WaterSubject

The leakage of pollutants from agricultural lands to aquifers has increased greatly, driven by increasing fertilizer and pesticide use. Because this increase is recent, ground water pollutant concentrations, loads, and exports may also be increasing as pollutants penetrate more deeply into aquifers. We established in an aquifer profile a ground water recharge and pollutant leakage chronology in an agricultural landscape where 30 m of till blankets a 57-m thick sandstone aquifer. Pollutant concentrations increased from older ground water (1963) at the aquifer base to younger ground water (1985) at its top, a signal of increasing pollutant leakage. Nitrate-N increased from 0.9 to 13.2 mg/L, implying that leakage increased from 1.9 to 16.5 kg/ha/year. Nitrate load and export could increase from 130% to 230% before reaching a steady state in 20 to 40 years. Chloride increases were similar. Pesticide residues alachlor ethane sulfonic acid (ESA), metolachlor ESA, and atrazine residues partially penetrated the aquifer profile. Their concentration-age-date patterns exhibited an initial increase and then a leveling corresponding to the timing of product adoption and leveling of demand. Unlike NO₃, projecting pesticide residue steady states is complicated by the phasing in and out of pesticide products over time; for example, neither alachlor nor atrazine is currently used in the area, and newer products, which have not had time to transit to the aquifer, have been adopted. The circumstances that resulted in the lack of a pollutant steady state are not rare; thus, the lack of steady states in agricultural region aquifers may not be uncommon.

The leakage of pollutants from agricultural landscapes to ground water degrades drinking water resources and exports contaminants that threaten aquatic ecosystems. This pollutant leakage (mainly NO₃ and pesticide residues) increased greatly over the last several decades, driven by increasing fertilizer and pesticide use, particularly over the period of about 1960 to 1990. Given that the increase in pollutant leakage (pollutant mass per landscape area-time) is a recent phenomenon compared with the residence time of ground water in many aquifers, aquifer pollutant loads (the pollutant mass per aquifer volume in aquifer storage) may not generally be at a steady state with pollutant leakage in agricultural regions. “Steady state” is defined here as the condition in which the mass of a pollutant and its spatial distribution in an aquifer remain more or less constant over time. In the absence of a steady state, pollutants will continue to penetrate more deeply into an aquifer, increasing the aquifer’s pollutant load and its pollutant export to surface water. A steady state develops when modern ground water and its accompanying pollutants penetrate an aquifer’s entire thickness or when pollutants in an aquifer degrade at a rate equal to their leakage.

Many studies of NO₃ and pesticide residues (”pollutants”) have surveyed their presence and concentrations at local to subcontinental scales or determined how pollutant presence and concentration vary with land use and physical setting. Few studies have examined how current pollutant conditions evolved in aquifers, whether pollutants are at steady state, or failing a steady state, what pollutant conditions will arise in the future. A notable early work by linked the evolution of increasing ground water NO₃ to increasing fertilizer use in the Big Springs Iowa watershed, and similar linkages have since been affirmed in other locales. Analogous linkages have not been as strongly established for pesticide residues, though great progress has been made in identifying atrazine, alachlor, and metolachlor residues (parent plus degradates) as widespread ground water pollutants.

Some questions of penetration and steady state have been addressed for NO₃ mainly in areas with thin aquifers , but again, analogous work for pesticides is mostly lacking. The Locust Grove Maryland Aquifer serves as a useful archetype of NO₃ studies. Ground water near the water table of the 20-m-thick Locust Grove Aquifer was less than 5 years old and contained more than 10 mg/L NO₃-N, while ground water at the aquifer’s base was 35 years old and contained less than 5 mg/L NO₃-N. The concentration age and depth pattern was attributed to increased NO₃ leakage from the landscape during 1960 to 1986, driven largely by increased fertilizer use. Barring changes in land use and land management, the Locust Grover aquifer NO₃ load would be expected to increase until modern ground water penetrated its entire thickness. Studies conflict as to the importance of aquifer denitrification in limiting NO₃ penetration and forcing a steady state. Denitrification did not limit penetration in the Locust Grove study area, or in the glacial drift aquifer west of Kitchener and Waterloo, Ontario, Canada. Denitrification partially attenuated NO₃ in Chesapeake Bay ground water, and greatly limited NO₃ penetration in the Otter Tail Minnesota glacial outwash aquifer, aided by a large sediment organic carbon content (median = 0.15%).

In this study, we examined questions of agricultural pollutant penetration and steady state at a study site in the northern midwestern United States where a till mantle covers the thick sandstone “upper bedrock aquifer.” Specifically, we determined whether a modern pollutant load penetrated the entire aquifer thickness, if degradation mechanisms limited pollutant penetration, and what pollutant conditions might develop in the future. Our approach employed “ground water stratigraphy” methods, establishing and interpreting in an aquifer profile the depth structure of NO₃, denitrified NO₃, pesticide residues, ground water recharge age-date, and other dissolved constituents. A companion paper by provides additional information about redox chemistry, denitrification mechanisms, and ground water impacts collateral with pollutant leakage, as well as details about dissolved gas and chlorofluorocarbon (CFC) age-date sampling and analysis.

The leakage chronology at Springfield Corners displayed increasing pollutant leakage coinciding with an era of increasing agricultural inputs. Net NO₃-N concentrations in ground water recharge increased from 0.9 mg/L in 1963 to 13.2 mg/L in 1985, and were still increasing at 0.50 mg/L/year at the end of the chronology. These concentrations correspond to an increase in net NO₃-N leakage rates from 1.9 to 16.5 kg/ha/year during the same period. Leakage of NO₃-N could plausibly reach 24.5 kg/ha/year before a steady state is achieved. As a result of increasing NO₃ leakage, a limited ability for denitrification to control NO₃, and the time required for modern ground water to penetrate the aquifer thickness, aquifer NO₃ load and export could increase by 130% to 230% before a steady state is reached in 20 to 40 years. Chloride leakage, which we attribute mainly to increasing KCl fertilizer, grew similarly. The ground water stratigraphy approach was also able to establish what pesticide residues leaked to ground water (alachlor ESA, metolachlor ESA, and those of atrazine), their leakage chronology, the degree to which they penetrated the aquifer, and that their in-aquifer degradation was slow or lacking. However, projecting pesticide residue futures and steady states is complicated by the phasing in and out of pesticide products over time; for instance, neither alachlor nor atrazine is currently used in the study area, and newer products, which have not had sufficient time to transit to the aquifer, have been adopted.

Pollutant penetration may be far from a steady state in the aquifers of some agricultural areas. The cause is that ground water residence times are frequently long compared with the relatively recent increase in pollutant leakage from agricultural landscapes brought about by increasing fertilizer and pesticide use. Characteristics of areas importantly far from a steady state may include the following: (1) prevailing farming systems involve large amounts of inputs that offer a potential pollutant source; (2) moisture in excess of evapotranspiration is available as leaching agent; (3) landscape runoff is conducted mainly by subsurface flow through an aquifer to drainages (i.e., overland runoff and tile drain flow are small or lacking); (4) large thicknesses of sediments conduct landscape runoff to discharges resulting in long subsurface residence times; (5) conditions supporting denitrification and pesticide degradation are absent or slow, perhaps due to a lack of solid phase or dissolved electron acceptors. As these characteristics are not rare, agricultural locales with aquifers lacking pollutant steady states may not be uncommon. Progression toward a steady state at local scales may be manifested as increasing pollutant concentrations in drinking water. At large scales, the progression could be manifested by increased surface water pollutant loads with corresponding ecological harms, such as an intensification of marine hypoxia.