Quantitative relationships between soil macropore characteristics and preferential flow and transport

Authors: LUO, LIFANG - University Of California; LIN, HENRY - Pennsylvania State University; Schmidt, John

Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/25/2010
Publication Date: 11/15/2010
Citation: Luo, L., Lin, H., Schmidt, J.P. 2010. Quantitative relationships between soil macropore characteristics and preferential flow and transport. Geoderma. 74:1929-1937.

Interpretive Summary: Macropore networks contribute to preferential flow in soils that is not always consistent with our expectations about different land uses. Our objective was to compare the macropore characteristics between cropland and pasture and to relate these characteristics to common measurements of hydraulic conductivity. Macropore parameters for multiple soil columns were obtained with X-ray computed tomography and compared to measurements of hydraulic conductivity and solute transport. While the pasture soil had greater pore space and more macropores than the cropland soil, the network of macropores was more complex and less contiguous than observed for cropland, so water transport through the soil profile was greater in the cropland soil. Hydraulic conductivity measurements from selected soil horizons were not consistent with whole soil profile measurements because horizon samples tended to disrupt macropore channels that were responsible for most of the water transport in soils.

Technical Abstract: Quantitative relationships between soil structure, especially macropore characteristics, and soil hydraulic properties are essential to improving our ability to predict flow and transport in structured soils. The objectives of this study were to quantitatively relate macropore characteristics to saturated hydraulic conductivities (Ksat) and solute dispersion coefficients and to identify major macropore characteristics for estimating soil hydraulic properties. Intact soil columns (102 mm in diameter and about 350 mm in length) were taken from two land uses (cropland and pasture) of a typical agricultural soil of central Pennsylvania. The soil columns were scanned using X-ray computed tomography to obtain macropore parameters including macroporosity, length density, mean tortuosity, network density, hydraulic radius, path number, node density, and mean angle. The Ksat of the whole soil column as well as each soil horizon within the column were measured and the breakthrough curve (BTC) of calcium bromide was determined for each column. Both the traditional convection-diffusion equation (equilibrium model) and the two-region model (non-equilibrium model) were used to fit the BTCs and inversely estimate the solute transport parameters. For all soil columns studied, macroporosity and path number explained 71-75% of the variability in the Ksats of the whole soil columns as well as individual soil horizons. We found that the Ksat of individual soil horizons measured from the soil columns were 3 to 46 times greater than those measured from standard small soil cores (55 mm in diameter and 60 mm in length). The equilibrium model simulated BTCs well for all soil columns except for one with an earthworm hole passing through the entire column, for which the non-equilibrium model was required. The cropland soil showed a higher degree of preferential flow compared to its pasture counterpart because soil structure in the pasture soil was better developed and macropores were more abundant and evenly distributed in the soils. The path number, hydraulic radius, and macropore angle were the best predictors for the solute dispersion coefficient, explaining 97% of its variability. Good correlation between the solute dispersion coefficients of the whole soil columns and the Ksats of the Bt horizons (but not other soil horizons) implied that the dispersion coefficient was mainly controlled by the horizon with the lowest conductivity in a soil column.