Nutrient removal and greenhouse gas emissions from 20 year old constructed wetlands receiving tile drainage water: a biogeochemical analysis

Agricultural nutrient loss (N and P) from the Midwestern United States is an important issue, especially from tile drained land. There are a wide range of in-field and edge-of-field practices to reduce losses that have been studied and are currently recommended, such as cover crops, fertilizer management, drainage water management, wetlands, wood chip bioreactors, and saturated lateral buffers. In-field practices can lead to reduced crop yields and may cause problems with annual implementation, so edge-of-field methods are viewed as an alternative to avoid these problems. Constructed wetlands receiving tile drainage have the potential to remove and retain both N and P from agricultural runoff. This has been shown through many studies. However, little is known about how the N and P removal processes and efficiencies change as constructed wetlands age. In addition, there have been few measurements of greenhouse gas, GHG, emissions from agricultural constructed wetlands. This study evaluated N and P removal and greenhouse gas emissions in three 20 year old constructed wetlands that receive tile drainage from corn and soybean fields in southern Champaign County, Illinois. All three wetlands were equipped with data loggers and pressure transducers at their inlets and outlets to measure flow, using Agri Drain control structures. Water samples were collected to determine N and P concentrations and loads. Wells were also installed in and outside of the berms of wetlands A and B in order to determine the volume of seepage water, and the nitrate load lost through this pathway. Using the inlet, outlet, and seepage nutrient load data, nutrient budgets were constructed, and N and P removal rates were calculated. Greenhouse gas fluxes, including carbon dioxide, methane, and nitrous oxide, were measured from the inundated and terrestrial portions of the wetlands via floating and static chambers, respectively. These fluxes were measured throughout the year, and were linearly interpolated in order to construct a cumulative flux. The wetlands removed approximately 46% of the nitrate-N and 2% of the total P inlet load when they were first established. During two years of study (2012-2013) I determined that the wetlands removed on average 59% of the nitrate-N and 32% of the total P inputs. Hydraulic loading had a strong, positive relationship with the mass of N removed per hectare with a R2 value of 0.73. The predictability of N removal increased when the nitrate inlet flow weighted mean was also considered in a multiple linear regression. Together, the hydraulic loading and average nitrate concentration explained 85% of the variation in N removal. Total P wetland retention was more difficult to explain, and varied greatly between water years and wetlands, with removal ranging from -7 to 100%. Unlike N removal, P retention was not strongly related to hydraulic loading and inlet P flow weighted means. The dominant GHG emitted from the wetlands was carbon dioxide, which made up between 75 and 96% of the total GHG emissions. The nitrous oxide flux, which was of special concern due to the denitrification process, only contributed between 3.7 and 13% of the total cumulative GHG flux. Methane made up even less, between 0.08 and 12%, of the total cumulative GHG flux. Further, the terrestrial portions of the wetlands emitted the majority, between 86 and 99%, of the total GHG emissions. Nitrous oxide emissions were 7.6 and 3.1% of the total nitrate loss from wetlands A and B in 2012 and 2013, respectively, with the larger percentage when the wetlands were mostly dry because most of the losses were from terrestrial portions of the wetlands. The GHG fluxes, from both the inundated and terrestrial portions of the wetland had a threshold with water and soil temperature respectively. GHG samples collected from water below 18⁰C or soil below 15⁰C typically had low concentrations, which ultimately translated to low fluxes. All of the large methane and nitrous oxide fluxes observed took place above these temperature thresholds. Soil moisture was also correlated to terrestrial GHG fluxes. Terrestrial GHG fluxes with a soil moisture level above 25% typically had larger fluxes than when they were drier. Overall, these wetlands continue to function well in controlling nitrate and total P losses from tile drained agricultural fields in Illinois. Wetland age has not affected nitrate removal. Maximizing hydraulic loading leads to the greatest nitrate removal per ha of wetland. Nitrous oxide emissions were a small percentage of nitrate removed, although this was larger during the drought year of 2012