Application of Drainage Water Management and Saturated Buffers for Conservation Drainage in South Dakota

Edge of field practices such as drainage water management (DWM) and saturated buffers can reduce nutrient transport from croplands to surface waterbodies. DWM uses stackable weir boards in a control structure to manipulate the water table depth throughout the cropping season and reduce the amount of nutrient rich water draining out from the field. Saturated buffers, on the other hand, use a control structure to divert water draining out from the cropland to a vegetative strip via a subsurface tile installed parallel to a waterway. For the saturated buffer systems, a combination of natural denitrification, nitrogen mineralization, and plant uptake are the major causes of nitrate reduction. This study was conducted at three field sites across eastern South Dakota and the overall goal was to evaluate the effectiveness of DWM and saturated buffers in reducing nitrate loads transported from the field. A DWM site was established near Alexandria, SD in the fall of 2015 and two saturated buffer sites were established near Flandreau, SD and Baltic, SD in the summer of 2016. Water samples were collected weekly when water was flowing through control structure and analyzed for nitrate concentration in the lab. Flow records from the DWM sites were divided into two periods, the free drainage period and the managed period. Results for the Alexandria site showed that DWM reduced the total annual outflow by 8mm for 2016 during the managed period. In addition, nitrate concentrations between the two halves of the site were compared and it was observed that the DWM half had lower nitrate concentrations as compared to conventionally drained half for most of the study. Annual nitrate loads for DWM and conventional half were calculated to be 3.3 kg ha-1 and 4.4 kg ha-1 during 2016 and 1.4 kg ha-1 and 2.3 kg ha-1 during 2017, respectively. Overall, DWM resulted in a load reduction of 26% during the managed period. Results for the Baltic site show an overall nitrate concentration reduction of 95% for 2017, during which time 100% of water was diverted to the buffer. For Flandreau, the overall nitrate concentration reduction for 2016 was 86%, during which time 97% of water was diverted to the buffer and 65% for 2017 when 83% of water was diverted to the buffer. The lower reduction rate for 2017 was attributed to the high flow volumes that were diverted to the buffer zone throughout most of the season, resulting in inadequate nutrient uptake by the plants and insufficient time for natural denitrification. In addition to the field study, a SWAT model was developed to assess the impact of the variability in soil properties and tile design parameters on flow volume reduction for DWM. The model was developed for the research site at Alexandria and daily measured flow data from the field study for 2016 were used for calibration, and 2017 was the validation period for the project. Model performance was evaluated using three statistics, NSE, RSR and PBIAS. The evaluation statistics ranged from 0.54 to 0.84 for NSE, -23% to 61% for PBIAS and 0.40 to 0.68 for RSR. It was concluded that SWAT simulations accurately represented the hydrological processes for the research site and that DWM resulted in an increase in ET, lateral flow and surface runoff while decreasing tile flow during relatively wet years. During dry years however, DWM resulted in an increase in tile flow. Apart from climatic conditions, DWM performance was also affected by variability in soil properties such as bulk density and available water capacity, and tile design parameters such as drain tile lag time and time to drain to field capacity. A financial comparison between the two systems showed that DWM had a higher cost per pound of nitrate removed per acre at $28 lb-1 ac-1 observed for Alexandria as compared to $22 lb-1 ac-1 and $0.6 lb-1 ac-1 for Baltic and Flandreau respectively. The lower cost for the buffer systems can be related to a higher cumulative load reduction for the study period. Overall, both management practices were successful in reducing nitrate loads from drained croplands and expanding the model to a sub watershed or watershed scale could facilitate in decision making for agricultural water management in South Dakota