geometric mean of particle size, and bulk density of four catenas in north-central Iowa are strongly correlated with percent distance from the summit in any given catena (r2 > 0.90 in most cases), but combining the data results in a weak soil-landscape relationship (r2 < 0.25 for most properties). Therefore, extrapolating from one hillslope to the next is problematic because differences in management history make soils more unpredictable. This indicates ongoing field work is especially important for creating precision soil maps. It further indicates those precision maps will need to rely upon a new catena model that integrates human impacts vis-Ã Â -vis pedology, hydrology, and geomorphology.

Another assumption often made in soil science is that soil particle density is 2.65 g cm-3. I tested whether this assumption is accurate in Iowa by measuring the particle density of some soils from across the state as well as from the four catenas in north-central Iowa. Soil particle density of the samples ranges from 2.10-2.84 g cm-3 with a mean of 2.66 g cm-3 and a median of 2.68 g cm-3. In the upper 10 cm, the mean particle density is 2.58 g cm-3 with a median of 2.62 g cm-3. Particle density is highly correlated with soil organic carbon content (r2 = 0.85) in these samples and increases with depth in most of the soil profiles. Across the catenas, lowland soils have lower particle density due to larger quantities of soil organic carbon and other low density components compared to their uphill counterparts. Soil particle density is used in calculations for other parameters such as porosity, thermal properties, and sedimentation, but is rarely measured. My results indicate a wide range in particle density that has ramifications for calculations and models that routinely estimate the value.

Last, I evaluated whether properties are truly homogeneous throughout a standard soil horizon by subdividing pedons into 3-cm increments. While some standard soil horizons are relatively homogeneous at the 3-cm level, at least for the properties analyzed, other horizons show considerable variability in morphology, pH, and/or total carbon (TC). The pH and TC for a given horizon can have a large range in values at the 3-cm level while the overall mean is similar to that found using a bulk standard horizon sample. Likewise, morphology can exhibit variability within standard horizons. The greatest variability in soil properties occurs either near the soil surface or near the weathering front in the lower B horizon and upper C horizon. Therefore, the method of sampling, whether it be by standard horizon or specific depth increment, can lead to different results for a given soil property. Standard horizons can be thought of as averaging within-horizon variability, which could hamper our understanding of soil formation and change. The challenge is what to do as an alternative. The answer may be different depending on which soil property is being evaluated.