Constructed wetland used to treat nitrate pollution generated from agricultural tile drainage waters in Southern Minnesota

Nitrate molecules are highly soluble in water and are bioavailable to plants. These properties are why excess nitrates in water are one of the main causes of hypoxia in the northern Gulf of Mexico. Over 90% of these nitrates originate from non-point sources such as agricultural fields. In fields with tile drainage systems nitrates have a swift passageway from field to surface waters. This study focuses on one Midwestern farm field located in southern Minnesota, along Elm Creek, a Blue Earth tributary. Tile drainage water from this field discharges into Elm Creek at a concentration averaging 23.0 mg/L NO3 as NO3-N. During the spring of 2013 a three celled treatment wetland was constructed adjacent to Elm Creek. The tile drainage system was re-routed to discharge into the constructed wetland. In the 2013 field season water volumes were monitored continuously and water samples were taken from the inlet, the wetland cells, and the outlet on a periodic basis. During the season the volume of tile drainage water that reached Elm Creek as surface water was reduced by 82%. The concentration of NO3-N in the water was not significantly reduced. However, the total load of NO3-N that reached Elm Creek as surface water was reduced by 262 to 332 pounds (14.4-18.2 lbs./acre). Most of the water that did not reach Elm Creek infiltrated into the subsurface soils and still contained NO3-N. Using the MPCA's (2013) estimates of groundwater denitrification for agroecoregions, a 45% reduction rate was applied at this location. When the 45% reduction rate is applied to the subsurface load it is estimated that 113.0 to 134 lbs. (6.21-7.36 lbs./acre) of NO3-N were removed from the infiltrated water. Thus a total of 124 to 172 lbs. (6.81-9.45 lbs./acre) of NO3-N were removed from the entire wetland system which accounts for 37.1-43.3% of the NO3-N.A concurrent laboratory experiment was set up in 2013 to test the effectiveness of different soils and vegetation at removing nitrate loads. Wetland mesocosm experiments were set up with soil collected from the field site and the design vegetation used in the field cells. Three vegetated mesocosm tanks were planted in Coland soils with Switchgrass (Panicum virgatum), Fringed Sedge (Carex crinita) and a tank with an equal mix of Dark Green Bulrush (Scirpus atrovirens), Panicum virgatum, and Carex crinita. The results showed that the mixed vegetation regime and the Panicum virgatum had significantly greater nitrate removal than the control (Coland bare soil). The mixed vegetation mesocosm had the highest amount of nitrate removal after 10 days at 34.9%. There was no significant difference in the nitrate removal rates in the soils tested.