Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Agricultural and Biological Engineering

First Advisor

Laura C. Bowling

Committee Chair

Laura C. Bowling

Committee Member 1

Jane R. Frankenberger

Committee Member 2

Indrajeet Chaubey

Committee Member 3

Reuben R. Goforth


A two-stage ditch involves modifications of a traditional drainage ditch to resemble more the features of a natural stream. The idea is to create or simulate extended benches on both sides of the ditch that would develop naturally over a period of time in a stream because of geomorphological processes. Previous research in Indiana and Ohio has shown that two-stage ditches offer the potential to reduce sediment load and extend the interaction time between water and vegetation on the benches allowing larger uptake of nutrients from the vegetation on the extended benches, and increasing the denitrification rates in the bench soil during high flow events. Purdue University constructed a two-stage ditch on September 26, 2012 at Throckmorton Purdue Agriculture Center (TPAC), 16 km south of the city of Lafayette, Indiana. It drains an area of approximately 2.7 km2 of farmland used for corn and soybean production. The study focused primarily on four main objectives: 1) quantify differences in observed concentrations/loads of total suspended sediment (TSS) and nutrients (nitrate-N, soluble reactive phosphorus (SRP) and total phosphorus (TP)) between two equal length reaches of a traditional agricultural ditch of which one was converted into a two-stage channel with the traditional reach acting as the control reach and the two-stage reach as the treatment reach, 2) monitor the performance of five different mixes of sedges, forbs and grasses with regard to nutrient uptake, resistance to invasive species and establishment, in order to determine if there is a certain mix of plants that performs better on the benches of two-stage ditches and that can be recommended for use in other two-stage ditches, 3) represent the two-stage ditch in a hydrologic model as a conservation practice, comparing the model outputs with observed water quality from the two-stage ditch that was part of this study, and 4) simulate the dominant mechanisms controlling nutrient retention in the two-stage ditch at the watershed scale using the modified hydrologic model.

To achieve the first objective of quantifying the impact of the two-stage ditch on sediment and nutrients, the ditch was divided into a 185 m upstream control reach, and a 200 m downstream treatment reach and flow and water quality data upstream and downstream of both the control reach and the treatment reach were collected. Analysis using paired t-tests and ANCOVA showed that the two-stage ditch reduced TSS loads and concentrations by 22% and 50%, respectively, TP loads and concentrations by 40 and 50% respectively, compared to increases that occurred in the control reach. SRP loads were reduced by 11% in the two-stage ditch compared to a 2% increase in the control reach. There were no statistically significant reductions in Nitrate-N loads, but concentrations decreased by x% in the treatment reach. The five vegetation mixes that were tested and monitored on the benches of the two-stage ditch were separated in twenty different plots to account for location differences. For the second objective, all the plots showed good vegetation establishment, with only the buffer strip mix plots (that included grasses mainly designed for use in upland buffer strips) showing a statistically significant lower vegetation coverage when compared to the other four mixes. The buffer strip mix plots were also outcompeted by more aggressive plants in the first year of growth. Overall biomass harvested from all the plots ranged between 7 – 17 tons/ha. No statistically significant difference in biomass was found among the different plant mixes. The biomass samples were analyzed for nitrogen and phosphorus content and the results showed that the nutrient content of the biomass collected did not differ between the plots. Total nitrogen and phosphorus content for each plot depended more on the total biomass weight and less on the nitrogen or phosphorus concentrations in a particular plant. The Soil and Water Assessment Tool (SWAT) was the model chosen to represent the two-stage ditch. The model’s source code was modified to account for the geometry of the two-stage ditch and the changes in nutrient transport that result from the modification of the stream channel. Model results supported the findings of the field study, showing reductions of the same scale in suspended sediments and total phosphorus and no reductions in nitrate loads when a reach was converted into a two-stage channel. The reductions in sediment and phosphorus were mainly driven by sedimentation on the benches of the two-staged ditch which accounted for 95% of reductions in total phosphorus. Mean denitrification on the benches was 0.1 kg/ha/day and 0.62 kg/ha/day when benches were flooded. Denitrification’s contribution to load reduction is dependent on both water depth and temperature and confined to a small time window (April – June). Mean annual plant uptake of nitrogen and phosphorus was 47 and 11 kg/ha, respectively. Arguably, the two-stage ditch could be one of the best management practices in reducing/retaining sediment and phosphorus loads going downstream.