• hydrology;
  • biogeochemistry;
  • agricultural watersheds;
  • spectral analysis;
  • wavelet analysis;
  • spatio-temporal scaling

[1] Conceptualizing catchments as physicochemical filters is an appealing way to link streamflow discharge and concentration time series to hydrological and biogeochemical processing in hillslopes and drainage networks. Making these links explicit is challenging in complex watersheds but may be possible in highly modified catchments where hydrological and biogeochemical processes are simplified. Linking hydrological and biogeochemical filtering in highly modified watersheds is appealing from a water quality perspective in order to identify the major controls on chemical export at different spatial and temporal scales. This study investigates filtering using a 10 year data set of hydrological and biogeochemical export from a small (<500 km2) agricultural watershed in Illinois, the Little Vermilion River (LVR) Watershed. A number of distinct scaling regimes were identified in the Fourier power spectrum of discharge and nitrate, phosphate, and atrazine concentrations. These scaling regimes were related to different runoff pathways and spatial scales throughout the catchment (surface drainage, tile drains, and channel flow in the river). Wavelet analysis indicated increased coupling between discharge and in-stream concentrations at seasonal-annual time scales. Using a multiresolution analysis, nitrate, phosphate, and atrazine loads exported at annual scales were found to exhibit near-linear scaling with annual streamflow, suggesting that at these scales the export dynamics could be approximated as chemostatic responses. This behavior was pronounced for nitrate and less so for phosphate and atrazine. The analysis suggests that biogeochemical inputs built up legacy loads, leading to the emergence of chemostatic behavior at annual time scales, even at the relatively small scale of the LVR.