HYDROGEOLOGIC AND WATER QUALITY INVESTIGATIONS IN THE BIG SPRING BASIN, CLAYTON COUNTY, IOWA WATER YEAR 1983
G.R. Hallberg, R.D. Libra, E.A. Bettis III, B.E. Hoyer
Iowa Department of Natural Resources, Geological Survey Bureau,
Open File Report 84-4, 1984, 231 p.
Water Year 1983, the second complete year of monitoring in the Big Spring basin, provides some interesting contrasts to the first year of study in the region. Significantly more detailed hydrologic and water-chemical monitoring were done for particular hydrologic events.
There was a dramatic change in land management in the basin because of the PIK program. Reduction in total corn acreage and slight reductions in fertilization rates produced about a 30-40% decrease in N-fertilizer application for the basin, and a somewhat lesser decrease in pesticide use.
Precipitation in Water Year 1983 totaled 44.5 inches (1130 mm), an increase of 31% from Water Year 1982. The total range of (instantaneous) discharges measured at Big Spring, from 32 to 295 cfs (0.9-8.4 cms), were nearly identical for the two years. However, the greater precipitation increased the total water discharged from the basin by about 47% over Water Year 1982; surface water discharge increased 100% and groundwater discharge (disregarding Water-Year 1982 storage changes) increased 37%. The total water-yield equaled about 49% of precipitation. For groundwater, the runin, conduit-flow component increased 34% while the infiltration, base-flow component increased 8% (Water Year 1982 storage effects cannot be removed from this figure). Even with the greater discharge these groundwater components still occur in the same relative balance over the course of a water year, with the runin, conduit-flow component comprising only 11% of the groundwater discharge, while the infiltration component comprised 89% of the discharge.
The greater water discharge, and particularly the greater groundwater movement through the soil in infiltration caused a significant increase in chemical discharge from the basin. Total nitrate-N discharged from the basin in Water Year 1983 increased 58% from Water Year 1982, and totaled more than 1430 tons (13 x 105 kg) of N; approximately 755 tons (6.9 x 105 kg) in groundwater, and 675 tons (6.1 x 105 kg) in stream-flow. This equals about 43 lbs-N/ac (49 kg-N/ha) for the entire basin. The flow-weighted mean nitrate concentration in groundwater increased from 39 mg/L in Water Year 1982 to 46 mg/L in Water Year 1983. Thus, the mean nitrate concentration for the basin exceeded the 45 mg/L drinking water standard.
Similar trends were recorded on the regional basis as well. The discharge of the Turkey River at Garber increased 40% from Water Year 1982, and the total nitrate-N discharged from the Turkey River basin was approximately 13,400 tons-N (12.1 x 106 kg-N), up from 9400 tons of N in Water Year 1982. This is an equivalent of 27 lbs-N/ac (30 kg-N/ha) for the entire region.
The discharge of the herbicide atrazine in groundwater increased 120% over Water Year 1982. However, this still only amounted to about 31 lbs (14 kg) of atrazine. Atrazine was the only pesticide detected in groundwater year-round, with concentrations ranging from 0.1-5.1 µg/L, with a flow-weighted mean of 0.28 µg/L. Four other commonly used herbicides were intermittently detected in groundwater, but primarily during runin-recharge events in May, June and July. These herbicides were (maximum concentrations noted in parentheses, before common name): Lasso (0.63 µg/L, alachlor); Bladex (1.2 µg/L, cyanazine); and Dual (0.62 µg/L, metolachlor). Lasso appeared in groundwater as late as 08/30/83. The insecticide Dyfonate (fonofos) also appeared in groundwater (0.11 µg/L) during a large runin-recharge, storm event. Even though the amount of and concentration of pesticides in water increased substantially in Water Year 1983, the total mass discharged is still estimated at only about 5% of that normally applied. Numerous other pesticides that were applied in the basin were not detected in groundwater.
Large spring or early summer runoff and discharge events can significantly affect the total pesticides lost in water. During the two-week period of large runoff-discharge events in late-June and early-July about 35% of the total-discharge of atrazine occurred. The amount discharged during this period alone equaled about 80% of the atrazine discharged in Water Year 1982.
The contributions of the components of the groundwater discharge were approximately the same in Water Year 1983 as in Water Year 1982; the infiltration , base-flow component comprising about 90% of the water discharge, and the runin, conduit-flow component only 10%. The highest concentrations and largest mass of nitrate are delivered through the infiltration-component (95%) while the runin component delivered only 5% of the nitrate-N; similar to Water Year 1982. With the large runoff-runin events, the relative delivery of pesticides (based on atrazine) to groundwater was substantially different. In Water Year 1983 the runin component delivered 47% of the pesticides, with a flow-weighted mean atrazine concentration of 1.2 µg/L, while the infiltration component delivered 53% of the pesticides, with a flow-weighted mean atrazine concentration of only 0.16 µg/L. Even with the different conditions of Water Year 1983, the infiltration component still delivers the largest mass of contaminants into the groundwater system. The respective contributions of these components must be considered in any planning of remedial measures or management practices.
Nitrate concentrations monitored from tile-lines, surfacewater sites, Big Spring and the Turkey River all fluctuate in harmony throughout the years of monitoring. The parallel nature of the records demonstrate that similar mechanisms and responses to recharge deliver nitrate to all parts of the hydrologic system. The close coincidence of both nitrate and discharge records between Big Spring and the Turkey River show that the processes and relationships documented at Big Spring are applicable on the regional scale as well.
The total nitrate-N losses from the Big Spring basin increased from an equivalent of 27 lbs-N/ac (31 kg-N/ha) in Water Year 1982 to 43 lbs-N/ac (49 kg-N/ha) in Water Year 1983. Obviously, there is no direct relationship between the nitrate-N discharged and the decreased application of fertilizer-N which resulted from the PIK program in Calendar Year 1983. This is because of the time lag between changes in chemical land treatment and changes in the chemical quality of the groundwater (combined with differences between crop, or calendar years and water years). As shown by a review of various agronomic studies, at moderate to high N-fertilization rates the nitrate-N is stored in the soil, and the amount builds up in direct proportion to the amounts applied and the number of years of application. The nitrate-N leached in any year, such as Water Year 1983, is in large part, related to this storage, which masks the effects of individual years in the short term. Any impact, or decrease in nitrate-N leaching resulting from PIK, would be expected to show in future monitoring. Landuse in the Big Spring basin has been relatively constant between 1979 and 1983, and thus the nitrate-N losses can be put in the context of the acreage that has been in corn-production over that time (3-5 years in various rotations). In this perspective the amount of N lost from this base acreage increased from 47 lbs-N/ac (52 kg-N/ha) in Water Year 1982 to 74 lbs-N/ac (83 kg-N/ha) in Water Year 1983. The Water Year 1982 N-losses were equivalent to 33% of the fertilizer-N applied in 1982; the Water Year 1983 N-losses would be equivalent to 53% of those same 1982 N-fertilizer amounts. Note that this is a minimum figure because it only accounts for nitrate-N losses. Thus, in a relatively wet year, such as Water Year 1983, a minimum equivalent to about 50% of the chemical fertilizer-N applied may be lost into groundwater and surfacewater combined. Particularly considering how well the behavior in the Big Spring system reflects general conditions, the magnitude of N-losses would certainly seem to constitute an economic as well as environmental loss.
As noted these figures are minimums for the amount of N lost from the basin. Piezometer studies and stream monitoring in the basin show that denitrification occurs in local-settings in the soil environment, and possibly in the streams themselves, accounting for additional lost N, that cannot be quantified. Monitoring of ammonium-N and organic-N show that these forms of N are also discharged in groundwater during runin-recharge events. They are discharged in higher quantities in streamflow which leaves the basin. In the alluvial aquifers where denitrification removed the nitrate, pesticides were still present in the groundwater.
In the unique karst-carbonate aquifer system suspended-sediment also occurs in groundwater causing water-quality problems, particularly during runin, conduit-flow periods. During peak, conduit-flow the sediment loads are essentially equal to surface runin water, reaching concentrations of nearly 5000 mg/L and discharge rates of over 190,000 lbs/hr (87,000 kg/hr). The total discharge of suspended sediment at Big Spring alone was about 3500 tons (3.2 x 106 kg). These sediment loads create serious problems for the ICC fish hatchery operations.
Detailed monitoring of discharge and water chemistry, dye tracing, and hydrograph analysis of major discharge events during Water Year 1983 provide many insights to the behavior of the karst-hydrogeologic system. These data verify, and amplify, the prior findings about the flow system and the nature of the contributing components. While the recharge-discharge mechanisms are complex, different analytical and chemical hydrograph separating techniques show that the two fundamental components--infiltration base-flow and runin conduit-flow--can be consistently quantified within about 10%. The chemical monitoring and separation techniques also verify the complexity of components which contribute to major discharge-hydrograph events, and contribute to the nature of water-quality fluctuations.
Monitoring of rainfall-runin events shows that the high surfacewater concentrations of parameters such as suspended sediment (e.g. 5000 mg/L), pesticides (5-20 µg/L), organic and ammonium-N move through the conduit-groundwater system as a 'slug' discharging from the groundwater in essentially the same concentration as they entered. These events also introduce bacteria and potentially pathogenic organisms into groundwater. While the runin component delivers contaminants to the groundwater which are of concern for public health on the local level, the infiltration component is responsible for regional aquifer contamination. Also, infiltration is the recharge mechanism common to all aquifers, which gives these data much broader implications.
Numerous soil cores were collected from the basin, to varying depths, under various landuse, after corn-harvest in 1982 and 1983. The soil samples were analyzed for nitrate-N and pesticides. The amount of nitrate-N stored in the profile vary directly with landuse and increased proportionately with the number of years of fertilized corn. Measured to a depth of 10 feet (3 meters) the amount of nitrate-N stored in soils under forest, pasture, fertilized-pasture and alfalfa-meadow (in rotation with corn) ranged from 48-80 lbs-N/ac (50-90 kg-N/ha) while under high fertilization corn the amounts ranged from 135 lbs-N/ac (150 kg-N/ha) under second-year corn to 400 lbs-N/ac (450 kg-N/ha) under 'continuous' corn.
In the plow-layer, maximum concentrations of 120.0 µg/kg atrazine, 5.8 µg/kg Bladex (cyanazine), 23.0 µg/kg Lasso (alachlor) and 10.0 µg/kg Dyfonate (fonofos) were recorded. Atrazine concentrations of 1.0 µg/kg were noted to depths of nearly 10 feet (3.0 meters), and 0.3 µg/kg atrazine were detected at a depth of 14.5 feet (4.4 meters) at one location. Other pesticides were not detected at great depths.
Other water-quality and crop-use data collected in the region further support the direct, linear relationship, between the increase in nitrates in groundwater with the large increase in N-fertilization that has taken place since the 1960's. A review of agronomic studies which have related N-fertilizer application rates to the N-buildup in soils or the N-losses in the tile drainage, show that this linear relationship is the response that should be predicted in a setting such as northeastern Iowa. The direct relationships between total fertilizer-N applications and groundwater nitrate concentrations also indicate that any significant decrease in the amounts of fertilizer-N applied (or increased efficiency of N-use, such that less could be leached) would be accompanied by a proportional decrease in groundwater nitrate, at least when integrated over a 2.5 year period.
On the short term, seasonal or monthly basis, the concentration of nitrate and mass of nitrate-N discharged show significant, positive, linear relations to the amount of water discharged. Over the long term the excess nitrate-N is stored in the soil. Because infiltration through the soil is the principle component of recharge, the timing of nitrate fluctuations in water supplies is related to seasonal recharge periods, and generally not to the timing of seasonal agricultural practices.
Management alternatives, or 'best-management practices' (BMPs), need to be formulated that will couple standard concerns for soil erosion and surfacewater quality with the need to reduce chemical losses in infiltration to groundwater. This will need to be done to balance our need for efficient agricultural production with the need for safe drinking water. Some measures will control soil erosion and runoff but will promote greater infiltration and additional chemical leaching. Over time these problems can be addressed through new technology. However, a review of many studies suggests that through new combinations of many current and accepted practices these goals can be compatible. Better chemical and nutrient management must be coupled with systems for soil management as well.