EXECUTIVE SUMMARY

In Indiana, there are extensive agricultural areas developed on glacial outwash deposited in floodways of major streams and on outwash fans and wind-blown sands. Because infiltration is rapid and direct in these hydrogeologic settings, ground water is subject to degradation by nitrates derived from agricultural sources such as inorganic fertilizers, crop residues, applied manure, and manure from grazing cattle. However, a previous investigation in Jackson County, Indiana, indicated that denitrification occurs as ground water migrates downward, so that shallow aquifers in this type of hydrogeologic setting are candidates for remediation.

As part of this investigation, a variety of experimental practices were employed by participating farmers in an attempt to reduce the leaching of nitrates into ground water. Ground-water chemistry was monitored in six irrigated test plots to determine if such practices resulted in measurable improvements in water quality. The experimental practices that were employed included (1) systematic reductions in the application of nitrogen fertilizers from 185 lbs/acre of total chemical nitrogen (N) to only 80 lbs/acre N, (2) application of nitrification inhibitors (N-serve and Agrotain), and (3) aerial seeding of rye in corn fields late in the growing season.

The results of the investigation indicate that, following application of standard quantities of inorganic fertilizers (typically about 185 lbs/acre N), nitrate concentrations in the uppermost 18 inches of ground water naturally decline to values that are less than the Drinking Water Standard limits (10 mg/l) in a period of about 6 to 18 months.

When nitrification inhibitors are used in conjunction with greatly reduced fertilizer applications (to as little as 80 lbs/acre N), concentrations of nitrate-nitrogen exhibit relatively small increases--or even slight decreases--in ground water in post-application rounds of sampling. However, it was determined that such large reductions in the application rate of fertilizers are associated with reductions in corn yields of 10 to 20 bushels/acre. Yields might have been even more adversely affected if nitrate-bearing ground water had not been used for irrigation.

An attempt to evaluate the effects of planting rye was inconclusive.

Reliable and long-term reduction of nitrate concentrations in ground water to less than 10 mg/l might require even greater reductions of fertilizer applications to less than 80 lbs/acre N. Such further reductions would lead to declines in crop yields that would be unacceptable to farmers. Consequently, with current agricultural practices, persons who wish to use shallow ground water as a domestic supply of drinking water in such hydrogeologic settings should consider the creation of protection areas around their wells. The results of the investigation indicate that setbacks of 200 feet, and possibly less, may be adequate, particularly if drainage ditches can be installed between the protection areas and cultivated fields.

INTRODUCTION

Ground-water quality in the major glacial-outwash aquifer system that occupies the floodplain of the East Fork of the White River, northwest of Seymour in Jackson County, Indiana, has been seriously degraded by anthropogenic nitrate. Most of the shallow (less than 25 feet deep) domestic drinking-water wells produce water having nitrate concentrations that approach or exceed the Drinking Water Standard limits of 10 mg/l, established by the U.S. Environmental Protection Agency (EPA).

An earlier investigation, conducted between September 1999 and September 2003 in an area north of Cortland, identified the sources of nitrogen that are responsible for contamination of the aquifer. The study area was a 15-square-mile section that lies northwest of the river, where the aquifer is 30 to 60 feet thick and consists of sand and fine gravel, overlain by 6 to 12 feet of somewhat less permeable overbank (alluvial) material (fine sand and silt with some clay). The soils are thick, level, well drained, and moderately fine- to coarse-grained in texture, with relatively low organic content. The aquifer lies on lake clay and shale, both of which have very low permeability. Samples collected from wells installed in farm fields implicated inorganic sources (i.e., fertilizers) or mixed inorganic and organic sources (i.e., animal waste), whereas wells adjacent to and down-gradient from concentrated animal feeding operations (CAFOs) indicated animal waste sources.

The earlier investigation also indicated that denitrification occurs in the aquifer as the vertical component of the ground-water flow system carries the nitrate downward into the anoxic zone at a depth of 25 to 30 feet. It was suggested that, with the implementation of best management practices (BMPs), naturally occurring denitrification and recharge from precipitation make this aquifer a potential candidate for remediation. For a more detailed technical report on the earlier investigation, go to the Jackson Nitrate Project Web page.

The hydrogeologic setting described above is common to the floodways of many of the major streams that flow in glacial meltwater channels in Indiana. Also, there are other geologic settings in glaciated portions of Indiana that create similar hydrologic conditions. For example, outwash fans in the northeastern part of the state and wind-blown sands in the northwestern part are similar hydrologically in that they consist of thick deposits of coarse-grained materials (Fig. 1). Ground water in each of these settings is quite sensitive to contamination from surface-applied chemicals.

Figure 1. Map showing areas in Indiana that have hydrogeologic settings that are the same or similar to the setting of the study area. The total area is almost 3 million acres, of which 26 percent was planted in corn in 2003. The map was derived from a statewide map showing hydrogeologic settings by Fleming and others, 1995 (Atlas of hydrogeologic terrains and settings of Indiana, Indiana Geological Survey Open-File Report 95-7). Fleming used existing maps of surficial geology, as well as logs of water wells on file at the Indiana Department of Natural Resources, Ground Water Section, to construct three-dimensional interpretations of the hydrogeology of Indiana.

This report discusses a cooperative effort of the Indiana Geological Survey (IGS), local farmers, and local farm service companies to establish a program to implement BMPs for fertilizer application and to continue ground-water monitoring in the same area as the earlier investigation. The purpose of the project, which lasted from October 2003 to October 2007, was to assist farmers with optimizing the rate, method, and timing of their nutrient applications and thereby minimizing the entry of nitrate into the aquifer. Six test plots were established, in which monitoring wells were installed to collect samples of water that leached from those fields to the aquifer in order to track any changes in nitrate concentrations through time that might indicate which experimental procedures are most effective. Selected monitoring wells from the earlier investigation continued to be sampled on a quarterly basis to provide information on changes in nitrate concentrations in the traditionally managed areas surrounding the test plots.

GROUND-WATER MONITORING

Monitoring Wells. In consultation with cooperating farmers, six fields, each of which is approximately 40 acres, were selected to serve as test plots (Fig. 2). Six shallow monitoring wells (Wells 51, 52, and 54 through 57) were installed in these fields in November 2003 for the purpose of monitoring nitrate concentrations in the uppermost 18 inches of the aquifer. As a control, a well was also installed in an adjacent, uncultivated wood lot (Well 53). In addition, several preexisting monitoring wells from the earlier investigation were included in the study.

The wells, which were 2 inches in diameter, were constructed with 3 feet of PVC casing above 10 feet of screen (see schematic diagram). The annulus around each screen was packed with silica sand and sealed at the surface with 2 feet of bentonite. The general setting of each monitoring well can be seen in photographs of monitoring wells.

Figure 2. Aerial photograph of the study area, showing the locations of test plots (magenta outlines) and monitoring wells. Green squares indicate monitoring wells in test plots. Red circles indicate preexisting monitoring wells from an earlier investigation in surrounding areas.

Depending on precipitation, it was observed that the level of the water table can vary seasonally by as much as 5 to 9 feet. On November 9, 2004, seven new wells (referred to as "Wells 51A through 57A") were installed immediately adjacent to the existing wells. These wells, which had 6 inches of casing above 10 feet of screen, were installed to provide a screened interval within the previously unscreened depths between 0.5 and 3 feet below the surface (see schematic diagram). These wells were sampled during periods when the water table was exceptionally high.

Precipitation within the study area (measured as "inches of water") was continuously recorded using a tipping-bucket rain gauge connected to an electronic data logger, and daily total accumulations were then calculated. In selected wells, water levels (measured as "feet above mean sea level") were measured hourly using a Solinst Levelogger Gold instrument, which is a non-vented pressure transducer with onboard data logging, corrected for atmospheric pressure fluctuations with a barometric logger. Daily average values were then calculated.

Sampling Method. Samples of ground water were extracted from the monitoring wells by using a low-flow submersible pump; samples from the wells in the test plots were obtained from the water column immediately below the water table (within 18 inches) by insertion of an inflatable bladder at the base of the pump. Collection of samples from the uppermost part of the saturated zone provided evidence of nitrates that are leached from the overlying, unsaturated soil profile by water that infiltrates after precipitation or irrigation events.

Sampling Schedule. Approximately four water samples were collected annually from the monitoring wells in the test plots and from selected preexisting wells in surrounding areas. Samples were collected before and after the application of nitrogen fertilizers (described below), around harvest time, and during winter. The post-application samples were not collected until after the area had experienced a significant rainfall or irrigation event. Sampling dates are listed in Table 1.

TABLE 1. Collection of Ground-water Samples
Date	Sampled Wells	Reason for Sampling
2004 CALENDAR YEAR
March 2	5, 51 through 57	Before application of nitrogen fertilizers
March 3	3, 4, 6, 9, 11, 22
May 5	51 through 57	After application of nitrogen fertilizers associated with Wells 54 and 55
June 1	51 through 57	After application of nitrogen fertilizers associated with Wells 56 and 57
June 2	3, 4, 5, 9, 11, 22	After application of nitrogen fertilizers
October 21	3, 4, 5, 9, 11, 22, 51 through 57	After harvest
2005 CALENDAR YEAR
January 18	51 through 57	Mid-winter sampling
April 15	3, 4, 5, 6, 9, 11, 22, 51 through 57	Before application of nitrogen fertilizers
June 13	51 through 57	After application of nitrogen fertilizers
June 28	3, 4, 5, 6, 9, 11, 22
November 2	3, 4, 5, 6, 9, 11, 22, 51 through 57	After harvest
2006 CALENDAR YEAR
February 1	3, 4, 5, 6, 9, 22, 51 through 57	Mid-winter sampling
April 18	3, 4, 5, 6, 9, 22, 51 through 57	Before application of nitrogen fertilizers
May 16	9, 51 through 57	After application of nitrogen fertilizers
May 22	3, 4, 5, 6, 22	After application of nitrogen fertilizers
June 5	56 and 57	After application of pesticides
June 20	56 and 57	After application of pesticides
November 9	3, 4, 5, 6, 9, 22, 51 through 57	After harvest
2007 CALENDAR YEAR
May 1	51 through 57	Before application of nitrogen fertilizers
June 18	51 through 57	After application of nitrogen fertilizers
September 18	3, 4, 5, 6, 9, 22, 51 through 57	After harvest

Chemical Analyses. Water samples were submitted to the Indiana State Department of Health (ISDH) for analysis. Analyses performed by the ISDH included nitrate plus nitrite as N, chlorides as Cl, sulfates as SO₄, phosphates as PO₄, sodium, and potassium. All analyses were reported as "milligrams per liter" (mg/l).

Isotopic Analyses. Isotopic ratios of nitrogen (¹⁵N/¹⁴N) can be used as indicators of nitrogen sources. Within the study area, several such sources existed, including inorganic fertilizers (applied primarily to corn crops), nitrogen fixed by soybeans, crop residues, and manure from cattle. Relatively low values of nitrogen isotopes indicate relatively greater influence of inorganic fertilizers (-5 to +5 per mil), while relatively high values indicate a greater influence of organic sources such as manure (10 to 20 per mil) or vegetable matter (3 to 9 per mil).

In order to qualitatively evaluate the relative importance of various nitrogen sources in the test plots, samples were collected on May 16, 2006, from each of the test wells. At the time the samples were taken, inorganic fertilizers had recently been applied to Fields 54 through 57, and soybeans had been harvested in Fields 54 and 55 during the previous growing season. Also, cattle had been pastured in Fields 51 and 52 in spring 2006, and in Fields 54 and 55 in spring 2005. The samples were submitted to the Environmental Isotope Lab, Department of Earth Sciences, University of Waterloo for analyses.

The sample from Well 53 (7.49 per mil) in the wood lot, where inorganic fertilizers had never been applied and where cattle had never been pastured, indicated predominantly organic sources of nitrogen (presumably decaying wood and leaf litter) (Table 2). In contrast, samples from Wells 56 and 57 (0.19 and 0.09 per mil, respectively), where inorganic fertilizers had been heavily applied, but where cattle had never been pastured, indicated exclusively inorganic sources (presumably from inorganic fertilizers). Isotopic values from Wells 51, 52, 54, and 55 (ranging from 1.69 to 6.46 per mil), whose associated fields received large applications of inorganic fertilizers whenever corn was cultivated and where cattle were pastured during winter and spring months following corn harvests, indicated the presence of mixed organic and inorganic sources of nitrogen.

TABLE 2. Isotopic Analyses
Well	15N/14N
51	2.98
52	4.33
53	7.49
54	6.46
55	1.69
56	0.19
57	0.09

Thus, the results of the isotopic analyses confirmed the assumptions of the investigators, based on conversations with the participating farmers, regarding the relative importance of various nitrogen sources in the six test plots.

STANDARD MANAGEMENT PRACTICES

In this region, farmers typically plant their fields in an annual rotation of corn and soybeans (Table 3). With conventional practices, when corn is planted, farmers typically apply relatively large quantities of nitrogen fertilizers (as much as 210 lbs/acre N) shortly before, during, and (or) after planting. In the past, because fertilizer was relatively inexpensive, farmers applied an amount that was designed to exceed the maximum potential yield for the most productive soil series in each field. During years when soybeans are being cultivated, nitrogen fertilizers are not applied. (As a rule of thumb, the participating farmers consider a soybean crop to be the equivalent of applying 25 to 30 lbs/acre N in the following year.) Other crops are occasionally planted, including green beans. When green beans are planted, their harvest is normally followed by the planting of soybeans, providing two crops during a single growing season. Like corn, green beans require nitrogen fertilizers.

TABLE 3. Dates of Chemical Applications and Plantings
Field	Date	Chemical Application	Application Rate	Planting
2004 GROWING SEASON
51, 52	5/28/04			Soybeans
55	4/20/04	Anhydrous ammonia (82% N) N-Serve	N - 180 lbs/acre N N-serve - 1 qt/acre
54	4/28/04	Urea (46% N)	N - 172 lbs/acre N
54, 55	5/10/04	19-18-0	N - 36 lbs/acre N	Yellow corn
56	4/16/04	19-18-0	N - 30.4 lbs/acre N	White waxy Japanese corn
57	4/17/04
56	5/18/04	Anhydrous ammonia (82% N)	N - 185 lbs/acre N
57			N - 185 lbs/acre N (westernmost 24 rows) N - 125 lbs/acre N (west-central 24 rows) N - 150 lbs/acre N (east-central 24 rows) N - 185 lbs/acre N (easternmost 24 rows)
55, 56	9/2/04			Rye
2005 GROWING SEASON
51	4/30/05	Anhydrous ammonia (82% N) N-serve	N - 150 lbs/acre N N-serve - 1 qt/acre
51, 52	5/3/05			Corn
52	6/8/05	Anhydrous ammonia (82% N)	N - 120 lbs/acre N (eastermost 96 rows) N - 150 lbs/acre N (west-central 48 rows) N - 180 lbs/acre N (westernmost 48 rows)
54, 55	6/4/05			Soybeans
56, 57	4/12/05			Corn
56	5/18/05	Anhydrous ammonia (82% N)	N - 185 lbs/acre N (6 easternmost rows immediately east of well) N - 150 lbs/acre N (12 east-central rows) N - 125 lbs/acre N (12 west-central rows) N - 185 lbs/acre N (12 westernmost rows)
57			N - 125 lbs/acre N (inside 150-ft radius around well) N - 185 lbs/acre N (elsewhere)
2006 GROWING SEASON
51, 52				Soybeans
54	5/5/06	Anhydrous ammonia (82% N)	N - 147 lbs/acre N
55	5/8/06	Agrotain (stabilized urea)(46% N)	N - 120 lbs/acre N
55	5/22/06	10-34-0	N - 13 lbs/acre N
54, 55	5/22/06			Corn
56, 57	5/7/06	19-18-0 Ammonium sulfate (21% N)	N - 62 lbs/acre N
56, 57				Green beans Soybeans
2007 GROWING SEASON
51, 52	5/5/07	Anhydrous ammonia (82% N) N-serve	N - 80 lbs/acre N (120 ft on either side of wells) N - 120 lbs/acre N (elsewhere) N-serve - 1 qt/acre
51, 52	5/10/07			Corn
54, 55	4/24/07 (green beans) 6/22/07 (soybeans)			Green beans Soybeans
54, 55	5/14/07	Ammonium sulfate (21% N)
56, 57	4/18/07			Corn
56	5/14/07	Anhydrous ammonia (82% N)	N - 80 lbs/acre N (outside 75-ft radius around well) N - 160 lbs/acre N (inside 75-ft radius around well)
57	5/14/07	Anhydrous ammonia (82% N)	N - 80 lbs/acre N (inside 75-ft radius around well) N - 160 lbs/acre N (outside 75-ft radius around well)

Fields 51, 52, 54, and 55 were tilled using a chisel plow and a seed-bed preparation tool. Nitrogen fertilizer was applied to these fields before planting by using a toolbar. Additional fertilizer was applied with the corn planter during planting. Fields 56 and 57 were also tilled with a chisel plow. Fertilizer was applied in the row with the corn planter during planting, and again later when the corn was 10 to 12 inches tall by using a toolbar.

Fields 54, 55, 56, and 57 (associated with Wells 54 through 57, respectively) were irrigated during summer growing seasons, as needed, using center-pivot irrigation systems whose water supply was derived by pumping from relatively deep levels (45 feet) of the same aquifer that was being studied. Irrigation was used in Fields 51 and 52 (associated with Wells 51 and 52, respectively) only during the 2007 growing season. The farmers used irrigation on as many as seven days per growing season, typically applying about 1 to 1.5 inches of water per irrigation event. The irrigation water contains as much as 9 mg/l of nitrate-nitrogen, so that as much as 21 lbs/acre N was recycled by applying irrigation water during the growing season.

A cooperating farmer had the practice of pasturing his cattle in Fields 51, 52, 54, and 55 during winter and spring months whenever corn residue was present from the preceding harvest. Consequently, cattle were present in Fields 54 and 55 in spring 2005 and spring 2007, and in Fields 51 and 52 in spring 2006. It was observed by the investigators--particularly in spring 2005--that the cattle tended to congregate around Wells 54 and 55, where they used the stickups of the wells as scratching posts. On that occasion, the wells were thoroughly purged and flushed before sampling, using water from nearby wells in the same aquifer. No cattle were ever pastured in Fields 56 and 57.

In order to obtain baseline information regarding soil characteristics, soil conductivity measurements within the test plots were made in early 2004, prior to spring planting. Maps of soil conductivity were provided by a participating farm-service company. Based on the results of these measurements, soil sampling to determine cation-exchange capacity was conducted by the farm-service companies. An example of the relationship between the locations of soil samples and the distribution of soil-conductivity measurements is shown in a map of soil samples around Wells 51 and 52. From the soil samples, maps were created showing soil pH and concentrations of phosphorous and potassium.

EXPERIMENTAL MANAGEMENT PRACTICES

During the period of the 4-year study, cooperating farmers implemented a variety of management practices within the test plots, including (1) planting of winter rye, (2) application of urea, as an alternative to anhydrous ammonia, (3) use of nitrification inhibitors (N-serve) and stabilized urea (Agrotain), and (4) variable rates and timing of fertilizer applications, both in discrete subsections of fields and as continuously variable applications by use of GPS-enabled variable-rate nitrogen applicators. The effects of variable rates of fertilizer application on crop yields were evaluated using GPS-enabled yield monitors on combines.

Fertilizers and Stabilizers. During the 2004 growing season, Field 57 was divided into four subsections, where different rates of fertilizer were applied, ranging from 125 to 185 lbs/acre N (Table 3). Well 57 was located within the subsection that received 125 lbs/acre N. Nitrification inhibitor (N-Serve) was applied to Field 55 (approximately 1 quart per acre).

During the 2005 growing season, Fields 52, 56, and 57 were divided into subsections that received different rates of fertilizer application, ranging from 120 to 185 lbs/acre N (Table 3). Wells 52, 56, and 57 were in subsections that received 180, 185, and 125 lbs/acre N, respectively. Nitrification inhibitor (N-Serve, approximately 1 quart per acre) was applied to Field 51.

During the 2006 growing season, anhydrous ammonia was applied to Field 54 using a variable-rate nitrogen applicator. The variable-rate applicator creates a map showing fertilizer distribution across the field; the average rate of application for the entire field was 147 lbs/acre N (Table 3). Fertilizer was applied to Field 55 in the form of stabilized urea (Agrotain) at a rate of 120 lbs/acre N.

During the 2007 growing season, Fields 51, 52, 56, and 57 were divided into subsections that received different rates of fertilizer application, ranging from 80 to 160 lbs/acre N (Table 3). Wells 51, 52, and 57 were in subsections that received 80 lbs/acre N. Well 56 was in a subsection that received 160 lbs/acre N. Nitrification inhibitor (N-Serve, approximately 1 quart per acre) was applied to Fields 51 and 52. Because of droughty conditions during the 2007 growing season, Fields 51 and 52 were irrigated on seven occasions, while Fields 56 and 57 were irrigated on four occasions.

Rye Seeding. Prior to harvest in 2004, aerial seeding of rye was conducted in Fields 55 and 56 on September 2, in order to determine if such a cover crop might reduce concentrations of nitrate-nitrogen in ground water by the uptake of nitrogen in the soil. Fields 54 and 57 were left unseeded as controls. Because of the long growing season in summer 2004, the rye was seeded prior to harvest in order to maximize its growth and nitrogen uptake before onset of winter dormancy and maximum recharge. The condition of the rye on October 21, 2004, can be seen in photographs of the ground cover. This was a management practice that was recommended by personnel of the cooperating farm-service agencies. However, because of the short growing season in autumn 2004, growth of the rye was less than 6 inches (with commensurately shallow root systems) when it was plowed under before the 2005 planting season.

EFFECTS ON CORN YIELDS

During the 2004 harvest, Fields 54 and 55, which had received fertilizer applications of 208 and 216 lbs/acre N, respectively, had corn yields of 179 and 186 bushels/acre, respectively (Table 4). Field 57 showed no variation in productivity of white waxy Japanese corn (165 bushels/acre), despite significantly different rates of nitrogen application (ranging from 125 to 185 lbs/acre N) within different subsections of the field.

TABLE 4. Crop Yields
Field or Field Subarea	Crop	Nitrogen application rate (lbs/acre N)	Yield (bushels/acre, dry)
2004 GROWING SEASON
54	Yellow Corn	208	179
55	Yellow Corn	216	186
56	White Corn	185	165
57 Westernmost 24 rows	White Corn	185	165
57 West-central 24 rows	White Corn	125	165
57 East-central 24 rows	White Corn	150	165
57 Easternmost 24 rows	White Corn	185	165
2005 GROWING SEASON
51, 52	Yellow Corn	150	Data lost - Combine fire
56 Westernmost 12 rows	Yellow Corn	185	205.3
56 West-central 12 rows	Yellow Corn	125	205.2
56 East-central 12 rows	Yellow Corn	150	205.2
56 Easternmost 6 rows	Yellow Corn	185	201.1
2006 GROWING SEASON
54, 55	Yellow Corn	147	151
2007 GROWING SEASON
51, 52	Yellow Corn	80	208
56 Inside 75-ft radius around the well	Yellow Corn	160	189.7
56 Outside 75-ft radius around the well	Yellow Corn	80	173.8
57 Inside 75-ft radius around the well	Yellow Corn	80	179.6
57 Outside 75-ft radius around the well	Yellow Corn	160	185.8

During the 2005 harvest, the use of electronic yield monitors indicated that different rates of fertilizer application in Field 56, varying from 125 to 185 lbs/acre N, had no measurable effects on corn yields, which were about 205 bushels per acre for all portions of the field (Table 4). Yield data for the fields associated with Wells 51 and 52 were lost because an equipment fire destroyed the combine harvester and its yield monitor before the data were downloaded.

During the 2006 harvest, Fields 54 and 55 had average corn yields of 152 bushels/acre (Table 4). Yield monitors were used to map variations in yield across the fields. The yield map of Fields 54 and 55 shows the effects on productivity of irrigation, with lower productivity being evident in areas beyond the reach of the center-point irrigation system, as well as local features such as the presence of a sandy hill of low relief, where coarser-grained soil was less productive.

During the 2006 growing season, several of the participating farmers conducted their own tests on corn fields outside the study area. In one test area, they varied nitrogen application rates from 80 to 150 lbs/acre N, but yields only varied from 152 to 162 bushels/acre. In another test area, they varied rates from 100 to 160 lbs/acre N, and yields only varied from 152 to 163 bushels/acre.

During the 2007 harvest, corn yields in Fields 56 and 57 varied from 174 to 190 bushels/acre (Table 4). In the subsections that had received fertilizer applications of only 80 lbs/acre N, the average yield was 177 bushels/acre, compared with an average yield of 188 bushels/acre in subsections that had received fertilizer applications of 160 lbs/acre N.

Also during the 2007 harvest, corn yields from Fields 51 and 52 averaged 208 bushels/acre; these were the highest average yields measured during the investigation, even though fertilizer applications in the vicinities of Wells 51 and 52 were only 80 lbs/acre N, and 120 lbs/acre elsewhere. In the yield map for Fields 51 and 52, the influence of the center-pivot irrigation system is apparent. In the immediate vicinity of the wells, however, the participating farmer observed that yields were approximately 20 bushels/acre less than adjoining areas where the application rate was 120 lbs/acre.

EFFECTS ON GROUND WATER

Well 53. In order to provide a control well, Well 53 was installed in an uncultivated woodlot. The well was located about 190 feet (58 m) south of Field 52, and the boundary between the woodlot and the cultivated field is a drainage ditch that is about 5 feet (1.5 m) deep. Throughout the investigation, the concentration of nitrate-nitrogen in the shallow ground water ranged from only 1.2 to 8.5 mg/l, never exceeding the Drinking Water Standard.

Wells 51 and 52. In 2003, which was the year preceding the commencement of the study, Fields 51 and 52 had been planted in corn. When ground-water samples were first collected from Wells 51 and 52 in March 2004, concentrations of nitrate-nitrogen exceeded the Drinking Water Standard of 10 mg/l (Fig. 3). But by May 2004, concentrations had fallen below that level and remained low throughout the 2004 growing season, when soybeans were cultivated, and the following winter.

Figure 3. Graphs showing concentrations of nitrate-nitrogen (mg/l) from monitoring wells, crops in associated test plots, the water level (feet above mean sea level) in Well 52, and total daily precipitation (inches of water). Blue and red triangles represent the application of nitrification inhibitor to the field associated with Well 51 and to the fields associated with Wells 52 and 55, respectively. Brown triangles indicate periods (between harvest and subsequent spring planting) during which cattle were sometimes pastured in fields associated with Wells 51, 52, 54, and 55.

In May 2005, corn was planted in Fields 51 and 52, and anhydrous ammonia was applied at a rate of about 150 lbs/acre N. A nitrification inhibitor (N-serve) was applied to Field 51 but not to Field 52. Subsequently, concentrations of nitrate-nitrogen in both Wells 51 and 52 rose to levels that were slightly above 10 mg/l, and they remained elevated throughout the summer (Fig. 3).

Cattle were occasionally pastured in Fields 51 and 52 in winter 2005 and spring 2006, and high concentrations of nitrate-nitrogen were detected in Wells 51 and 52 during the mid-winter sampling. Concentrations of nitrate-nitrogen generally remained at elevated levels through the 2006 growing season, even though soybeans were cultivated and no additional inorganic fertilizers were added (Fig. 3).

In May 2007, corn was again planted in Fields 51 and 52, but Wells 51 and 52 were included in subsections that received anhydrous ammonia at a rate of only 80 lbs/acre N. Also, nitrification inhibitor (N-serve) was applied to both fields. Subsequently, concentrations of nitrate-nitrogen in both Wells 51 and 52 actually declined slightly, so that the concentration in Well 52 was less than 10 mg/l (Fig. 3). Thus, the results provide evidence that the use of reduced fertilizer applications (80 lbs/acre N), together with the use of a nitrification inhibitor, may have been effective in reducing contamination of the ground water, even during a corn-growing season.

Wells 54 and 55.In 2003, Fields 54 and 55 had been planted in soybeans. Nevertheless, concentrations of nitrate-nitrogen in Wells 54 and 55 exceeded the Drinking Water Standard (10 mg/l) when the wells were first sampled in March 2004 (Fig. 3), prior to application of fertilizer.

When Fields 54 and 55 were planted in corn in May 2004, fertilizers were applied at a high rate of about 210 lbs/acre N. The fertilizer that was applied to Field 54 was in the form of urea, while that applied to Field 55 was in the form of anhydrous ammonia. Nitrification inhibitor (N-serve) was applied to Field 55 but not to Field 54. Subsequently, concentrations of nitrate-nitrogen in Well 55 actually declined slightly, while concentrations in Well 54 rose to very high levels (39 mg/l) (Fig. 3). These data indicate that the nitrogen stabilizer, when used in conjunction with anhydrous ammonia, may have been effective in reducing contamination of the ground water. On the other hand, the use of urea in Field 54 was associated with the highest levels of ground-water contamination ever observed during the study.

During spring 2005, the concentration of nitrate-nitrogen in Well 54, which had been very high during the previous growing season, fell to low levels, while concentrations in Well 55 rose to high levels (Fig. 3). Clearly, the rye that had been aerially seeded in Field 55 on September 2, 2004, was ineffective in reducing nitrate contamination of Well 55. Because of the lateness of the harvest that year, the rye may have been planted too late in the fall to produce any effect. More importantly, however, pastured cattle tended to congregate around Wells 54 and 55 in spring 2005 (as discussed above). The presence of the rye around Well 55 attracted the foraging cattle, and it is probable that their prolonged presence in that area might have contributed to the highly elevated concentrations of nitrate-nitrogen in Well 55.

Soybeans were planted in Fields 54 and 55 during the 2005 growing season, and concentrations of nitrate-nitrogen in the ground water from Wells 54 and 55 fell below 10 mg/l until spring 2006 (Fig. 3). Even though no cattle had been pastured in Field 55, an increase in the concentration of nitrate-nitrogen occurred in Well 55 (to more than 20 mg/l), prior to planting of corn and application of fertilizer.

Corn was planted in Fields 54 and 55 in May 2006. The fertilizer that was applied to Field 55 was primarily in the form of stabilized urea (Agrotain) at 120 lbs/acre N. A variable-rate applicator was used to apply anhydrous ammonia to Field 54, where the average application rate was 147 lbs/acre N. Throughout the growing season, concentrations of nitrate-nitrogen in both wells were only moderately elevated (Fig. 3). Thus, the use of stabilized urea was associated with significantly lower levels of contamination than was the earlier use of urea without any nitrification inhibitor.

In 2007, green beans were planted in Fields 54 and 55, and concentrations of nitrate-nitrogen in Wells 54 and 55 fluctuated around 15 mg/l.

Wells 56 and 57. In 2003, Fields 56 and 57 had been planted in soybeans. As with Wells 54 and 55, however, concentrations of nitrate-nitrogen in Wells 56 and 57 exceeded the Drinking Water Standard (10 mg/l) when the wells were first sampled in March 2004 (Fig. 3).

In 2004, Field 56 received anhydrous ammonia at a rate of 185 lbs/acre N, while various portions of Field 57 received rates that ranged from 125 to 185 lbs/acre N, with an overall average application rate of 160 lbs/acre N. The difference in the overall application rates of fertilizers (185 versus 160 lbs/acre N) did not result in any obvious difference in the concentration of nitrate-nitrogen in Wells 56 and 57 (Fig. 3). Concentrations of nitrate-nitrogen in both wells declined through winter 2004 and spring 2005. Well 56, where rye had been planted and where growth of the cover crop was relatively vigorous--and where no cattle had ever been pastured--did experience a greater decrease in contamination during winter 2004 and spring 2005.

In 2005, Fields 56 and 57 were again planted in corn, and anhydrous ammonia was applied at rates varying from 125 to 180 lbs/acre N. Concentrations of nitrate-nitrogen experienced a spike in Well 56 shortly after the application of fertilizer, and concentrations in both wells rose throughout the growing season, then tended to decline during winter 2005 and spring 2006. Concentrations in Well 57, which was located in a subsection of Field 57 that received a fertilizer application of only 125 lbs/acre N, actually exceeded the concentration in Well 56, which was located in a subsection of Field 56 that received from 150 to 185 lbs/acre N.

In 2006, green beans were planted in Fields 56 and 57, which received applications of anhydrous ammonia at a rate of 62 lbs/acre N. By November 2006, concentrations of nitrate-nitrogen in Wells 56 and 57 were relatively high (21 and 22 mg/l, respectively) (Fig. 3).

A large spike in the concentration of nitrate-nitrogen occurred in Well 56--but not in Well 57--in late spring 2006. Cattle were never pastured in Field 56, so that no explanation for this spike in concentration is apparent.

In 2007, Fields 56 and 57 were again planted in corn, and anhydrous ammonia was applied at rates of 80 and 160 lbs/acre N, with the smaller amount being applied in the vicinity of Well 57 and the greater amount being applied in the vicinity of Well 56. Subsequently, the concentration of nitrate-nitrogen rose to high levels in Well 56, while the concentration in Well 57 barely exceeded the Drinking Water Standard. This response in Well 57 is consistent with the response observed during the same period in Wells 51 and 52, where anhydrous ammonia had also been applied at the greatly reduced rate of 80 lbs/acre N, and where the concentrations of nitrate-nitrogen remained low, even at the beginning of the corn-growing season.

SUMMARY

Cultivation of corn and the concomitant application of relatively large quantities of nitrogen fertilizers (150 to 185 lbs/acre N) are associated with elevated concentrations (above 10 mg/l) of nitrate-nitrogen in shallow ground water (upper 18 inches) for periods of 6 to 18 months following the application.

As the investigation progressed, participating farmers systematically reduced their application rates of nitrogen fertilizers. During the final year of the study, fertilizer application rates in Fields 51, 52, and 57 were reduced to only 80 lbs/acre N in subsections of the fields that were in the immediate vicinity of monitoring wells. A nitrification inhibitor was also applied in Fields 51 and 52. However, because of droughty conditions during the 2007 growing season, additional nitrate was supplied to the fields through nitrate-bearing ground water that was recycled through the irrigation system. In post-application sampling, the concentrations of nitrate-nitrogen in Wells 51, 52, and 57 exhibited only very small increases, or even decreases. The application rate of only 80 lbs/acre N was associated with decreases of corn yields of approximately 10 to 20 bushels/acre in Field 57 and in Fields 51 and 52, respectively. While it is unlikely that the participating farmers would consider any further reductions in fertilizer application rates, the long-term use of a rate as low as 80 lbs/acre, particularly when used with nitrification inhibitors, could potentially result in significant long-term improvement of the aquifer.

Compared with anhydrous ammonia, nitrification and leaching of urea may occur more rapidly, so that urea may introduce more nitrate into the shallow ground water more quickly. But compared with urea, anhydrous ammonia may result in levels of nitrate-nitrogen that exceed the Drinking Water Standard for longer periods. Compared with urea, stabilized urea may have less negative effects on ground water, effects that may be more similar to anhydrous ammonia.

Because horizontal flow in the upper part of the aquifer is slow and vertical recharge is rapid and direct, it is possible that domestic drinking-water supplies could be protected by establishing wellhead protection areas around shallow wells (15 to 25 feet deep) that extend as little as 200 feet up-gradient of such wells. The exact size of such protection areas would depend on factors such as well depth and maximum production rate. Also, such wellhead protection areas would need to eliminate all sources of contaminants, including farm animals and septic systems, as well as fertilizer applications.