Stream restoration and sewers impact sources and fluxes of water, carbon, and nutrients in urban watersheds

Abstract. An improved understanding of sources and timing of water, carbon, and nutrient fluxes associated with urban infrastructure and stream restoration is critical for guiding effective watershed management globally. We investigated how sources, fluxes, and flowpaths of water, carbon (C), nitrogen (N), and phosphorus (P) shift in response to differences in urban stream restoration and sewer infrastructure. We compared an urban restored stream with two urban degraded streams draining varying levels of urban development and one stream with upland stormwater management systems over a 3-year period. We found that there was significantly decreased peak discharge in response to precipitation events following stream restoration. Similarly, we found that the restored stream showed significantly lower (p < 0.05) monthly peak runoff (9.4 ± 1.0 mm day⁻¹) compared with two urban degraded streams (ranging from 44.9 ± 4.5 to 55.4 ± 5.8 mm day⁻¹) draining higher impervious surface cover, and the stream-draining stormwater management systems and less impervious surface cover in its watershed (13.2 ± 1.9 mm day⁻¹). The restored stream exported most carbon, nitrogen, and phosphorus at relatively lower streamflow than the two more urban catchments, which exported most carbon and nutrients at higher streamflow. Annual exports of total carbon (6.6 ± 0.5 kg ha⁻¹ yr⁻¹), total nitrogen (4.5 ± 0.3 kg ha⁻¹ yr⁻¹), and total phosphorus (161 ± 15 kg ha⁻¹ yr⁻¹) were significantly lower in the restored stream compared to both urban degraded streams (p < 0.05), but statistically similar to the stream draining stormwater management systems, for N exports. However, nitrate isotope data suggested that 55 ± 1 % of the nitrate in the urban restored stream was derived from leaky sanitary sewers (during baseflow), statistically similar to the urban degraded streams. These isotopic results as well as additional tracers, including fluoride (added to drinking water) and iodide (contained in dietary salt), suggested that groundwater contamination was a major source of urban nutrient fluxes, which has been less considered compared to upland sources. Overall, leaking sewer pipes are a problem globally and our results suggest that combining stream restoration with restoration of aging sewer pipes can be critical to more effectively minimizing urban nonpoint nutrient sources. The sources, fluxes, and flowpaths of groundwater should be prioritized in management efforts to improve stream restoration by locating hydrologic hot spots where stream restoration is most likely to succeed.