Whole Hillslope Irrigation Reveals Differential Interflow Behavior of Dye Tracers, Conservative Solutes and Nutrients

Previous investigations of perching and interflow behavior in low angle hillslopes in the SC Coastal Plain have suggested a high threshold for interflow occurrence. Here we report a new irrigation experiment designed to quantify interflow thresholds and reveal subsurface mixing processes during steady state flow conditions over a 12m x 16.5m plot draining to an interflow interception trench. Dye tracers were applied on surface transects prior to irrigation, and bromide (conservative tracer), nitrate, ammonium, and phosphorus (reactive tracers) were added at constant concentrations to the irrigation water drawn from a deep aquifer with a distinct isotopic signature. 417mm of water were applied over 51 hours, and drainage conditions were monitored for a week following irrigation. Interflow in the two drains commenced after 131 and 178mm, and flow rates diminished immediately after irrigation ceased, although interflow continued for four more days. Over the experiment, 199mm of water (49% of applied water) appeared as interflow. Dye tracers moved rapidly with the wetting front, with peak concentrations measured shortly after flow commencement, suggesting saturated topsoil conductivities of 0.5 to 1.5 m/hr. No preferential flow was observed during this experiment or previously during rainfall events at the trench face. Bromide concentrations and the new water fraction rose steadily throughout irrigation, peaking about 16 hours after irrigation ceased. Ammonium and phosphorus concentrations at the trench face were low, suggesting rapid uptake or sorption, while nitrate concentrations were higher, suggesting more conservative transport. Our two collection drains showed identical temporal variation in bromide concentrations but consistently different new/old water fractions, indicating differences in flow paths and storages within the plot. These data suggest that tightly bound soil water exchanged with new water throughout the experiment, and that a significant portion of flowing old water must have originated from the argillic layer beneath the sandy loam topsoil. Inferences from the isotopic signature of the outflow water were consistent with the bromide data. Piezometric and outflow data all point to fill and spill effects on interflow generation in this low relief site.