Nitrogen balance in and export from agricultural fields associated with controlled drainage systems and denitrifying bioreactors

Abstract

Nitrate loss from drainage tiles across the cornbelt of the upper midwestern US is a result of intensive agriculture with limited crop diversity, extensive periods of fallow soil, and the need for high fertilizer applications to corn, all located on a hydrologically modified landscape. Two methods proposed to reduce tile nitrate export are managed or controlled drainage to limit tile flow and bioreactors to enhance denitrification. Nitrogen budgets and tile flow monitoring were conducted over two- to three-year periods between 2006 and 2009. We estimated N budgets in a seed corn-soybean rotation farming system near DeLand, east-central Illinois, USA, with free (FD) and controlled drainage (CD) patterned tile systems. In addition, wood chip filled trenches (bioreactors) were installed below the CD structures, one lined with plastic and one unlined. We measured daily tile flow and nitrate-N (NO₃-N) concentrations and calculated cumulative N loss from the tile water at both FD and CD areas for a period of three cropping years. We also monitored the tile flow and nitrate concentration in inlet and outlet of the bioreactor associated with a CD system and evaluated the efficiency of the bioreactor for two cropping years. Most components of the N balance were unaffected by CD (yields and therefore N harvested, surface soil denitrification), and there was a negative N balance in the soybean cropping year (−165 and −163 kg N ha⁻¹ at FD and CD areas, respectively), whereas seed corn cropping in the following year resulted in positive N balances (29 and 34 kg N ha⁻¹ at FD and CD areas, respectively). For two years, the overall N balances were −136 and −129 kg N ha⁻¹ at FD and CD areas, respectively, consistent with other recent corn belt studies showing a small net depletion of soil organic N. Controlled drainage greatly reduced tile N export, with a three-year average loss of 57.2 kg N ha⁻¹ yr⁻¹ from FD compared to 17 kg N ha⁻¹ yr⁻¹ for CD. There was high uncertainty in denitrification measurements and thus the fate of missing N in the CD system remained unknown. Nitrate reduction efficiency of the bioreactor varied greatly, with periods where nearly 100% of the nitrate was denitrified. The overall efficiency of the bioreactor associated with the CD system in reducing the tile N load was 33%. When nitrate was non-limiting, the nitrate removal rate of the bioreactor was 6.4 g N m⁻³ d⁻¹. Little N₂O emission was found from the bioreactor bed and is not thought to be a problem with these systems. Both the tile bioreactor and controlled drainage greatly reduced tile nitrate export in this leaky seed corn and soybean agricultural field.

Introduction

Agricultural intensification is recognized as a major source of increased N concentrations in surface and ground waters (McIsaac and Libra, 2003, Puckett, 1995). Increased nitrate loading in the Mississippi River has been thought to be a primary cause of the large hypoxic zone in the Gulf of Mexico (Rabalais et al., 2001, USEPA, 2008). A number of studies in the Midwest have developed field N budgets to evaluate the effects of agricultural practices on N leaching losses (e.g., David et al., 1997, Gentry et al., 1998, Andraski et al., 2000, Jaynes et al., 2001) that are the source of riverine N loads. David et al. (1997) evaluated agricultural N fluxes and sources of river nitrate in a predominantly tile-drained agricultural watershed in east-central Illinois and reported that about 49% of the field inorganic N pools was leached through tile drains and seepage and was exported by the Embarras River. Estimating the net N inputs for a period of 20 years, McIsaac and Hu (2004) reported that 100% of the residual N, the remaining N in soil after harvest, was discharged to the rivers in a tile-drained region of Illinois.

Most of the cropland in the Midwest is intensively tile-drained and corn (Zea mays L.)/soybean (Glycine max L.) rotations are the predominant cropping system (USEPA, 2008). Because N fertilizer management alone is not likely to reduce nitrate pollution sufficiently (Jaynes et al., 2001, Hong et al., 2007), additional methods of nitrate removal from subsurface drainage water are needed (USEPA, 2008). One possible method is controlled drainage (CD), sometimes called drainage water management, where structures are placed in tile lines to control the outlet depth and allow water to be temporally backed up into the field (Gilliam et al., 1979, Skaggs and Youssef, 2008, Cooke et al., 2008). Many of these systems have been installed in the upper Midwest and southern Ontario, and the reports available suggest they greatly reduce the volume of tile flow and concomitantly the amount of nitrate (e.g., Lalonde et al., 1996, Fausey et al., 2004, Drury et al., 2009). Wetlands placed at the end of tile lines have also been shown to be an effective method to reduce tile export of nitrate, but can be quite costly to build (Kovacic et al., 2000, USEPA, 2008).

Other edge-of-field methods such as setting up riparian buffer strips in areas where lateral seepage is the dominant flow (Blattel et al., 2009, Woodward et al., 2009) or constructing denitrification walls or trenches to intercept flow (Schipper and Vojvodic-Vukovic, 2001, Jaynes et al., 2008) fosters biological denitrification to remove nitrate. Use of denitrifying biofilters or bioreactors at the end of the pipe is common for treating industrial wastewater or reducing pollution from landfill sites (e.g., He et al., 2007, Morita et al., 2007), and have now been proposed for controlling tile nitrate losses. There have been some results from installation of trenches and bioreactors to reduce nitrate loss due to agriculture (Blowes et al., 1994, Schipper et al., 2010), as well as recent evaluations conducted under laboratory conditions (Greenan et al., 2009, Chun et al., 2009) or to establish field-scale flow and transport parameters (Chun et al., 2010).

Many of the systems (both walls, trenches and bioreactors) that have been designed to remove nitrate have proposed using sawdust or wood chips as the carbon (C) source to promote denitrification, and have reported that these systems did or could reduce the nitrate concentration in water flowing through the C bed (Schipper and Vojvodic-Vukovic, 2001, Jaynes et al., 2008, Greenan et al., 2009). One question concerning the use of bioreactors is the degree of N₂O production. If the nitrate is fully reduced to N₂, then there is no environmental degradation. However, if nitrate is only reduced to N₂O, a powerful greenhouse gas, then one environmental problem could be substituted for another. Greenan et al. (2009) in their laboratory column study with wood chips reported little N₂O emission and indicated complete denitrification to N₂.

In the study reported here, reductions in tile nitrate loss were evaluated from CD and subsurface, end of tile denitrifying bioreactors. The study was conducted on typical corn and soybean fields with patterned tile drainage. We also sought to determine the effect of the drainage management on the overall field N balances. Therefore, the objectives of our study were to (1) compare the field N balance with free drainage (FD) and without the CD system, (2) determine the reduction in tile nitrate export due to CD, and (3) measure how efficient a tile bioreactor was in reducing the N load.