The latitude and longitude of the 1918 water depth readings were used to extract the 2010 water depth that corresponded to the same location. The difference in water depth between 1918 and 2010 is a measure of the sedimentation that occurred at that cross section during the intervening 92 years. The thickness of sediment between the radiometrically dated 1918 core horizon and the basal fluvial sediment provides an estimate of sedimentation from the time of dam construction to 1918. The volume of impoundment sediment was

calculated in segments centered on each 1918 cross section. First, the area of impoundment sediment on each cross section was multiplied by the longitudinal distance between cross sections. Second, all the segment volumes were summed. The Middle Cuyahoga River Watershed Action Plan (Peck, 2012) estimates average annual sediment load for the watershed using the US EPA Spreadsheet Technique for Estimating Pollutant Rigosertib Loading (STEPL)

model (US EPA, 2010). STEPL is one of several models widely recommended by state agencies to estimate sediment loading at a watershed scale, primarily Tanespimycin cell line to compare average loadings before and after changes to land use, best management practices, and restoration projects. The model is relatively simple, based in Excel spreadsheets. Inputs to the model are readily available land cover data and embedded county default data for soil characteristics and average precipitation. STEPL does not incorporate detailed watershed flow modeling or routing to estimate sediment load. STEPL combines two widely used methods: the Revised Uniform Soil Loss Equation (RUSLE) is used to estimate Epothilone B (EPO906, Patupilone) sediment from agricultural land, annual rainfall runoff times pollutant concentration are used to estimate pollutant loading from developed land. For the Middle Cuyahoga River Watershed, 30 m × 30 m resolution land cover data from 2006 (NOAA, 2006) was clipped to the watershed (Fig. 3). Portage County, Ohio, soil and rainfall characteristics were used because most of the watershed lies in Portage County. The

model also allows modifications to be made to better reflect the landscape characteristics, including known agricultural practices (75% of the area uses reduced tillage), livestock, use of best management practices, and severe streambank erosion. Based upon field inspections throughout the watershed and review of 2006 aerial photography, it is estimated that approximately 24 km of the 188 km of stream channel in the watershed is severely eroding. The STEPL model indicated that 7490 tonnes per year was the average annual sediment loading under 2006 land cover and stream erosion conditions. We converted the model output from tons to tonnes (i.e., metric tons). Primary sediment sources in the south were agricultural land and streambank erosion (Fig. 3).