Anion transport in a Piedmont ultisol .2. Local-scale parameters

Field-scale transport parameters obtained from tile drain data in two large conventionally tilled plots were reported in the first paper in this series. Our objectives in this paper were to determine the relationship between local-scale transport parameters and macropore distributions measured with columns in one of these plots and to compare local-scale transport parameters with the field-scale parameters. Forty undisturbed soil columns (15 cm in diameter, 33-64 cm in length) were taken from Plot 1 described in the first paper. Transport parameters were determined from a Br breakthrough curve (BTC) with CXTFIT. Macropore distributions were determined from computed tomography (CT) scanning and by sectioning the columns after adding methylene blue dye. Local-scale saturated hydraulic conductivity (K-s), pore water velocity (V), and dispersivity (lambda) were highly variable (coefficient of variation of 77, 74, and 41%, respectively) and lognormally distributed. The K-s was positively correlated with the number of 5- to 10-mm-diameter CT scan macropores at a depth of 35 to 40 cm near the top of the Bt horizon. Dye patterns indicated that preferential how began at about this depth. This may be the zone where flow becomes restricted and highly dependent on macropores. Mean local-scale lambda (6.6 cm) was similar to the estimated field-scale values (3.4 and 5.3 cm) for the unsaturated zone to a depth of 20 cm. This implied that variation among local-scale V in the field experiment was much less than that observed in the columns, probably because the surface was not ponded in the field. A deterministic approach with mean values of local-scale parameters adequately described the estimated field-scale BTC.