Big Spring Run is a 1.68-square mile watershed underlain by mostly carbonate rock in a mixed land-use setting (part agricultural and part developed) in Lancaster County, Pennsylvania. Big Spring Run is a subwatershed of Mill Creek, a tributary to the Conestoga River. These watersheds are known contributors of nutrient and sediment loads to the Chesapeake Bay and several stream reaches are on the Pennsylvania impaired waters list. Big Spring Run is listed as impaired and was selected by the Pennsylvania Department of Environmental Protection to evaluate a novel best management practice to restore natural aquatic ecosystems by removing legacy sediment. The study was designed to quantify sediment and nutrient contributions in pre- and postrestoration periods (water years 2009–11 and 2012–15, respectively) using an intensive monitoring approach at three surface-water sites within the watershed. Instrumentation at each site continuously measured (15-minute intervals) streamflow, water temperature, and turbidity. Water-quality samples were collected routinely (generally monthly and during selected storms); sampling frequency varied by site and constituent at the three monitoring sites.

Effects of legacy sediment removal and restoration on nutrient concentrations varied in surface water samples depending on the form (particulate, dissolved, organic, inorganic). For example, total phosphorus concentrations at the downstream site decreased from a median of 0.19 milligram per liter (mg/L) to 0.04 mg/L, pre- and postrestoration periods, respectively. Concentrations of orthophosphate, the dissolved form of phosphorus, were not significantly different pre- to postrestoration at the downstream site. Similarly, nitrate concentrations, the dominant form of nitrogen in Big Spring Run surface-water samples (92.3 percent of total nitrogen) were not significantly different in the pre- compared to the postrestoration periods.

Legacy sediment removal and restoration had significant effects on suspended-sediment concentrations and loads. Median suspended-sediment concentrations at the downstream site decreased from 556 mg/L prerestoration to 74 mg/L postrestoration even though streamflow hydrographs during the two periods were similar. In the postrestoration period, the mean annual suspended-sediment load conveyed to the restoration area from the upstream sites was 839 tons, whereas mean annual suspended-sediment load at the downstream site was reduced to 242 tons.

Streamflow during storms transports a large proportion of the suspended-sediment load; there were a total of 320 storms over the study period. In Big Spring Run, a single storm event can transport more than 25 percent of the annual suspended-sediment load. The greatest single-storm contribution to suspended-sediment load was 38 percent in water year 2015 at the downstream site. Although streamflow magnitudes during storms varied greatly over the study period, median streamflow was 17.5 cubic feet per second and median duration was about 3 hours and 24 minutes.

Results observed for this study using the newly proposed best management practice were compared with other best management practices intended to reduce sediment. For example, during a previous study, statistically significant reductions in suspended-sediment concentration were observed when streambank fencing was implemented in an adjacent watershed; however, suspended-sediment reductions were an order of magnitude less than the reductions observed in the current study. Median suspended-sediment concentration at the downstream site was reduced by 482 mg/L in the current study compared to only 30 to 46 mg/L as a result of streambank fencing.