AGRICULTURAL DRAINAGE WELL CLOSURES AND GROUNDWATER QUALITY RESPONSE: FLOYD COUNTY, IOWA
As of May 1997, the Floyd County ADW Closure Project has documented significant improvements in the groundwater quality of the middle and lower Devonian aquifers (Quade and Seigley, 1997). There has been a less dramatic decline in nitrate and pesticide concentrations in the upper part of the upper aquifer. Stratification of nitrate concentrations occurs in the upper aquifer; nitrate concentrations are higher and pesticide are more frequently detected in the uppermost part of the upper aquifer. While nitrate concentrations have declined in the upper part of the upper aquifer since closure of the ADWs, average nitrate-N concentrations remain above 10 mg/L. Removal of the point source inputs of the ADWs has allowed for the discrimination of point sources versus nonpoint source inputs.
This report summarizes the water-quality response of the Devonian aquifer system to closure of the three ADWs closest to the bedrock well nest.
Background on ADWs in Iowa and Floyd County
Many of Iowa's rich agricultural soils, particularly those in north-central Iowa, are poorly drained and at times contain excess water that can hinder field operations or ruin crops. In these areas, farm fields are often artificially drained by buried tiles leading to drainage ditches or streams. Another method, less commonly used, is the agricultural drainage well (ADW; Figure 1), a drilled shaft that funnels excess drainage water into underlying bedrock. The upper parts of these wells are often cistern-like structures that form the discharge point for tile-drainage lines; some ADWs are also designed to take surface runoff. ADWs are generally 5 to 10 inches in diameter and are cased from the land surface into the underlying bedrock. Construction of new drainage wells has been illegal since 1957. Virtually all ADWs in Iowa discharge into fractured carbonate aquifers; these strata can accept large quantities of drainage water without clogging. These aquifers are also excellent sources of groundwater for domestic, industrial, and municipal water supplies. For additional information on ADWs, see Libra and Hallberg (1993), Baker and Austin (1984), or Baker and others (1985; 1996).
There are an estimated 292 ADWs in Iowa (Mumm and Heathcote, 1997). Reported areas drained by individual ADWs range from 2 to 720 acres, and reported depths range from 12 to 400 feet. From a statewide perspective, these wells are relatively minor features. However, over 80% of the registered ADWs are concentrated within only four counties: Pocahontas, Humboldt, Wright, and Floyd. The significantly greater number of ADWs and area drained within these counties suggest a potential to affect groundwater quality.
Previous investigations of ADWs in Iowa show that while ADWs negatively impact groundwater within 0.5 to 1.5 miles of numerous ADWs, not all wells within this distance show ADW impacts, and the apparent impacts vary with time (Baker and Austin, 1984). ADW impacts were most noticeable following runoff and/or infiltration generating conditions, when surface and/or tile drainage is delivered to the groundwater via ADWs. During extended dry periods, drainage inputs are insignificant, and ADW impacts are lessened, or not noticeable. These investigations also showed that ADW effects were difficult to identify in the areas where the receiving aquifer is naturally susceptible to contamination. Water entering ADWs from tile drainage typically contains 15-50 mg/L nitrate-N and 1-10 µg/L of commonly used herbicides. Direct surface runoff into ADWs may contain herbicides in the 10-100 µg/L range, while nitrate-N concentrations are commonly less than 10 mg/L. Influent surface water could contain bacteria and potentially pathogenic organisms that are less likely to occur in tile effluent. Beyond the routine delivery of drainage water with typical agricultural contaminants to aquifers, ADWs pose other risks to groundwater quality. Some ADWs are connected to drainage systems that accept water from road ditches. Therefore, spills or leaks of harmful substances into these ditches could quickly and directly impact groundwater supplies. In addition, large-scale hog confinement facilities have been built in close proximity to ADWs, specifically in Wright County.
A variety of Best Management Practices (BMPs) have been proposed and some implemented to mitigate the water-quality impact of ADWs. Because of the variability in the geologic setting surrounding these ADWs, not all of the proposed BMPs are suitable for all ADW areas. A study by the Iowa Department of Agriculture and Land Stewardship (IDALS) evaluated the following four management alternatives for ADWs (IDALS, 1994):
1) Closure of the ADWs and development of alternative drainage
outlets to surface water.
As of August 1994, Floyd County had an estimated 68 active ADWs; 23 wells have been closed with project assistance as part of the Floyd County Groundwater Protection Project (Moore, 1996). Most of the ADWs in Floyd County are less than 100 feet deep, while a few exceed 250 feet in depth. Three of the ADWs identified for closure, and eventually closed, were the three nearest the bedrock well nest (FM3; Figure 2).
In the fall of 1994, the Iowa Department of Natural Resources - Geological Survey Bureau (IDNR-GSB) initiated the "Floyd County ADW Closure Project" to monitor the expected groundwater-quality improvements resulting from closure of the three drainage wells (Figure 2). In addition to sampling the bedrock well nest (FM3) and a nearby glacial till well (site FM3-T), the combined tile drainage from the three ADWs was sampled, as well as water from the drainage ditch prior to its entering Beaver Creek. The quality of the diverted drainage was also monitored to assess impacts on the receiving stream and address concerns about downstream leakage of the diverted drainage from Beaver Creek to the underlying groundwater system. Five nearby private wells were also sampled.
In December 1994, the three drainage wells (ADW-1, ADW-2, and ADW-3) nearest the bedrock well nest were closed (Figure 2). The project provided cost-share assistance for development of an alternate surface water outlet for tile drainage. After the three wells were closed, the tile water was diverted to a constructed drainage ditch, along which the water travels for approximately three miles before emptying into Beaver Creek.
Hydrogeologic Setting of the Floyd County ADW Closure Area
Figure 3 is a map of the geologic regions of Floyd and Mitchell counties. Most of the ADWs in Floyd County are located in the deep bedrock geologic region. Libra and others (1984) reported depth to bedrock at greater than 50 feet in this area. However, revised depth to bedrock maps of Floyd County, as well as an updated sinkhole coverage (Hoyer and Hallberg, 1991), showed that the deep bedrock area is not as large as once mapped. In actuality, the areal extent of the deep bedrock area may be more limited than Figure 3 indicates.
Stratigraphic and hydrogeologic studies (Witzke and Bunker, 1984 and 1985; Witzke et al., 1988; Libra et al., 1984) suggest that the Devonian strata in Floyd and Mitchell counties are best described as a three-part aquifer system with the major water-producing carbonate strata separated by intervening shales and shaley carbonates. Figure 4 illustrates a vertical section of the three-part Devonian aquifer, depths of wells at the FM3 site, and the depth of the three ADWs closed in December 1994. Three major aquifers are noted on Figure 4; the Spillville Formation forms the lower aquifer, the Basset Member of the Little Cedar Formation acts as the middle aquifer, and the upper aquifer consists of the Hinkle and Eagle Center members of the Little Cedar, Coralville, Lithograph City, and Shell Rock formations. Major confining beds include the Chickasaw Shale Member of the Little Cedar Formation and the Pinicon Ridge Formation. These confining units are believed to be fairly extensive in north-central Iowa (Witzke and Bunker, 1985).
Water-Quality Monitoring Design
To eliminate the point source inputs from these ADWs, the Floyd County Soil and Water Conservation District closed the deep ADW (ADW-1; >300 feet) at this site as well as two nearby ADWs (ADW-2 and ADW-3) that are approximately 65 feet deep. The three ADWs were within a one-half mile radius of the IDNR-GSB well site (FM3; Figure 2). The ADW closure and eventual diversion affected 280 acres of land and involved three landowners who agreed to closure of the ADWs.
The Floyd County ADW Closure Project monitored water quality at nine sites on a monthly basis. Monitoring sites include the bedrock well nest, a glacial till well, the tile-line drainage from the three closed ADWs, two downstream surface water sites on Beaver Creek, and five private wells that are within a two-mile radius of the three closed ADWs (Figure 2). (Since December 1994, two additional ADWs within two miles of the bedrock well nest have also been closed.) The FM3 site is within 500 feet of one of the closed ADWs and has a long-term monitoring record. Tile drainage from the ADWs now drains into a drainage ditch referred to as ADW-DRAIN. The ADW-DRAIN eventually drains into a tributary of Beaver Creek. Surface water sampling points are located at a bridge site on a tributary to Beaver Creek (TRIB-Up) and at another bridge site immediately downstream on Beaver Creek (BC-Down). The selection of these various sites allows for the monitoring of groundwater, shallow groundwater, and surface water.
Quality of Shallow Groundwater Entering the ADWs
Prior to closure, shallow groundwater from the tile lines entering the three ADWs were sampled for nitrate-N and pesticides. Typically, the tile lines do not run during very dry periods or in extremely cold weather. At such times, there is little or no delivery of water via the ADWs to the underlying bedrock. During periods when the tiles were discharging, nitrate concentrations of the water varied from 2 to 35 milligrams per liter (mg/L) nitrate-N. The highest nitrate concentrations occurred during major rainfall periods and the lowest during drier periods. Figure 5 shows the nitrate-N and pesticide concentrations from ADW-1 for the period 1984 to closure in December 1994. (Note: Lack of funding prevented monitoring of ADW-1 and the FM3 wells from 1987 through 1989.) The nitrate-N and pesticide concentrations varied through time in response to seasonal flow as well as variations in rainfall. Water samples were also collected from ADW-2 and ADW-3; sampling of these sites occurred on a monthly basis from mid-1993 until their closure in December 1994 (Figures 6 and 7).
The tile lines entering these ADWs had a high frequency of pesticide detections. Sixty-nine percent of the samples from the tile line entering ADW-1 had at least one pesticide detected. Eighty to 100% of the samples from the tile lines entering the two shallow ADWs had one or more pesticides detected. Atrazine (Aatrex) was the most commonly detected pesticide. Metolachlor (Dual), alachlor (Lasso), cyanazine (Bladex), and metribuzin (Sencor) were also detected.
Groundwater Quality in the Glacial Till Well
Nitrate-N concentrations at FM3-T have varied from below detection to 22 mg/L during the pre-closure and post-closure monitoring. Atrazine concentrations have varied from below the detection limit (0.10 µg/L) to 0.12 µg/L, with the majority of samples below the detection limit. Figure 8 illustrates the trends in nitrate and atrazine concentrations at FM3-T.
Nitrate-N concentrations, atrazine concentrations, and frequency of atrazine detections have increased at FM3-T since closure of the ADWs. The mean nitrate-N concentration at FM3-T increased from 4.5 mg/L prior to closure of the ADWs to 7.5 mg/L once the ADWs were closed; mean atrazine concentration increased from below the detection limit of 0.10 µg/L to 0.11 µg/L, and the frequency of atrazine detections increased from 35 to 80 percent. Improvements in water quality from the glacial till well were not expected with the closure of the ADWs, as none of the closed ADWs injected water to the shallow water table. The increase in nitrate-N and atrazine concentrations are likely related to year-to-year changes in climatic conditions and changing farm practices (i.e., fertilizer and chemical use) on the field adjacent to the well.
Groundwater Quality in the Bedrock Study Wells
During pre-closure monitoring, the maximum nitrate-N concentrations in the "nested" bedrock wells varied from 54 mg/L in the upper part of the upper aquifer (Figure 4) (FM3-1; see Figure 9), to 9 mg/L in the lower part of the uppermost aquifer (FM3-2; see Figure 10), to 0.5 mg/L in the middle aquifer (FM3-3; see Figure 11), and to 8.5 mg/L in the lower aquifer (FM3-4; see Figure 12). During relatively dry periods, nitrate-N concentrations from FM3-1 were usually less than 10 mg/L. The tile draining to ADW-1 (and other observed tile outlets) was dry or supplying a minimal amount of water during these periods.
Pre-closure monitoring during wet periods documented increases in water levels in the "nested" bedrock wells; these increases were in response to tile water input from the ADWs. At these times, nitrate-N concentrations in FM3-1 (shallowest well) exceeded 30 to 45 mg/L (see Figure 9), and these changes occurred within just a few days. The deepest well, FM3-4 (see Figure 12) also showed increased nitrate-N concentrations during rainfall events, but concentrations were not as high as in FM3-1. The earlier documented increases in nitrate concentrations in some of the bedrock wells are interpreted as the result of inputs from nearby ADWs during wet periods. Higher nitrate-N concentrations at FM3-1 may have occurred for several reasons. First, the two nearby shallow ADWs (ADW-2 and ADW-3) may not have delivered water to bedrock much deeper than the depth of well FM3-1 (103 feet), and the water delivered may only have impacted the shallowest part of the upper aquifer. Shallow bedrock at this site is unsaturated, that is, fractures or cracks in the bedrock are air-filled and do not contain water. As water enters the bedrock through the ADWs, it is not subject to mixing and dilution with water already present in the bedrock.
Post-closure declines in mean nitrate-N concentrations at all bedrock wells at site FM3 are illustrated in Figure 13. The decline in nitrate-N concentrations was statistically significant (p=0.05; Kruskal Wallis Analysis of Variance) at wells FM3-2, FM3-3, and FM3-4. Post-closure mean nitrate-N concentrations at these sites were negligible. The mean nitrate-N concentration at the FM3-1 site also declined from 19 mg/L during the pre-closure period to 12 mg/L during the post-closure period. The majority of samples collected from FM3-1 during the post-closure period, however, exceeded 10 mg/L nitrate-N. Of note is the decrease in the range of nitrate-N concentrations of this well from the pre- (0.8 to 54.0 mg/L) to post-closure period (7.2 to 20.0 mg/L). Nitrate-N concentrations in tile water inputs to ADWs vary significantly, both temporally and spatially. As a result, the range in nitrate concentrations tends to be large. With closure of the ADWs, the point source inputs have been removed. The more consistent, yet elevated, nitrate concentrations most likely reflect the inputs from nonpoint sources.
Pesticide data from the bedrock wells paralleled trends in the nitrate-N data. During the pre-closure period, pesticides were detected primarily in the shallowest (FM3-1; see Figure 9) and deepest wells (FM3-4; see Figure 12), those wells directly affected by ADWs. The highest pesticide concentrations at FM3-1 and FM3-4 occurred during wet periods when tiles were discharging water via the ADWs into the bedrock aquifers. During the pre-closure period, 84% of the water samples from FM3-1 had at least one pesticide detected, as did 46% of the samples from FM3-4. The frequency of pesticide detections was lower at FM3-2 (14%) and FM3-3 (5%).
Post-closure monitoring has documented a decrease in the mean atrazine concentration at all of the bedrock wells at FM3 (Figure 14). Mean atrazine concentrations for wells FM3-2, FM3-3, and FM3-4 were below the detection limit (0.1 µg/L) during the post-closure period; the decline in mean atrazine concentration is statistically significant (p=0.05) for well FM3-4 (see Figure 14). The frequency of atrazine detections declined in all of the bedrock wells except FM3-1, where the percent of detections increased from 90 to 95 percent. During the pre-closure period, atrazine, atrazine metabolites, cyanazine, metribuzin, and metolachlor were detected in the wells. During the post-closure period, only atrazine and atrazine metabolites were detected, and those were found only in FM3-T and FM3-1. One sample collected from FM3-4 in April 1995 did contain metolachlor; it has not been detected since.
Water Quality at ADW Drainage and Surface Water Sites
Water Quality Response in Private Wells
Funding for this project has been provided by the Iowa Department of Agriculture and Land Stewardship Water Quality Protection Funds, administered through the Floyd County Soil and Water Conservation District. Water quality analyses have been provided by The University Hygienic Laboratory.