RESULTS FROM THE BIG SPRING BASIN WATER QUALITY MONITORING AND DEMONSTRATION PROJECTS
by Robert D. Rowden, Huaibao Liu and Robert D. Libra
The agricultural practices, hydrology, and water quality of the 267 km2 Big Spring groundwater basin in Clayton County, Iowa, have been investigated since 1981. Landuse is dominated by corn and alfalfa production, along with numerous small livestock operations; nitrate-nitrogen (-N) and herbicides are the resulting contaminants in ground- and surface water. The Ordovician Galena Group carbonates are the main aquifer in the basin. Recharge to this shallow, moderately karsted aquifer is dominantly by infiltration, augmented by sinkhole-captured runoff. Groundwater is discharged to the surface at Big Spring, where the quantity and quality of the discharge is monitored.
Monitoring has shown a three-fold increase in groundwater nitrate-N concentrations from the 1960s to the early 1980s, following a similar increase in nitrogen fertilizer applications. The nitrate-N discharged from the basin by ground- and surface water typically is equivalent to over one-third of the nitrogen fertilizer applied, with larger losses and greater concentrations occurring during wetter years. Atrazine is present in the groundwater year round. Contaminant concentrations in the groundwater respond directly to recharge events, and the unique chemical signatures of infiltration versus runoff recharge are detectable in the discharge from Big Spring.Education and demonstration efforts have decreased pesticide use and have reduced nitrogen fertilizer application rates by one-third since 1981, while crop yields have been maintained. Relating the declines in nitrogen and pesticide inputs to nitrate and pesticide concentrations at Big Spring is problematic, and confounded by year-to-year variability in recharge, which strongly affects concentrations. Annual recharge, as inferred by discharge from Big Spring, has varied five-fold during the monitoring period, overshadowing any water quality improvements resulting from incrementally decreased inputs.
The agricultural practices, hydrology, and water quality of the Big Spring basin (Figure 1), a 267 km2 groundwater basin in Clayton County, Iowa, have been studied by the Iowa Department of Natural Resources-Geological Survey Bureau (GSB) and cooperators since 1981 (Hallberg et al., 1983, 1984, 1985, 1986, 1987, 1989; Libra et al., 1986, 1987, 1991; Rowden et al., 1993, 1995, 1998, 2000; Rowden 1995; Liu et al., 1997). Historic water-quality data have shown regional increases in nitrate-N in the groundwater of the basin paralleling a three-fold increase in nitrogen fertilizer use from the mid-1960s to the early 1980s (Figure 2). A network of sites, including tile lines, streams, springs, and wells, was established to monitor water-quality changes accompanying changes in farm management. The network was designed in a nested fashion, from small-scale field plots to the basin groundwater and surface-water outlets (Littke and Hallberg, 1991). Water samples from as many as 50 sites have been analyzed for various forms of nitrate, herbicides, and other parameters on a weekly to monthly basis. Key sites were instrumented for continuous or event-related measurement of water discharge and chemistry and for automated sample collection. The development of monitoring sites within the basin has been a cooperative effort among the GSB, the U.S. Geological Survey, Iowa State University, the U.S. Department of Agriculture-Natural Resource Conservation Service, and the U.S. Environmental Protection Agency. Water-quality analyses were performed by the University of Iowa Hygienic Laboratory, using standard methods and an U.S.EPA-approved quality assurance/quality control plan.
Landuse within the basin is about 97% agricultural. The basin includes about 200 farms with an average size of about 330 acres. Small dairy and hog operations are common. Typically 50% of the basin area is planted to corn, 35-40% to alfalfa, and 10% of the basin is pasture. Purchased fertilizer, manure and herbicides are applied to corn primarily during the spring planting period. There are no significant urban or industrial point sources within the basin that impact groundwater quality. These conditions along with the ability to gage the volume of groundwater passing through the Galena aquifer allow unambiguous study of the agricultural ecosystem. By surveying farmers annually for application rates of pesticides and fertilizers, and monitoring the water quality and discharge of surface water and groundwater in the basin, the mass flux of nutrients and chemicals applied within the basin can be quantified, allowing assessment of chemical balances on a basin-wide scale.Nitrate-N, and the most commonly used corn herbicides, particularly atrazine, are the major agricultural contaminants. Initial investigations in the area (Hallberg et al., 1983, 1984; Libra et al., 1986) showed atrazine is present (>0.1 µg/L) year-round in surface- and groundwater, except during extended dry periods. Detectable concentrations of several other herbicides generally occur following the spring application period, but are also present year-round following runoff events. The amount of nitrate-N discharged from the basin by groundwater and surface water typically is equivalent to one third of that applied as fertilizer, and during wetter years, exceeds one half of that applied. In-stream denitrification and nitrogen uptake also occur before surface waters exit the basin, suggesting that even more nitrogen is lost from agricultural practices (Crumpton and Isenhart, 1987). In an effort to reduce these losses, a multi-agency group initiated the Big Spring Basin Demonstration Project, which was funded by the Iowa Groundwater Protection Act and conducted from 1986-1992. The project integrated public education with on-farm research and demonstration projects that stressed and monitored the environmental and economic benefits of efficient chemical management, with a particular focus on nitrogen management. Education and demonstration activities continued from 1992-1999 as part of the Northeast Iowa Demonstration Project, under the direction of the Iowa State University Extension Service.
The results of the Big Spring Project have become an integral part of the states water quality monitoring program. In 1995 the U.S. General Accounting Office selected the Big Spring Project as one of nine (from a nation-wide field of 618) particularly innovative and successful nonpoint-source pollution prevention efforts. In this article we briefly discuss the Big Spring basin's hydrogeologic system, the response of the system to recharge, and the results of 18 years of hydrologic and nonpoint-source contaminant monitoring in a responsive hydrogeologic environment.
Northeast Iowa is characterized by a midcontinental subhumid climate. Mean annual precipitation (based on the period 1951-1980) is about 84 cm, with 70% of the annual precipitation occurring during the April-September growing season. The mean annual temperature is 6.7o C, with a winter average of -5.6o C and a summer average of 22.2o C.
The oldest rocks exposed in the Big Spring basin are the carbonates of the Ordovician Galena Group (Figure 1). These strata, which have an uneroded thickness of about 76 m, generally dip to the southwest at about 5 m/km and form the major aquifer used by rural residents in the area. The Galena outcrops low on the landscape in the central and eastern parts of the basin. Big Spring, which is the largest spring in Iowa, discharges near the base of the aquifer in the valley of the Turkey River at a state owned fish hatchery. The Galena aquifer is underlain by shales and shaley carbonates of the Decorah, Platteville, and Glenwood Formations, which hydrologically isolate it from underlying aquifers. Across most of the basin, the Galena Group is overlain by the Maquoketa Formation. The lower part of the Maquoketa consists of silty and shaley carbonates with an uneroded thickness of 23 to 30 m. These strata are not a barrier to groundwater flow and are hydrologically connected with the Galena aquifer. The lower Maquoketa forms the uppermost bedrock over much of the central part of the basin. In the western 20% of the basin, claystones and shales of the upper Maquoketa Formation overlie the lower Maquoketa strata, and act as an effective confining bed with a thickness of up to 30 m. Along the western margin of the basin, outliers of Silurian rock overlie a complete section of the upper Maquoketa strata.
The bedrock in the basin is mantled by thin (typically less than 5 m) Quaternary deposits and the rolling landscape is generally controlled by the relief of the bedrock surface. Thin, eroded remnants of Pre-Illinoian glacial till occur on the uppermost drainage divides. The uplands and hillslopes are draped with 2 to 8 m of Wisconsinan Peoria loess, or loess-derived deposits, and loamy alluvium occurs in stream valleys and drainageways. The Quaternary materials within the basin are not an effective barrier to groundwater flow.
The carbonates of the Galena and lowermost Maquoketa are fractured and exhibit moderate karst development where they form the uppermost bedrock. Figure 1 shows locations of sinkholes within the basin generally occurring within the Galena outcrop belt, and near the Galena-Maquoketa contact, where only a limited thickness of the Maquoketa is present. About 10% of the surface area of the basin drains to sinkholes, the majority of which are soil-filled. The individual sinkhole drainage basins are generally less than 2 km2 in area. Within the Big Spring basin, the aquifer has both diffuse- and conduit-flow systems (Hallberg et al., 1983), using the terminology of White (1969, 1977). The diffuse-flow system is recharged by slow infiltration through overlying materials into joints, fractures, and bedding planes that have experienced little or moderate solutional modification. The diffuse-flow parts of the aquifer, along with the overlying shales and Quaternary materials, conceptually form a relatively low transmissivity-high storage part of the groundwater system (Hallberg et al., 1989). Groundwater from this less transmissive part of the system discharges to the more solutionally-modified conduit-flow system, and sustains the discharge from Big Spring during periods of low recharge. The conduit-flow system is directly recharged by diversion of surface water into sinkholes. Application of hydrograph separation techniques (Singh and Stall, 1971; Gustard et al., 1992) to the Big Spring discharge record suggests that infiltration to the less transmissive parts of the groundwater system accounts for 75% to 95% of the recharge to the Galena aquifer on an annual basis.
The extent of the groundwater basin was defined by mapping the potentiometric surface of the Galena aquifer, dye tracing via sinkholes, and gaging gaining and losing stream reaches (Hallberg et al., 1983). A water distribution system at the hatchery allows the spring's discharge to be monitored and stream gaging stations within the basin allow monitoring of surface-water discharge. Over 85% of the groundwater discharged from the basin flows through Big Spring. Surface water is discharged by various streams, but dominantly by the Roberts Creek watershed, which accounts for 65% of the basin's surface area and about 75-80% of the surface-water flow leaving the basin. Typical base-flow discharge rates for Big Spring are about 0.8 cubic meters/second (cms), with peak discharge rates of over 6.0 cms occurring one to two days after rainfall or snowmelt events. Typical base-flow discharge for Roberts Creek is about 0.7 cms.
Figure 3 shows an east-west cross section within the southern part of the basin. In western portions of the basin, the aquifer is confined and the potentiometric surface is above the top of the Galena. Across much of the remainder of the basin, the aquifer is unconfined, and a variable thickness of the aquifer is unsaturated. There are two north-south trending troughs in the potentiometric surface of the aquifer which head in sinkhole areas in the north central and northeastern portions of the basin and converge at Big Spring. These troughs reflect the presence of highly transmissive conduit-zones, where solutional activity has enlarged fractures and bedding planes, increasing the permeability of the rocks (Hallberg et al., 1983). These conduit-zones transmit groundwater to the discharge point at Big Spring, and act as drains for the diffuse-flow parts of the aquifer. In the western part of the basin, where the Galena is overlain by a relatively thick Maquoketa sequence, solutional activity has been limited and the Galena is generally less permeable than it is in the central and eastern parts of the basin.
The basins main surface drainages head in the north and western parts of the basin (Figure 1). These streams gain groundwater in their upper reaches, where a relatively thick sequence of Maquoketa Formation and Quaternary materials is present. Here, these units are saturated, and the water table is above the Galena aquifer, near the land surface and graded to the streams. Downstream, the Maquoketa-Quaternary sequence has been erosionally thinned and the Galena-lowermost Maquoketa strata are more permeable. Here, the water table is located within the aquifer, and surface drainages lose water to the aquifer (Hallberg et al., 1983). Although these streams lose water, through the central and eastern portions of the basin, they maintain perennial flows except during extended dry periods. This is possible when recharge provided by shallow groundwater (including tile drainage) in the streams headwaters is greater than leakage into the groundwater system downstream. Relatively fine-grained materials in the alluvial deposits retard downward leakage through the streambeds (Hallberg et al., 1983; Rowden and Libra, 1990).
The Big Spring groundwater system receives both infiltration and runoff recharge, which have unique chemical signatures that can be tracked through the nested monitoring network, from the water table beneath individual fields to the basins surface- and ground-water outlets (Hallberg et al., 1983, 1984). Infiltration recharge is enriched in nitrate and other chemicals that are mobile in soil. Runoff recharge has lower concentrations of such compounds, but is enriched in herbicides and other chemicals with low soil mobility. The concentrations of organic- and ammonium-N tend to increase directly with suspended sediment concentration during runoff events. Typically Big Spring yields groundwater delivered through infiltration, but following significant precipitation or snowmelt, sinkholes within the basin may direct surface runoff into the aquifer, mixing it with the groundwater. As this runoff recharge moves through the groundwater system and discharges from Big Spring, relatively low nitrate and high herbicide concentrations occur during peak discharge periods. This is typically followed by higher nitrate and lower herbicide concentrations as the associated infiltration recharge moves through the hydrologic system.
During prolonged recession periods, nitrate and atrazine concentrations generally show a slow, steady decline. This decline likely occurs as an increasing percentage of the discharge is relatively older groundwater from the more diffuse-flow parts of the flow system (Hallberg et al., 1984).
Figure 4 shows discharge, nitrate and atrazine concentrations for Big Spring during March 5-13, 1990 (Rowden et al., 1993). On March 8, 1990, precipitation totaling about 20 mm and warming temperatures generated a snowmelt event at Big Spring. Three days before the event, the groundwater discharge was 0.9 cms, the nitrate-N concentration was 4.9 mg/L and the atrazine concentration was 0.82 µg/L. On March 8 at 17:30 the discharge was 4.9 cms, the nitrate-N concentration was 3.8 mg/L, and the atrazine concentration was 1.0 µg/L. At this time, little change in chemical concentrations had occurred, indicating that most of the water discharged during the rising limb of the hydrograph was pre-event water. This is typical of large recharge events at Big Spring, where the arrival of event-water is nearly coincident with peak discharge (Hallberg et al., 1984). The displacement of large volumes of pre-event water is a common occurrence at many karst springs (Atkinson, 1977; Driess, 1989). Discharge peaked at 7.3 cms at 20:30, while the maximum measured atrazine concentration, 7.8 µg/l, occurred earlier at 19:15, and the minimum measured nitrate-N concentration, 1.3 mg/L, occurred at 01:00 the following morning. By 07:00, the discharge had declined to 2.8 cms and the nitrate-N concentration had increased to 3.3 mg/L. On March 10 at 08:15, the groundwater discharge was 0.9 cms, the nitrate-N concentration had increased to 3.8 mg/L and the atrazine concentration had decreased to 2.7 µg/L. The slowly declining discharge following this and a smaller event on March 11 was derived from infiltration recharge that occurred broadly across the groundwater basin. Note that concentrations of atrazine and nitrate-N in the infiltration-derived recharge following the March 11 event were greater than the concentrations prior to the March 8 recharge event. Ultimately, the infiltration recharge component supplies the bulk of the water and chemicals discharging from Big Spring (Libra et al., 1986).
Results of Long-Term Monitoring
The average fertilizer-N rate on all corn rotations within the Big Spring basin was reduced from 79 kg/acre in 1981 to 52 kg/acre in 1993, a 34% decrease with no decline in corn yields (Table 1; Rowden, 1995); this saved basin producers about $360,000 annually. The decreased inputs amounted to almost 907,200 kilograms of nitrogen per year, which is the energy equivalent of 1,570,775 liters of diesel fuel. In addition, an estimated 41% of the basin producers reduced herbicide rates, and 34% reduced insecticide rates over a three-year period (Iowa State University-Cooperative Extension, unpublished data).
During 1983, a Payment-in-Kind (PIK) set aside program provided the opportunity to evaluate the results of a one-year reduction in nitrogen applications of about 40% in the basin. Statistical analysis of discharge and nitrate concentrations at Big Spring suggests the decline in nitrate concentrations during 1985 was related to the reduction in nitrogen inputs during 1983 (Figure 2). Monitoring in the basin shows that time lags occur between the implementation of management practices and responses in water quality at monitoring sites. While fertilizer use has declined by one-third since 1981, the effects of these incremental reductions have been obscured by year-to-year variations in groundwater discharge resulting from climatic variability. The driest consecutive two-year period in Iowa's history, water years (WYs; October 1 through September 30) 1988 and 1989, preceded the wettest consecutive two-year period since the project's inception. Relating changes in pesticide use to water quality changes in the Big Spring basin has also been difficult. Annual atrazine concentrations and loads appear to have declined significantly, but this trend has only become apparent during the last few years of monitoring.
Annual groundwater discharge for the monitoring period varied from 15.6 millions of cubic meters (mcm) in WY 1989 to 71.7 mcm in WY 1993, and averaged 39.2 mcm. Recharge in the basin, and ultimately discharge from Big Spring, is a function of the amount, timing, and intensity of precipitation. On an annual basis, discharge generally follows precipitation trends, although time lags are evident. The drought conditions in WY 1988 did not cause a decline in the annual discharge, relative to the preceding year. This indicates there was sufficient groundwater within the less transmissive parts of the groundwater system to sustain discharge through WY 1988. Other lags are apparent following the drought. While precipitation in WY 1990 was 12.7 cm above average and 34.4 cm greater than that for WY 1989, discharge showed only a modest increase. This suggests that much of the potential recharge during WY 1990 went towards replenishing soil-moisture that was depleted during the drought. While WY 1993 had the second-greatest annual precipitation during the monitoring period, it had by far, the greatest annual groundwater discharge recorded at Big Spring following the wet WYs 1990-1992. Decreases in groundwater discharge as precipitation increased during WYs 1996 and 1997 suggest that some of the recharge during these years was held in storage, eliminating soil-moisture deficits incurred during WYs 1994-1996. In WY 1999, discharge increased as precipitation decreased, suggesting that groundwater within the less transmissive parts of the groundwater system was being released and supplementing the discharge. During the 18-year period, discharge accounted for 16% of precipitation, and ranged from 8% to 23% of precipitation for individual years.
On an annual basis, nitrate concentrations and nitrate-N discharge parallel the overall volume of water moving through the soil and into the groundwater system. From WY 1982 to WY 1989, annual fw mean nitrate-N concentrations at Big Spring declined from 8.7 to 5.7 mg/L, and nitrate-N loads decreased from 396,000 to 88,403 kg, while annual discharge declined from 46.0 to 15.6 mcm. While some of the decrease in nitrate concentrations may be due to reductions in application rates, this response cannot be separated from the decrease caused by the decline in water-flux through the hydrologic system. From WY 1990 to WY 1991 annual discharge increased from 21.5 to 52.4 mcm while the fw mean nitrate-N concentration increased from 8.2 to 12.5 mg/L and nitrate-N loads increased from 176,181 to 655,558 kg. These increases resulted from both the increased volume of water passing through the soil and groundwater system, and the leaching of unutilized nitrogen left over from the drought years. Any improvements in water quality that may have resulted from the incremental reduction of nitrogen applications were obscured by variations in groundwater flux. From WY 1992 to WY 1993, discharge increased from 46.0 to 71.7 mcm, and the nitrate-N load increased from 553,333 to 814,518 kg, while the nitrate-N concentration decreased from 12.0 to 11.4 mg/L. The decrease in concentration was probably related to having a greater than normal proportion of runoff composing the annual discharge, as well as having less nitrogen available due to increased leaching of nitrogen during WYs 1991-1992. Groundwater discharge and nitrate-N loads declined from WY 1993 to WY 1997. Annual fw mean nitrate-N concentrations decreased from 10.4 mg/L in WY 1994 to 10.1 mg/L in WY 1995, then increased to 10.3 mg/L in WY 1996. This increase may be related to limited runoff during the year leading to an increased proportion of infiltration- versus runoff-recharge constituting the annual discharge. Following the largest precipitation event of WY 1996 in late June, discharge receded and nitrate and atrazine concentrations both increased and remained at elevated levels through August. From WY 1997 to WY 1998, discharge increased from 28.2 to 44.0 mcm, the fw nitrate-N concentration increased from 9.7 to 12.5 mg/L and the nitrate-N load increased from 273,313 to 549,866 kg. During WY 1999, discharge increased to 45.8 mcm and the annual fw nitrate-N concentration and load decreased to 11.8 mg/L and 539,478 kg. Some of the decrease in nitrate-N concentration and load in WY 1999 may be related to increased leaching of nitrogen during WY 1998.
Atrazine has been detected in 86% of Big Spring groundwater samples since monitoring began in WY 1982. The next most frequently detected pesticides at Big Spring during the monitoring period include cyanazine at 11%, alachlor at 9%, acetochlor at 6% and metolachlor at 4%. Unlike nitrate-N, annual fw mean atrazine concentrations and loads do not increase and decrease with the water-flux through the Big Spring hydrologic system (Figure 5). In WY 1993, groundwater discharge increased to record levels while the annual atrazine concentration and load showed only minor increases.
Annual atrazine concentrations increased from 0.3 to 0.7 µg/L from WY 1982 to WY 1985, and loads increased from 6.4 to 21.6 kg, as discharge declined. The increases in atrazine concentrations and loads from WY 1988-1991, and general decreases from WY 1991-1995 were probably related to changes in the timing and intensity of rainfall and in the relative proportion of infiltration- versus runoff recharge composing Big Spring's discharge. Another factor influencing annual concentrations and loads may be pesticide degradation rates, which vary with environmental factors, such as soil moisture (USEPA, 1986). The increases following the drought may be due to the mobilization of atrazine that did not degrade during dry conditions, and the decreases following WY 1991 may reflect the smaller than normal mass of herbicide available for mobilization to groundwater, due to enhanced hydrolysis and microbial activity during the wet WY 1990-1991 period. The increase in atrazine concentrations and loads in WY 1996 were, in part, due to a large runoff event that occurred in late June, not long after most atrazine application occurs in the basin. June accounted for 19% of the groundwater discharge and 58% of atrazine discharge for WY 1996. In WY 1999, a discharge event on May 17 set a new record for instantaneous groundwater discharge from Big Spring. The daily discharge accounted for 2.6% of the annual groundwater total and 13.3% of the annual atrazine load. The month of May accounted for 20% of the annual groundwater discharge and 44% of the annual atrazine load during WY 1999.
Since atrazine concentrations increase and decrease along with runoff, the calculation of accurate fw mean atrazine concentrations and loads requires both routine and event-related sampling. The lack of event samples during WYs 1993-1999, as well as a change from weekly to biweekly sampling during WYs 1998 and 1999, due to funding cuts, probably reduced fw mean atrazine concentrations and loads during the last seven years of monitoring. The lower atrazine concentrations and loads may also be related to changes in application rates. In recent years there has been a tendency to reduce the number of pounds of atrazine applied per acre, and use it in lower concentrations, mixed with other pesticides over a larger area. The replacement of atrazine by other herbicides may also be leading to declining atrazine concentrations and loads. Acetochlor was added to the list of pesticide analyses for the Big Spring Project in August 1994. During August and September of WY 1994, none of the weekly samples from Big Spring contained detectable (>0.10 µg/L) acetochlor. In WY 1995, 2% of 52 samples contained detectable acetochlor, in WY 1996, 4% of 52 samples contained acetochlor and in WY 1997, 6% of 53 samples contained acetochlor. In WY 1998, 4% of 24 biweekly samples contained acetochlor and in WY 1999, 17% of 24 samples contained acetochlor. The increasing number of acetochlor detections through time suggests that its use within the basin is increasing.In January 1993, two atrazine metabolites, desethylatrazine and desisopropylatrazine, were added to the list of pesticide analyses for the Big Spring Project. Desethylatrazine was detected (>0.10 µg/L) in 72% of the samples, while desisopropylatrazine was not detected in samples from Big Spring during WYs 1993-1999. The large number of detections of desethylatrazine, along with the absence of desisopropylatrazine during the monitoring period at Big Spring (Figure 6) and other sites, support research that has shown desethylatrazine to be the more stable initial degradation product (Adams and Thurman, 1991; Geller, 1980). Desethylatrazine has shown trends similar to atrazine at both surface water and groundwater monitoring sites within the basin (Rowden et al., 1993, Liu et al., 1997). In WY 1993, desethylatrazine was detected in 96% of the 23 weekly samples taken at Big Spring during the latter half of the year. Detections of desethylatrazine at Big Spring declined along with groundwater discharge from 98% in WY 1994, to 88% in WY 1995, to 62% in WY 1996 and to 32% of the weekly samples in WY 1997. In WY 1998, desethylatrazine was detected in 58% of the 24 samples taken during the year and in WY 1999, desethylatrazine was detected in 83% of the biweekly samples from Big Spring. Since the number of detections decreased during drier years and increased during wetter years, this may support the theory that pesticide degradation rates vary with environmental factors, such as soil moisture. Desethylatrazine has been detected year round at Big spring, except in WY 1997 when it was not detected during April, October, January, and February, and in WY 1998 when it was not detected from November through March. At most surface water sites, concentrations of desethylatrazine were usually greater during May through July and at most groundwater 999sites, concentrations appeared to be more related to discharge fluctuations than seasonal variations. These differences may result from differences in biotic degradation processes and/or degradation rates between groundwater and surface water. Since high concentrations of desethylatrazine were detected in both surface- and groundwater in the Big Spring area, desethylatrazine may also be produced by abiotic degradation.
Analysis of annual groundwater data from Big Spring for WYs 1982-1999 indicates that fw mean nitrate-N concentrations and loads generally parallel groundwater discharge, and fw mean atrazine concentrations and loads do not. The affects of incrementally decreased nitrogen loading within the basin of over 30% have been overshadowed by three and four-fold changes in groundwater discharge during the monitoring period. Water Year 1993 was the first year that the annual nitrate-N concentration decreased as annual discharge increased, WY 1996 was the first year that the nitrate-N concentration increased as discharge decreased, and WY 1999 was the first year that both the nitrate-N concentration and load decreased as discharge increased. It is possible that the gradual reductions in nitrogen applied within the basin are affecting changes in the water quality of Big Spring, but the declines in nitrate concentrations during WYs 1993 and 1999 may also be due to increased leaching of nitrogen during preceding years. The general decline in nitrate-N concentrations and loads from WY 1993-1997 cannot be separated from decreases caused by declines in groundwater-flux through the basins hydrologic system.
The Big Spring hydrologic system receives both infiltration and runoff recharge, which have unique chemical signatures that can be tracked through the nested monitoring network, from the soil zone beneath individual fields to the basin water outlets. The pronounced short-term changes in nitrate and atrazine concentrations are responses to significant recharge events. The concentration changes at the larger watershed scales are not as great or immediate as changes at smaller scale monitoring sites, although they clearly occur. The incremental reductions in nitrogen fertilizer and pesticide use in the basin may not result in pronounced water-quality changes, but they will be detectable over time. At watershed scales such as Big Spring, many landuse and management practices are integrated, and water-quality responses are dampened and complicated by climatic variations, storage effects, and biochemical processing in both surface-water and groundwater systems.
The mass of nitrate-N leaving the basin annually in groundwater from Big Spring and surface water from Roberts Creek has varied 15-fold, from 96 thousand kilograms in WY 1989 following drought conditions to 1.4 million kilograms in WY 1993 following wet conditions. The nitrate-N losses are equivalent to 5% and 79%, respectively, of the chemical nitrogen fertilizer applied during those years. The long-term average loss is equivalent to 40% of the chemical fertilizer applied. Additional biochemical losses of nitrate-N occur within streams before they exit the basin, indicating the actual N-losses from fertilized fields are greater than those measured. The nitrate-N discharged by the basins hydrologic system enters the Turkey River and becomes part of Iowas nitrate-N contribution to the Mississippi River, and, ultimately, to the Gulf of Mexico. Since 1982, annual losses from the Turkey River basin area have also varied 15-fold, from less than 1 kg/acre during the drought of 1989 to around 14 kg/acre during the relatively wet years following the drought.
The results from Big Spring show that large-scale environmental monitoring projects require long-term commitment. Producers are cautious when decreasing chemical inputs, and water quality in groundwater- and surface watersheds responds slowly to incremental changes in inputs. Time lags in water-quality responses caused by antecedent hydrologic conditions have also obscured relationships between changes in management practices and water quality responses. Annual nitrate concentrations generally declined from WY 1993 through WY 1997, but have increased in WYs 1998 and 1999. Annual atrazine concentrations have declined significantly during the last eight years of monitoring, but this may result from the replacement of atrazine by other pesticides and changes in the way that atrazine is applied, as well as significant reductions in the number of pesticide samples collected in recent years.
The Big Spring Project of the Iowa Department of Natural Resources has been supported, in part, through the Iowa Groundwater Protection Act and Petroleum Violation Escrow accounts, and other sponsoring agencies: the U.S. Environmental Protection Agency, Region VII, Kansas City, Nonpoint Source Programs, the USDA-Natural Resource Conservation Service, the Iowa State University Cooperative Extension Service, the University of Iowa Hygienic Laboratory (UHL), and the Iowa Department of Agriculture and Land Stewardship, Division of Soil Conservation. Major funding through Oil Overcharge funds, ended at the end of state fiscal year (FY) 1997. Section 319 funds and in-kind contributions from UHL and the U.S. Geological Survey supported the project through FY 1999. Significant changes in agricultural management have occurred, and the environmental benefits of these changes will continue. During the last five years the project has been in a scaled-back, long-term trend assessment phase. For WY 2000, Big Spring and Roberts Creek will be sampled on a weekly to biweekly basis. The long-term record from Big Spring has been, and should continue to be, an integral part of the states water quality monitoring program. Continued monitoring will provide a better understanding of water-quality improvements resulting from changes in agricultural practices.
Adams, C.D. and Thurman, E.M., 1991, Formation and transport of desethylatrazine in the soil and vadose zone: Journal of Environmental Quality, v. 20, no. 3, p. 540-547.
Atkinson, T.C., 1977, Diffuse flow and conduit flow in limestone terrain in the Mendip Hills, Somerset (Great Britain): Jour. Hydrol., v. 35, p. 93-110.
Crumpton, W.G. and Isenhart, T.M., 1987, Nitrogen balance in a stream ecosystem: the role of algal assimilation: Jour. of Water Poll. Control Fed., v. 50, p. 2043-2046.
Dreiss S.J., 1989, Regional scale transport in a karst aquifer 1. Component seperation of spring flow hydrographs. Water Resources Research 25 (1): 117-125.
Geller, A., 1980, Studies on the degradation of atrazine by bacterial communities enriched from various biotypes: Archives of Environmental Contamination and Toxicology, v. 9, no. 3, p. 289-305.
Gustard, A., Bullock, A., and Dixon, J.M., 1992, Low flow estimation in the United Kingdom: Institute of Hydrology Report no. 108, Wallingford, U.K., 88 p.
Hallberg, G.R., Hoyer, B.E., Bettis, E.A., III, and Libra, R.D., 1983, Hydrogeology, water quality, and land management in the Big Spring basin, Clayton County, Iowa: Iowa Geological Survey, Open-File Report 83-3, 191 p.
Hallberg, G.R., Libra, R.D., Bettis, E.A. III, and Hoyer, B.E., 1984, Hydrologic and water-quality investigations in the Big Spring basin, Clayton County, Iowa: 1983 Water-Year: Iowa Geological Survey Report 84-4, 231 p.
Hallberg, G.R., Libra, R.D., and Hoyer, B.E., 1985, Nonpoint source contamination of groundwater in karst-carbonate aquifers in Iowa, in Perspectives on Nonpoint Source Pollution: Washington, D.C., United States Environmental Protection Agency, Environmental Protection Agency 440/5 85-001, p. 109-114.
Hallberg, G.R., Baker, J.L., and Randall, G.W., 1986, Utility of tile-line effluent studies to evaluate the impact of agricultural practices on groundwater, in Agricultural Impacts on Ground Water: Worthington, Ohio, National Water Well Association, p. 298-326.
Hallberg, G.R., Libra, R.D., Long, K.R., and Splinter, R.C., 1987, Pesticides, groundwater, and rural drinking water quality in Iowa, in Pesticides and Groundwater: A Health Concern for the Midwest: Navarre, MN, The Freshwater Foundation and the USEPA, p. 83-104.
Hallberg, G.R., Libra, R.D., Quade, D.J., Littke, J., and Nations, B., 1989, Groundwater monitoring in the Big Spring basin 1984-1987: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 16, 68 p.
Libra, R.D., Hallberg, G.R., Hoyer, B.E., and Johnson, L.G., 1986, Agricultural impacts on groundwater quality: the Big Spring Basin study in Agricultural Impacts on Ground Water: National Water Well Association, Worthington, OH, p. 253-273.
Libra, R.D., Hallberg, G.R., and Hoyer, B.E., 1987, Impacts of agricultural chemicals on groundwater quality in Iowa, in Fairchild, D.M. (ed.): Ground Water Quality and Agricultural Practices, Chelsea, MI, Lewis Publ. Inc., p. 185-217.
Libra, R.D., Hallberg, G.R., Littke, J.P., Nations, B.K., Quade, D.J., and Rowden, R.D., 1991, Groundwater monitoring in the Big Spring basin 1988-1989: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 21, 29 p.
Littke, J.P., and Hallberg, G.R., 1991, Big Spring basin water-quality monitoring program; design and implementation: Iowa Department of Natural Resources, Open-File Report 91-1, 19 p.
Liu, H., Rowden, R.D., and Libra, R.D., 1997, Groundwater monitoring in the Big Spring basin 1994-1995: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 37, 45 p.
Rowden, R.D., and Libra, R.D., 1990, Hydrogeologic observations from bedrock monitoring well nests in the Big Spring basin: Iowa Department of Natural Resources, Geological Survey Bureau, Open-File Report 90-1, 27 p.
Rowden, R.D., Libra, R.D., Hallberg, G.R., and Nations, B.K., 1993, Groundwater monitoring in the Big Spring basin 1990-1991: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 27, 36 p.
Rowden, R.D., 1995, Temporal trends in groundwater quality of the Big Spring basin 1982-1994: Proceedings, 40th Annual Midwest Ground Water Conference, Abstracts of Papers, October 16-18, 1995 Columbia, MO, p. 50.
Rowden, R.D., Libra, R.D., and Hallberg, G.R., 1995, Surface water monitoring in the Big Spring basin 1986-1992: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 33, 109 p.
Rowden, R.D., Libra, R.D., and Liu, H., 1998, Shallow groundwater and surface water monitoring of the Silver Creek sub-basin within the Big Spring basin 1986-1995: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 38 94 p.
Rowden, R.D., Libra, R.D., and Liu, H., 2000, Groundwater and surface-water monitoring in the Bugenhagen sub-basin 1986-1995: a summary review: Iowa Department of Natural Resources, Geological Survey Bureau, Technical Information Series 41, 150 p.
White, W.B., 1969, Conceptual models for carbonate aquifers: Groundwater, v. 7, p. 15-21.
Singh, K.P., and Stall, J.B., 1971, Derivation of base flow recession curves and parameters: Water Resources Res., v. 7, no. 2, p. 292-303.
U.S. Environmental Protection Agency, 1986, Pesticides in groundwater: background document: U.S.E.P.A., Office of Pesticide Programs, Washington, D.C.
White, W.B., 1977, Conceptual models for carbonate aquifers: revisited, in Dilamarter, R.R., and Csallany, S.C., editors, Hydrologic Problems in Karst Regions, Western Kentucky University, Bowling Green, Kentucky, p. 176-187.
Reprinted with permission from Hydrogeology Journal, 2000, 9: 487-497.