Effects of urban stormwater infrastructure on dissolved nutrient export from semi-arid, urban watersheds

The effects of urbanization on downstream ecosystems, particularly due to changes in nutrient inputs and altered hydrology are well studied. Less is known, however, about nutrient transport and processing within urban watersheds. Previous research has focused on the roles of land cover and land use but drainage system design and configuration also are apt to play a significant role in controlling the transport of water and nutrients downstream. Furthermore, variability in drainage systems within and between cities may lead to differences in the effects of urbanization on downstream ecosystems over time and space. We established a nested stormwater sampling network with 12 watersheds ranging in size from 5 to 17,000 ha in the Indian Bend Wash watershed in Scottsdale, AZ. Small (< 200ha) watersheds had uniform land cover (medium-density residential) but were drained by a variety of stormwater infrastructure including surface runoff, pipes, natural or engineered washes, and retention basins. We quantified discharge and precipitation at the outflow of each subwatershed and collected stormwater and rainfall samples for analyses of dissolved nitrogen (TDN) and phosphorus (PO₄), and organic carbon (oC) over two years.

On seasonal time scales, wash drained watersheds exported 80% less and retention basin drained watersheds exported 90% less TDN and PO₄ per area than piped watersheds. Total seasonal nutrient loads were correlated with total discharge across sites. Within sites, runoff and nutrient loads were predicted by storm size. Runoff coefficients from piped watersheds did not change with storm size, whereas runoff coefficients for wash and retention basin sites increased linearly with storm size. Event-based nutrient loads from the wash and retention basin sites were nonlinearly related to storm size, and for exceptionally large events, nutrient loads from the wash drained sites were as high as event loads from the piped watersheds. Due to less frequent flushing, nutrient concentrations were higher from retention basin and wash than piped sites. We compared the cumulative nutrient load with cumulative discharge to evaluate nutrient flushing. Flushing was evident for some storms at all sites, although it was more common in piped and smaller watersheds. Within the large, wash-drained watershed, flushing was predicted strongly by rainfall intensity. Overall, we find that urban drainage infrastructure is characterized by a range of hydrologic connectivity. Infrastructure significantly alters hydrologic fluxes which are the main driver of nutrient fluxes.