On-Farm Evaluation of Treatment Methods for Excess Nutrients in Agricultural Subsurface Tile Drainage


Agricultural subsurface tile drainage contributes nutrients to Minnesota waters. Constructed and/or restored wetland systems have proven successful in treating excess sediment and nutrients, especially nitrates, under a variety of conditions and have the potential to improve water quality in Minnesota’s agricultural watersheds. Restored and reconstructed wetlands in areas with high phosphorus content in the soils tend to release phosphorus (stored as residual P in the soils) to water in wetlands that may be discharged into streams. Perennial crops that are adapted to wet soil conditions, or along the fringe of wetlands, have the capacity to remove excess phosphorus from wetlands and thus provide an opportunity for addressing phosphorus in drainage water and wetlands used to treat those waters.


In Minnesota, most subsurface drainage water flows directly into streams without any treatment. This subsurface drainage water carries nutrients and contributes to overall streamflow volume, and in some areas can adversely impact the quality of nearby water bodies. There is a need to better understand storage and water quality treatment of agricultural drainage water before discharging to steams.

Over the past 40 years, constructed wetlands have become a widely accepted water treatment technology to deal with both point and non-point sources of water pollution. Constructed wetlands can improve the quality of subsurface drainage flow via a variety of physical and biological mechanisms, including sediment filtration in constructed wetlands to remove, transform or stabilize contaminants that are within agricultural runoff. The presence of aquatic plants can result in higher treatment efficiency relative to unvegetated constructed wetlands. Furthermore, removing or harvesting aboveground vegetative growth can enhance the nutrient removal capability.

Wetlands restored in western Martin County for multipurpose water quality, duck habitat and recreational goals proved very effective at removing nitrate and reducing peak flow of surface runoff, but phosphorus removal was less effective. Given the current high value of corn and soybeans there is strong economic pressure to find water storage and treatment approaches that fit into marginal farmland areas, such as stream valleys that are frequently flooded. Stream valleys in agricultural regions of Minnesota have distinct soil, topography and hydrology that will determine their effectiveness or lack of effectiveness at removing sediment and nutrients. For example the presence of clay soils would limit the infiltration capacity of drainage water in the treatment wetland, while sandier, coarse-textured soils would more quickly infiltrate drainage water.

Prior to this project there was little information on the effectiveness of treatment wetlands relative to other tile drainage Best Management Practices (BMPs). The installation of two bioreactors and a controlled drainage system on a farm in the Blue Earth River Basin provided an opportunity, in cooperation with the landowner and our partners at the Martin Soil and Water Conservation District, to assess the relative effectiveness of these three BMPs in treating tile drainage effluent. This project focused on evaluating a constructed wetland, while a parallel project at Martin SWCD evaluated the bioreactor and controlled drainage systems.


  1. Evaluate the effectiveness of constructed wetlands to remove sediment, nitrate, total phosphorus and ortho-phosphorus from tile drainage water as compared to other on-site tile drainage BMPs.
  2. Assess the contribution of denitrification, plant nutrient uptake and biomass harvest to the overall nutrient treatment efficiency of the constructed wetland.
  3. Assess hydrologic effects of treatment wetland to determine if there is any water volume reduction through evapotranspiration in the wetland.


Research was conducted at the farm of Darwin Roberts, located in Granada, Minnesota. The Roberts farm is located within the Blue Earth river basin and is bisected by Elm Creek.

A treatment wetland was constructed.  A 20-acre, tile drained field provided effluent water directly into the consturcted wetland. The landowner continued farming the field as normal, throughout the project.

Installation of water monitoring stations occurred following completion of wetland construction.

  • A monitoring station will be established above and below the constructed wetland.
  • ISCO area-velocity probes were installed at the inlet of the wetland and
  • Agri-drain water control structures with known stage-discharge equations were used to compute discharge in combination with a pressure transducer (Hobo, Solinst or other brand) to record the water level in the outlets.
  • Nest of wells and peizometers were used to determine the groundwater gradient (discharge or recharge) in the wetland.
  • A recording rain gauge and weather station were installed on site for accurate rain measurements and estimates of potential evapotranspiration based on weather data.

Water quality sampling methods

Water quality samples (analyzed for nitrate-nitrogen, total phosphorus, orthophosphorus and total suspended solids) were collected weekly for baseflow samples using grab samples in coordination with the Martin SWCD staff. Storm events were sampled using additional samples from an ISCO automatic water sampler installed at the inlet. Prior to wetland construction, soil and groundwater samples were tested for nutrient levels as well. Water samples went to MVTL labs in New Ulm while soil samples were taken to the University of Minnesota soils lab.

Vegetation methods

The treatment wetland was seeded at recommended rates with vegetation suited to the soil and moisture conditions. Details regarding species composition were determined once wetland design was finalized and flow volumes and residence time calculated.

Beginning in the second growing season, wetland vegetation was harvested to a 7 cm height in selected cells following a killing frost each year to simulate a biomass harvest regime. Nutrient uptake was calculated as the product of average biomass dry matter yield and average mineral concentration. This data will be used to estimate the relative contribution of biomass harvests to overall nutrient treatment in the wetland.

Project Completed 

Complete Report

Report Includes:
1a: Data on the percent removal of sediment, phosphorus and nitrogen for an on-farm constructed wetland within the stream valley.

1b: Direct comparison of the treatment effectiveness of three onsite BMPs for nutrient management: treatment wetland, bioreactor, and controlled drainage (in cooperation with Martin SWCD)

2a: Dry matter productivity of vegetation will be estimated each year following establishment. This information will be used to determine the potential of the system for bioenergy feedstock production.

2b. Above ground plant tissue nutrient concentrations will be for wetland vegetation in each year. This data will be used to calculate plant nutrient uptake and removal in biomass harvests.

2c. Plant nutrient uptake data will be synthesized with water sampling data to assess the effectiveness of plant nutrient uptake relative to other wetland removal processes.

3a: Data on the ability of the treatment wetland to reduce water flows through evapotranspiration.

Principal Investigator: Dean Current
Co-Investigators: Chris Lenhart, Craig Sheaffer, Don Wyse and Linda Meschke
Organizations: University of Minnesota and Rural Advantage
Sponsor: Clean Water Fund
Award Amount: $312,248
Start Date: 6/15/2012 | End Date: 12/31/2015
Project Manager: Heidi Peterson (Heidi.Peterson@state.mn.us)