Cooperative water quality testing and vulnerability factors for private rural wells in central Indiana.
The use of agricultural chemicals in some areas of the United States has resulted in the contamination of groundwater used by rural residents for drinking water. Nitrogen-containing fertilizers and synthetic herbicides and pesticides can leach into aquifers in concentrations that may increase public health risks. Numerous studies have demonstrated detrimental health effects associated with nitrate intake (3,7,9,11). Too much nitrate in drinking water, for example, can cause blue-baby disease (i.e., methemoglobinemia) in infants (11). Exposure to the herbicides alachlor and atrazine has been linked to adverse health effects including the formation of tumors. These chemicals pose health risks to rural residents who obtain their drinking water from private wells. The extent and severity of private well contamination in agricultural areas, however, remains inadequately studied and largely unexamined (1,4,5,10).
Data from a variety of sources provide some understanding of the extent of agricultural groundwater contamination problems, but these data suffer from a variety of drawbacks and limitations. An estimated 40 million people in rural, agricultural areas in the U.S. rely on individually-owned private wells as their primary source of drinking water (3). These wells are not regulated, but many individual well owners have their wells tested for the presence of chemical contaminants. Although the results of their testing could provide valuable groundwater quality data, this information is confidential and cannot be used in public studies (1). Some national sampling programs have surveyed rural community and private water suppliers that serve relatively large populations (5,6). These programs have failed to characterize the drinking water quality of the local rural subpopulations who obtain their drinking water from individually-owned wells (1). Other government water-testing programs have targeted "problem" areas with vulnerable aquifers where poor management practices have been used. These programs have typically relied on small numbers of samples, resulting in spatial biases that limit the degree to which results can be generalized. Finally, tests of groundwater quality conducted at controlled monitoring wells in aquifers, rather than private wells, may fail to capture localized contamination of well water because the contaminants in private well water may reflect entrance of contaminated surface water directly into the well rather than contamination of the aquifer.
One program that overcomes some of these limitations is the Cooperative Private Well Testing Program developed by the Water Quality Laboratory (WQL) at Heidelberg College in Ohio and the American Farm Bureau Federation. The WQL has completed studies in at least eight states, including Indiana (1). In these cooperative programs, sampling of drinking water supplied by private wells is done by individual well owners. Local or regional organizations sponsor and organize the sampling, and the WQL analyzes the samples (1,2). Participating well owners complete questionnaires about the presence of certain factors, called vulnerability factors, which can contribute to well contamination. They also agree to allow their test results and questionnaire responses to be compiled and used anonymously to assess the overall distribution and degree of contamination (1). Unlike studies focused on groundwater problem areas, cooperative programs typically provide a very large number of geographically-dispersed well samples within a broad region. Although the samples generated in the program are not truly random, the data allow researchers to assess local and regional groundwater quality, vulnerability, and potential threats from agrichemicals. Other advantages of cooperative programs are that they test the quality of drinking water from private wells, rather than aquifer quality, and that analyses are cheaper because of volume discounts.
Cooperative programs, however, have several potential limitations. Results may be biased because well owners who believe that their drinking water may be contaminated may be more likely to participate, and the quality of the information about vulnerability factors provided by participants cannot be guaranteed, The questionnaire, for example, asks participants to assess the type of soil, a determination best made by certified soil scientists. Also, some participants may not be able to accurately estimate the distances between their wells and potential sources of contamination. These types of limitations are not significant, however, given the potential benefits of the program.
The Hoosier Heartland Well Testing Program
This paper presents the results of a cooperative well testing study to investigate contamination from nitrogen and pesticides in the 10-county region of central Indiana known as the Hoosier Heartland. The objectives of this study were to establish baseline contaminant data for rural wells and to illustrate how land use, proximity of potential contaminant sources, soils, and well characteristics influence the susceptibility of individual wells to contamination. This article summarizes data for 2,252 samples and compares results with government standards, nationwide studies, and other studies in Indiana. The study was sponsored by the Hoosier Heartland Resource Conservation and Development Council (RCD) with financial support from the Indiana Department of Environmental Management (IDEM), and programmatic support from the Indiana Farm Bureau, the Soil Conservation Service (now Natural Resources Conservation Service), local health departments, and other local organizations.
Detection Levels and MCLs Tested for Chemicals of Interest
Chemical Detection Level MCL
Total Nitrogen 0.3 mg/L 10 mg/L Triazine 0.05 [[micro]gram]/L 3 [[micro]gram]/L Alachlor 0.2 [[micro]gram]/L 2 [[micro]gram]/L
1 mg/L = 1 milligram per liter, or 1 part per million.
1 [[micro]gram]/L = 1 microgram per liter, or 1 part per billion.
The Hoosier Heartland includes both glaciated and unglaciated areas, and a variety of rural land uses. The northern counties of the region are glaciated and are underlain by unconsolidated deposits of till that provide good agricultural land. The southern counties are unglaciated, hilly, and forested. Approximately 60 percent of the 2.4 million acres in the region is devoted to agriculture, mainly corn (39 percent), soybeans (32 percent), and wheat (4 percent). The remaining 40 percent of the land comprises the heavily urbanized Indianapolis metropolitan region and forested lands.
The cooperative methods used to collect well data were developed by Baker (1,2). Local sponsors in each of the 10 Hoosier Heartland counties coordinated and conducted the collection of samples. These sponsors advertised the program, recruited participants, arranged for distribution of sampling kits, and organized sampling stations and sampling dates. Participants picked up sampling kits and questionnaires from the local sponsors. These individuals collected samples from their taps according to instructions provided by the WQL and completed the questionnaires about vulnerability factors, including well characteristics, soil characteristics, land use, and proximity of contaminant sources.
Samples were sent to the WQL for analyses. Analytical methods are described in other reports (1,2). Although samples were analyzed for many chemicals, this study focuses on three agrichemical groups: nitrogen, triazine, and alachlor. "Total Nitrogen" refers to the sum of the nitrate and nitrite concentrations. The triazine levels primarily indicate the presence of atrazine, although other triazine herbicides such as cyanazine and simazine may also contribute to its measured concentration. Sample results provided by the WQL were analyzed, graphed, and classified in three categories:
* contaminant concentrations below detection levels;
* contaminant concentrations above detection levels, but below maximum contaminant levels (MCLs); and
* contaminant concentrations above MCLs.
A concentration below detection indicates that the contaminant is present in concentrations too small to be accurately or reliably detected by the analytical method employed. The second and third categories are based on maximum contaminant levels or MCLs. MCLs are drinking water standards set by the U.S. Environmental Protection Agency (EPA). Table 1 lists the detection levels and MCLs for each of the chemical groups analyzed in this study.
To obtain a relative understanding of the quality of well water in central Indiana, it is helpful to compare these results to the results of other regional and national studies. National values for well contamination were obtained from papers published by researchers at the EPA and WQL. The EPA has published results from a 1990 national pesticide survey and compiled data from monitoring studies completed between 1971 and 1991 into a pesticides database. EPA also worked with the Monsanto Company on the National Alachlor Well Water Survey (NAWWS). This study, conducted from mid-1987 through the end of 1989, sampled private rural domestic wells for alachlor, atrazine, nitrates, and other chemicals. The WQL recently summarized the results from its cooperative studies (2). Along with national averages, WQL reported results from individual states.
Levels of Contamination
Results provided by WQL for the Hoosier Heartland are presented here by chemical and by vulnerability factor. Figures 1a, 1b, and 1c depict the proportions of wells in which nitrogen, triazine, and alachlor, respectively, were detected. Table 2 compares the Hoosier Heartland results with results presented elsewhere. For each of the three indicator chemicals, relatively small proportions of the samples had chemical concentrations above the MCLs. Only nitrogen was detected in a relatively large proportion of wells.
Of the wells tested for total nitrogen, 2.5 percent had concentrations which surpassed [TABULAR DATA FOR TABLE 2 OMITTED] the MCL of 10 mg/L [ILLUSTRATION FOR FIGURE 1A OMITTED]. Approximately 14 percent of wells had nitrogen concentrations which were detectable but below the MCL, and the remaining 83.4 percent were below detection levels. The proportion of wells contaminated by nitrogen (i.e., the proportion that exceeded the MCL) was generally lower than the proportion elsewhere in Indiana and in cooperative programs in other states, and was comparable to proportions reported by EPA (Table 2).
Of the wells tested for triazine, only 0.1 percent had concentrations in excess of the MCL of 3 [[micro]gram]/L [ILLUSTRATION FOR FIGURE 1B OMITTED]. Triazine was detected at levels below the MCL in another 4.4 percent of the wells, but was not detected in the remaining 95.5 percent of the wells. The proportion of wells in which triazine was detected was lower than the proportions reported for other areas (Table 2). The proportion of samples which exceeded the MCL was lower than that in the WQL's cooperative program overall and in the EPA database, but was comparable to the proportions for NAWWS and the state of Indiana.
Of the wells tested for alachlor, 0.5 percent had concentrations in excess of the MCL of 2 [[micro]gram]/L [ILLUSTRATION FOR FIGURE 1C OMITTED]. Alachlor was detected in concentrations below the MCL in approximately 2.5 percent of the wells, but was not detected in the remaining 97 percent. The proportion of wells with alachlor in concentrations greater than the MCL was lower than the percentage in comparable studies reported by the WQL but higher than percentages in the EPA database and the NAWWS.
Information reported by participants provides insight into factors that contribute to well contamination. By correlating concentrations with vulnerability factors, potential causes of contamination can be determined. Three general categories of vulnerability factors are:
* well characteristics (construction, age, depth)
* soil characteristics (clay, mixed soil, sandy)
* proximity of contaminant source (e.g., crops, feedlots)
Figures 2a, 2b, and 2c represent the proportion of wells in which chemicals were detected (both above and below the MCL) by chemical and vulnerability factor. The results of this analysis generally confirm the expected: wells that are older, poorly constructed, shallow, and located in permeable soils have a higher probability of elevated contaminant concentrations. Wells in close proximity to possible sources of contamination are also more vulnerable. In addition, wells with two or more of these factors (e.g., shallow wells dug in sand) are more likely to be contaminated.
Certain characteristics of wells correlated positively with the frequency with which chemicals were detected [ILLUSTRATION FOR FIGURE 2A OMITTED]. The most striking result was the frequency with which nitrogen was detected in the 59 dug wells: nitrogen was detected in nearly one-half of them. Although the percentages were much lower, triazine and alachlor were also detected more frequently in dug wells. Driven and drilled wells were clearly less prone to contamination. Chemicals were also detected more frequently in older and shallower wells: 21.9 percent of the 215 wells installed before 1950 had detectable levels of nitrogen, and, of the 357 shallow wells in the study, nitrogen, triazine, and alachlor were detected in 33.1 percent, 8.7 percent, and 5.1 percent, respectively. Although the proportions were quite small, chemicals were also detected in "good" wells - wells that were drilled, developed since 1950, and more than 100 feet deep.
Each of the chemicals was detected more frequently in soils reported to be sandy than in soils reported as clay or mixed [ILLUSTRATION FOR FIGURE 2B OMITTED] Detection of nitrogen levels clearly correlated with the presence of sandy soils (30.1 percent below MCL, 11.2 percent above MCL). Triazine and alachlor were detected in 7.7 percent and 4.2 percent, respectively, of the wells with sandy soil. No concentrations of triazine in wells in sandy soil exceeded the MCL; however, more than one-half of the alachlor concentrations did. Chemicals were detected in very small proportions of wells in clay and mixed soils. Results for wells in these soils were very similar, possibly because participants may have been unable to classify soils correctly.
Proximity of Contaminant Source
Respondents were asked about the proximity of their wells to crops, feedlots, and other potential sources of pollution including animal waste and fertilizer dealers. Nitrogen, triazine, and alachlor were detected more frequently in wells closest to potential sources of contamination. Among the different sources of contamination, chemicals were detected most frequently in wells in close proximity to feedlots, followed by wells in close proximity to crops [ILLUSTRATION FOR FIGURE 2C OMITTED]. The proximity of wells to animal wastes and fertilizer dealers also correlated somewhat with the detection of contamination. Other potential sources of contamination were identified (e.g., chemical lawn treatment, septic systems, landfills) but none appeared to correlate with the frequency with which chemicals were detected.
Observations and Conclusions
This study indicates that well contamination is not widespread and that wells in the Hoosier Heartland are less contaminated than wells in a number of other states and wells included in a number of national studies. Although nitrogen was detected in more than 15 percent of the wells that were tested, the proportion generally was lower than proportions reported in other studies. Furthermore, the very low proportion of wells in which triazine and alachlor were detected indicates that central Indiana does not have a large pesticide contamination problem. When comparing the Hoosier Heartland results to results from other studies, caution must be exercised, however, because the studies vary in sample size, experimental controls, and landscape conditions.
This study also shows that wells which are shallow, old, dug, and located in permeable soils are more likely to have elevated concentrations of contaminants. All of these conditions allow contaminants to move more freely into the well. Wells in close proximity to particular sources of contamination, such as feedlots, are also more likely to have higher levels of contaminants.
Chemicals were also detected in very small numbers of deep wells drilled in soils with notable clay content and located at a distance from sources of contamination. These results may indicate contamination of aquifers, or they may reflect entrance of contaminants directly into the wellhead. Additional investigations to confirm findings are warranted.
Individual well owners have no control over some vulnerability factors such as soil type, but they may take steps to avoid other factors by drilling a new well or by instituting best management practices (BMPs) and keeping greater distances between wells and possible sources of contamination. In this particular application, individual test results were provided only to participants, who alone were responsible for determining whether action to address possible contamination was needed. Because data made available for analyses did not include addresses that could link results to particular individuals, local environmental professionals could not target education efforts to specific well owners. The RCD, however, provided copies of results to each of the 10 Soil and Water Conservation Districts in the region, many of which displayed the information at county fairs. Information on BMPs was also provided. A useful follow-up study would involve research to determine whether participants attempted to solve problems.
Like other cooperative well testing programs, this study produced a large set of samples which provided owners of wells with useful information about the quality of their well water. Despite limitations, such as the lack of rigorous experimental controls in sampling and subjective judgments by participants about vulnerability factors, analysts completed a comprehensive assessment of regional well water quality that local officials have used to educate rural residents about factors which increase the likelihood that their wells may be contaminated. Cooperative well testing programs provide a cost-effective approach to collection of basic information that local environmental health officials can use to help people protect their water supplies.
The authors wish to acknowledge Bill Beard of the Hoosier Heartland Resource Conservation and Development Council and Jack Wittman, Anne Kaufmann, Lisa Streisfeld, and Robyn Dommel of the School of Public and Environmental Affairs, Indiana University, who contributed to the report on which this paper is based.
1. Baker, D.B. (1990), "Groundwater Quality Assessment Through Cooperative Private Well Testing: An Ohio example," Journal of Soil and Water Conservation, 45(2):230-235.
2. Baker, D.B., L.K. Wallrabenstein, and R.P. Richards (in press), "Well Vulnerability and Agrichemical Contamination: Assessments from a voluntary well testing program," Proceedings of the Fourth National Conference on Pesticides: New directions in pesticide research, development and policy, Virginia Water Resources Center, Blacksburg, VA.
3. Bouwer, H. (1990), "Agricultural Chemicals and Groundwater Quality," Journal of Soil and Water Conservation, 45(2):184-189.
4. Conservation Foundation (1987), State of the Environment: A view toward the nineties, Washington, D.C.
5. Environmental Protection Agency (1987), National Pesticide Survey: Summary of statistical design, Washington, D.C.
6. Francis, J.D., B.L. Bower, W.F. Graham, O.W. Larson III, J.L. McCaull, and H. Moran Vigorita, National Statistical Assessment of Rural Water Conditions - Executive Summary, Dept. of Rural Sociology, Cornell University, Ithaca, NY.
7. Hartmann, G.R. (1986), "Overview of Agricultural Chemicals in Groundwater," Proceedings of the Conference on Agricultural Impacts on Groundwater, National Water Well Association, Dublin, OH, pp. 1-63.
8. Indiana Agricultural Statistics Service (1993), Indiana Agricultural Statistics - 1992-93, West Lafayette, Indiana: Purdue University.
9. Mirvish, S.S. (1977), "N-Nitroso Compounds, Nitrite and Nitrate: Possible implications for the causation of human cancer," Progressive Water Technology, 8:195-207.
10. Nielsen, E.G., and L.K. Lee (1987), The Magnitude and Costs of Groundwater Contamination from Agricultural Chemicals: A national perspective, U.S. Dept. of Agriculture, Washington, D.C.
11. Shuval, H.I., et al. (1980), "Infant Methemoglobinemia and Other Health Effects of Nitrates in Drinking Water," Progressive Water Technology, 12:173.
Corresponding Author: Greg Lindsey, Ph.D., Associate Director for Environmental Research, Center for Urban Policy and the Environment, 342 North Senate, Indianapolis, Indiana 46204-1708.
|Printer friendly Cite/link Email Feedback|
|Publication:||Journal of Environmental Health|
|Date:||Apr 1, 1996|
|Previous Article:||Primary prevention in Oregon.|
|Next Article:||Failure of routine restaurant inspections: restaurant-related foodborne outbreaks in Alabama, 1992, and Mississippi, 1993.|