Influence of sampling strategy on detecting preferential flow paths in water-repellent sand

Rapid response tensiometers or TDR (time domain reflectometry) probes in soil profiles, or early arrival of solutes in groundwater or drainwater after a rain event, provide direct evidence of preferential flow in soils. However, little information about the amount and magnitude of preferential flow paths is obtained by such measurement methods. Here, two intensively sampled vertical trenches illustrate the effect of sample spacing and the effect of sample size on the detection of preferential flow paths. In a water-repellent sandy soil, a sample spacing of up to 22 cm over a distance of several metres is just sufficient to collect information about preferential flow paths. Using larger sample spacings, the water content distributions apparently became more horizontally stratified. Increasing the sample size by pooling pairs of adjacent 100 cm 3 soil samples over a distance of several metres, still allowed the detection of preferential flow paths. Preferential flow paths were no longer observed for larger sample sizes. Enlarging the sample size reduces the calculated standard deviation and coefficient of variation. As preferential flow paths may vary in space and time, so the optimal number of samples to detect these paths in vertical trenches may vary, indicating that sampling strategies need to be flexible in design.