Elsevier

Ecological Engineering

Volume 28, Issue 3, 1 December 2006, Pages 258-270
Ecological Engineering

Use of created wetlands to improve water quality in the Midwest—Lake Bloomington case study

https://doi.org/10.1016/j.ecoleng.2006.08.002Get rights and content

Abstract

Agricultural watersheds of the Midwest typically export nitrogen (N) and phosphorus (P) to surface waters causing contamination of drinking water reservoirs and, ultimately, hypoxia in the Gulf of Mexico. Two agricultural runoff wetlands, W1 (area 0.16 ha, volume 660 m3) and W2 (area 0.4 ha, volume 1780 m3), intercepting surface and tile drainage in the Lake Bloomington, Illinois, watershed were constructed in 1996 on forest soils (alfisols) between upland cropland and Lake Bloomington. They were created to determine whether wetlands could reduce agricultural nonpoint source pollution before it entered the Lake Bloomington drinking water reservoir. Water (precipitation, tile inflow, surface inflow, outflow, seepage and evaporation) and nutrient (N, P and carbon [C]) budgets were determined from 1 April 1998 to 30 December 1999 for each wetland. Combined, the wetlands received 746 kg NO3-N as tile loading, 104 kg as surface loading and exported 545 kg of NO3-N as outflow and seepage. Mass NO3-N retention was 36%. Following wetland treatment, overall volume-weighted NO3-N concentrations were reduced by 42% (W1) and 31% (W2). Combined P mass retention was 53%, and combined total organic carbon (TOC) mass retention was 9%. Wetlands were constructed in a sloping drainage (5%) where surface runoff was a major component of flow. Nutrient dynamics of P and C were affected by site slopes. Calculations made by extrapolating these results indicate that a wetland area of 450 ha would be required in the Lake Bloomington watershed to reduce N loading by 46%, at a construction cost ranging from 3 to 3.5 million dollars. Results support the growing evidence that agricultural runoff wetlands can effectively reduce NPS pollution loading in the Mississippi River Basin.

Introduction

The United States leads the world in food production, annually growing 40% of the world's maize (Aquino et al., 2001) and 44% of the world's soybeans (FAS, 1999). The Midwestern states of Iowa, Illinois, Minnesota, Indiana, Ohio, Missouri, and Wisconsin (in order of production) account for 58 and 70% of the maize and soybean production, respectively (USDA, 2002). This 40 million ha of land receives 6 million metric tonnes of fertilizer annually (Goolsby et al., 1999). If a conservative leaching loss estimate of 25 kg N ha−1 (David et al., 1997) is used, the Midwest would contribute at least 1 million metric tonnes of nonpoint source agricultural N to the Mississippi River annually (mainly in the form of nitrate, David et al., 1997). As nutrient-enriched waters flow to the sea, however, eutrophication of Gulf coastal waters has resulted. This problem, known as hypoxia, occurs regularly not only in the Gulf of Mexico (Goolsby et al., 1999), but throughout the world (Turner and Rabalais, 1991, Mitsch et al., 2001). Studies suggest that hypoxia may have detrimental effects on benthic organisms and other Gulf biota (Rabalais et al., 1992, Justic et al., 1996, Justic et al., 1997). Hypoxic conditions have caused collapses or reductions in fisheries in the Kattegat Sea, the Black Sea and the Baltic Sea (Diaz, 2001) and along the Oregon Coast (Grantham et al., 2004).

State NPS assessment reports have clearly identified agriculture as the greatest NPS pollution problem in the U.S. (CAST, 1992, USEPA, 2002). Illinois water quality data reflect this same trend, with 96% of the annual nitrate load derived from agriculture (David et al., 1997). Similar results were found in Iowa (Keeney and Deluca, 1993). As a result of these high nitrate loads, many municipal drinking water reservoirs in the agricultural Midwest frequently exceed the EPA maximum contaminant level (MCL) for nitrate. Of primary concern is the prevention of methemoglobinemia (blue baby syndrome) which causes neural damage in infants 6 months old or less.

The original natural ecosystems of the Midwest were stable late successional systems that retained nutrients and were dominated by perennial plants. In contrast, the agricultural systems that replaced them are highly disturbed immature (early successional) ecosystems of annual plants that characteristically “leak” nutrients (Loucks, 1979). As a result, streams and lakes in agricultural regions often contain excessive nutrient levels. Soil and above-ground carbon stores are low and levels of nutrients are high as a result of fertilizer application and ineffective uptake by monoculture cropping (Keeney, 1982, Simonis, 1988). These “leaky” systems are further exacerbated in the Midwest by extensive tile drainage. Tile drainage has decoupled wetland systems from their wetland/riverine interface on 37% of the agricultural lands in the Midwestern states discussed above, shunting contaminated upland drainage water directly to main river channels (Fausey et al., 1995).

Bloomington, Illinois, is a prime example of a Midwestern city where drinking water from surface reservoirs is consistently plagued by high nitrate levels. The city relies on Lake Bloomington and Evergreen Lake for a large portion of its drinking water-sources which at times exceed the EPA MCL of 10 mg L−1 as a result of agricultural crop production on 86% of the watershed (USDA, 1991). From 1986 through 2003, the MCL for nitrate in Lake Bloomington was exceeded annually, and the concentration of the lake remained over the MCL from February through early July in 2001.

Approximately 70,000 people receive their water from the Lake Bloomington municipal water treatment facility. The average demand is 20 million gallons per day (mgd) with a maximum of 24 mgd. When water in Lake Bloomington exceeds 10 ppm (the EPA MCL), the “solution is dilution” of the contaminated water with cleaner water from Evergreen Lake (the city's backup water supply).

Reducing nitrate levels in drinking water is important from the aspect of protecting human health, and a variety of methodologies exist to reduce nitrate levels. Often, expensive technical treatment makes sense for cities continually plagued with high nitrate levels in their drinking water and with the potential for being charged expensive fines. Failure to maintain drinking water quality standards resulted in a 1992 Illinois Environmental Protection Agency directive (authorized by the USEPA) to the city of Bloomington, Illinois, to reduce its maximum drinking water nitrate contaminant levels to below 10 mg L−1. Although some progress was made, the city was not able to meet the directive's deadline of June 1997. Two major issues in meeting the MCL were the cost of engineered systems designed to remove nitrates from drinking water and the potential effects these treatment costs would have on the economies of many agricultural communities. Ion exchange, the most cost-effective treatment available to remove nitrate is expensive and produces a salt (NaCl) brine that is expensive to treat (Letterman, 1999). The estimated cost in 2003 for construction of a 20 mgd plant was $8 million, with additional costs of up to $1 million for yearly maintenance and brine disposal (Jill Mayes, Bloomington, Illinois, Water Treatment Plant, personal communication). The plant would treat 40% or less of the water required to provide a blend concentration of 7 ppm NO3-N.

In addition to being expensive, the construction of treatment facilities is also not in keeping with the EPA goal of “source water protection” (a watershed approach). Ion exchange solves the immediate problem of high nitrates, but it does not reduce overall nitrate loading to surface waters. Source water protection is the only way to reduce nitrogen loading to surface waters. Bloomington seeks less expensive, more ecologically sound watershed-based approaches to remove nitrate from water supply reservoirs. These include both fertilizer input management and wetlands (Mitsch et al., 2001, Mitsch et al., 2005).

Constructed wetlands in tile-drained agricultural systems may prove to be a long-term practical, economical, and effective method to reduce surface water nitrate contamination. These “wetlands” are formed by berming an area adjacent to a stream and forming a small detention basin or holding pond that intercepts tile and surface drainage water before it enters the stream. The basin acts to reduce export of nitrate in drainage water through plant uptake and microbial transformation and degradation. Following wetland “treatment,” drainage water is slowly released to the stream through outlet flow. Research in Champaign County, Illinois, has shown that constructed wetlands established on former tall grass prairie soils (mollisols) can reduce total nitrogen loading from tile effluent by as much as 46% before it enters surface waters (37% of the N was denitrified (Xue et al., 1999) by the wetland pool, and an additional 9% was denitrified as it passed through the berm in seepage water) (Kovacic et al., 2000).

Before wetlands systems are adopted as a universal approach to improve surface water quality in the United States, it is imperative to demonstrate their effectiveness and adaptability to key regions of the Midwest. The goal of this study was to construct, evaluate, and demonstrate the nutrient removal capacity of created wetlands on cropland soils (alfisols) that originally supported temperate deciduous forests. N, P and water input/output budgets were determined for two experimental wetlands constructed in conjunction with two experimental systematically tile-drained sub-basins in the Lake Bloomington watershed. The Bloomington constructed wetlands and their experimental watershed served as a valuable site to investigate nutrient runoff derived from both tile drainage and surface flow.

Section snippets

Description of the study area

The study site was located adjacent to Lake Bloomington in the northwest corner of McLean County, approximately 24 km north of Bloomington, Illinois. Lake Bloomington is characterized by numerous glacial moraines that form a rolling landscape. Slopes in the region normally range between 2 and 5% but can be as steep as 50%. The land area is 86% row crop agriculture in both maize and soybean, 4% pasture, 5% woodland, and 5% water and urban areas (U.S. Department of Agriculture, 1991). In this

Water budget

In 1998 and 1999, precipitation was 1038 and 957 mm, respectively. Flow monitoring for 1998 began 8 April, and ended 31 December 1998. During this period, precipitation totaled 699 mm. Although 63–67% of all precipitation occurred in winter and spring, the two seasons accounted for 92–100% of wetland hydraulic loading. During the course of the study, surface flow, tile flow, and precipitation were 47%, 43%, and 9% for W1; and 66%, 30% and 4% for W2, respectively. In 1998 and 1999, crop

Acknowledgements

The authors thank Jim Rutherford, Robert Hoffman, and the City of Bloomington for their cooperation and assistance in establishing the wetland study site. Thanks also go to Peggy Kovacic for editing this manuscript. This work was supported through grants from the Illinois Water Resources Center (Project Number: WRC-97-1) and EPA region 5 (Regional Geographic Initiative Program, grant # X985519-01).

References (56)

  • J.A. Baldock et al.

    Soil organic matter

  • K.N. Brooks et al.

    Hydrology and the Management of Watersheds

    (1991)
  • Clover, M.W., 2005. Impact of nitrogen management on corn grain yield and nitrogen loss on a tile drained field. M.S....
  • Council for Agricultural Sciences and Technology (CAST), 1992. Water Quality: Agriculture's Role. Task Force Rep. 120....
  • M.B. David et al.

    Anthropogenic inputs of nitrogen and phosphorus and riverine export for Illinois, USA

    J. Environ. Qual.

    (2000)
  • M.B. David et al.

    Nitrogen balance in and export from an agricultural watershed

    J. Environ. Qual.

    (1997)
  • C.B. Davis et al.

    Prairie pothole marshes as traps for nitrogen and phosphorus in agricultural runoff

  • R.J. Diaz

    Overview of hypoxia around the world

    J. Environ. Qual.

    (2001)
  • D.L. Dinnes et al.

    Nitrogen management strategies to reduce nitrate leaching in tile-drained Midwestern soils

    Agron. J.

    (2002)
  • Doering, O.C., Diaz-Hermelo, F., Howard, C., Heimlich, R., Hitzhusen, R.F., Kazmierczak, R., Lee, J., Libby, L., Milon,...
  • C.F. Drury et al.

    Influence of controlled drainage-subirrigation on surface and tile drainage nitrate loss

    J. Environ. Qual.

    (1996)
  • N.R. Fausey et al.

    Drainage and water quality in Great Lakes and Cornbelt states

    J. Irrig. Drain. Eng.

    (1995)
  • J.S. Feierabend

    Wetlands: the lifeblood of wildlife

  • D.F. Fink et al.

    Seasonal and storm event nutrient removal by a created wetland in an agricultural watershed

    Ecol. Eng.

    (2005)
  • S. Fleischer et al.

    Nitrogen removal in created ponds

    Ambio

    (1994)
  • Foreign Agricultural Service (FAS), May 1999. United States Department of Agriculture. World Soybean Production....
  • Goolsby, D.A., Battaglin, W.A., Lawrence, G.B., Artz, R.S., Aulenbach, B.T., Hooper, R.P., 1999. Flux and sources of...
  • B.A. Grantham et al.

    Upwelling-driven nearshore hypoxia signals ecosystem and oceanographic changes in the northeast Pacific

    Nature

    (2004)
  • Cited by (0)

    View full text