Assessing herbicide movement using soil samples versus percolate samples

Soil core samples are often used to assess herbicide movement through soil or to evaluate contaminant transport models. When macropore flow occurs, th e suitability of soil cores for these purposes is questionable. Our objecti ves were to evaluate the effectiveness of soil samples from a macroporous s oil as the sole means to assess herbicide movement in percolate and to asse ss a contaminant transport model. To accomplish this, atrazine and alachlor were surface-applied to 30 x 30 x 30 cm blocks of undisturbed, no-till sil t loam soil at three moisture levels (dry, intermediate, wet) then subjecte d to a 0.5-h, 30 mm simulated rain. Percolate was collected from the base o f the blocks and the soil was sampled by slicing the blocks into eight, 3.7 5 cm-thick, horizontal slabs. The contaminant transport model GLEAMS was us ed to predict herbicide concentration in the percolate and soil. The model was calibrated to equate the observed and predict percolate volume by adjus ting the initial water content. Further model calibration was performed for two modeling scenarios: (I) to equate the observed and predicted herbicide concentration in the bottommost soil layer (26.25-30 cm) or (2) to equate the observed and predicted herbicide concentration in the surface soil laye r (0-3.75 cm). No correlation was observed between herbicide concentration in soil and herbicide concentration in percolate. GLEAMS was calibrated to accurately predict percolate volumes and herbicide concentration in soil, b ut herbicide concentration in percolate was substantially under-predicted i n most instances by a factor of 6 to 123. Our results indicate that the soi l sampling strategy used in this study was a poor indicator of subsurface h erbicide movement in percolate and ineffective as the sole means to assess contaminant transport models in soils subject to macropore flow.