Step 1: Identify TMDL Project Area HUC10 Watersheds

The IEC model is designed to model a single watershed with up to 10 subwatersheds. For the TMDL progress tracking project HUC 8 watersheds are used for the large watershed and HUC 10 watersheds are used for the subwatersheds. GIS was used to identify the HUC 8 watersheds and HUC 10 watersheds to create IEC models for. There are 135 HUC 10 watersheds with TMDLs contained in this analysis which combined to 28 IEC models covering 26 HUC 8 watersheds.

Step 2: Build IEC models

The IEC model requires three inputs to calculate the base load of E. coli in each watershed: land use by subwatershed, agricultural animals per subwatershed, and septic systems per watershed. It also requires the BMP type, BMP landuse type, contributing acres, and subwatershed to calculate the reduction.

The first required input for the IEC are land use areas broken down into four categories: cropland, pasture, forest, and urban. These four categories are simplified from the 15 land use “classes” covered by the 2011 National Land Cover Dataset (NLCD). The following map shows the final delineation used to create the IEC land use acreages.

The 2017 USDA Indiana Agriculture Census was used to estimate the number of agricultural animals per watershed. The information was downloaded as a text file and analyzed to output the number of animals by type (cattle, hogs, dairy cows, poultry, horses, and goats) contained within the IEC.

The final input required by the IEC is the estimated number of septic systems within the watershed. The most recent accurate count from the US census was in 1990 and the results were taken from this webpage. The number of septic systems by county was translated to number of septic systems per HUC 8 watershed and input into the IEC.

Each IEC model had the number of septics, number of agricultural animals, and the county entered to calculate the base loading for the watershed. An example IEC input sheet is included below for HUC 4040001.

Step 3: Enter Indiana Conservation Partnership (ICP) BMP Data

The ICP collects the number and type of BMPs implemented by partners across the state of Indiana each year. Partners include the Indiana State Department of Agriculture (ISDA) and their various subsidiaries and IDEM. This dataset is the most comprehensive list of conservation BMPs implemented. IDEM requested and received data from 2013 through 2020 for this analysis.

The ICP data also needed to be formatted to fit the BMP classifications of the IEC. The following table shows equivalencies between IEC BMP classifications and ICP BMPs.

Once the BMPs were entered into the spreadsheets and the load reductions calculated the results were summed and divided out by HUC 10 and HUC 8 watersheds. The table below shows results for the HUC 8 watershed classification.

Step 4: Analysis

The largest E. coli reductions are found in the watersheds that flow to the Ohio River. The next largest reductions are seen in the watersheds that flow to the Upper Mississippi River, and the fewest reductions were seen in the Great Lakes drainages.

A further breakdown of the drainages show that the Wabash watersheds had the largest number of E. coli reductions followed by Patoka-White and the Great Miami. The smallest reductions were seen in the Lower Ohio watersheds.

The amount of E. coli reduced rests heavily on the acreage of BMPs installed in the watershed. The Wabash had 195,809.97 acres of BMPs installed between 2013-2020, almost quadruple the next closest watershed, Patoka-White, which had 53,118.77 acres installed over the same time period.

The amount of E. coli reductions has been increasing steadily since 2013 with the highest reductions cells occuring in 2020.

As an extension of the TDML progress tracking analysis, E. coli reductions were calculated using the same methodology for the rest of the watersheds in the state.

Indiana Statewide E. Coli Reductions

The below figure shows the breakdown of acreages of the land uses from BMP installations. Cropland constitutes the largest proportion and urban the lowest. The vast majority of BMPs are installed in cropland. However, pasture related BMPs have an outsized number of reductions. Urban BMPs have the smallest acreage and lowest impact. There are opportunities to increase implementation of urban and pasture BMPs to increase E. coli load reductions.

IDEM analyzed the quantities of E. coli load reductions in TMDL watersheds, non-TMDL watersheds, and watersheds with WMPs to determine progress in implementing bacteria practices. Watersheds with both a WMP and a TMDL have consistently higher load reductions than watersheds without, and reductions from watersheds with only a WMP have risen significantly since 2017. Additionally, there is a significant difference in load reductions for watersheds without a WMP but have a TMDL and watersheds with both, indicating that WMPs and implementation projects play a significant role in increasing load reductions. See the figures below for visualizations.


Implementation in TMDL-only watersheds has increased steadily from 2013 but these reductions are significantly smaller than reductions from watersheds with a WMP only or both a TMDL/WMP combination. Watersheds with a WMP only experienced a significant increase in reductions between 2017 and 2018 and have continued to increase.


TMDLs in Indiana have recommended percent reductions of E. coli. To facilitate comparisons, estimated percent load reductions were calculated for the eight-digit watersheds in Indiana. The largest percent reductions are seen in the Blue-Sinking watershed and the Middle Wabash-Busseron watershed with 12.75% and 10.33% respectively. The South Fork Blue River TMDL calls for 73-97% load reductions during high flow scenarios and the Wabash River TMDL calls for 88% reductions in the Middle Wabash-Busseron watershed.