| Fractal descriptions of soil aggregation are commonly accomplished through excavation and sieving processes. To achieve an in situ description of soil aggregation, we applied fractal theory to penetration resistance measurements with depth in undisturbed soil cores (Dpr). Sieved soil aggregates were also used to estimate a number-based aggregate fractal dimension (Did). The results showed no fractal dimension difference between no-tillage (NT) and moldboard plowed (MP) treatments regardless of the measurement method used. There were, however, significant differences in fractal dimension values between methods, where D id was greater than Dpr. The difference between fractal dimensions may be due to theoretical differences of the estimation techniques and an increased fragmentation from the sieving process. In a second experiment, completely fragmented fractal dimensions were obtained from the analysis of sieved soil by number-size (Dc) and mass-size distribution, assuming scale variant (Dcf) and invariant density of aggregates (Did). Incompletely fragmented fractal dimension was obtained from penetration resistance of a soil core and from the bulk fractal dimension (Dc) of the sieved soil. The estimation of Dc and Dcf produced very similar results, showing significant difference between soils (Rayne and Wooster series) and tillage (NT and MP) treatments. The assumption of scale invariant density of aggregates seems to decrease the sensitivity of Did to describe tillage induced differences in aggregation. The description of incompletely fragmented soil by both methods, Dpr and Dr, did not show differences between soils and tillage treatments.; Estimation of unsaturated hydraulic conductivity near saturation, using Dcf and Dr as inputs was not successful, possibly due to unreasonable assumption of the model concerning connectivity of the pore system, and a lack of fractality of the soils in the scale range of the measurements. The fractal dimension of aggregate distribution has, apparently, a limited application in modeling soil hydraulic processes. The successful description of the aggregate distribution by a power law ( Dc, Dcf and Did) using sieved soil, does not necessarily indicate the presence of aggregate and pore size self-similarity of the Wooster and Rayne soils in the size range studied. |
