Kansas Geological Survey, Open-file Report 1999-45

Preliminary Assessment of the Distribution and Sources of Nitrate-Nitrogen in Kansas Ground Water

by
M. A. Townsend and D. P. Young

KGS Open-file Report 1999-45

Abstract

Nitrate in ground water is a potential problem in many parts of Kansas. From 1989 to 1998, 747 water samples were collected from domestic, irrigation, monitoring, and public water supply wells and analyzed for nitrate-nitrogen by the Kansas Geological Survey. Of all the samples 29% have nitrate-N values less than or equal to 3 mg/L; 51% have values between 3 and 10 mg/L; and 20% have values greater than or equal to 10 mg/L.

Nitrogen-15 isotope values were used in several studies to determine sources of the observed nitrate-N. Of the 112 samples analyzed in this data set, 29% of the samples indicate a fertilizer source, 22% indicate mixed sources, 42% indicate an animal waste source, and 7% indicate denitrification processes have occurred.

Factors that show a statistically significant relationship with the occurrence of nitrate in Kansas ground water in the preliminary stages of this study include geographic area of the state, depth of well, and age of well.

Introduction

Ground water is the major source of drinking water for 70% of the residents of Kansas. In rural areas, 85% of the population relies on ground water. Contaminants in ground water that may cause health problems, such as nitrate, are of significant concern.

Nitrate is the most common inorganic contaminant of ground water in Kansas. Background level for the natural occurrence of nitrate-N is considered less than or equal to 3 mg/L (Madison and Brunett, 1985). Values above 3 mg/L are considered indications that man-induced sources of contamination have occurred. Areas that are heavily agricultural, such as much of the Midwest, appear to have a greater number of detections of nitrate-N above the background level in the ground water (Spalding and Exner, 1993; Kolpin et al., 1994).

The objectives of this preliminary assessment were to evaluate available ground-water nitrate-N data for the time period of 1990 to 1998 to determine if any trends in concentrations within the state existed, or if there were any factors that appeared to have an impact on the observed distribution of nitrate in ground water. The work also has been useful in identifying areas where more study is needed.

Methods

Sources of Data

The Kansas Geological Survey (KGS) is in the process of evaluating the occurrence of nitrate-N in ground water throughout the state. The sources of the data for this preliminary assessment were studies where the chemical analyses were done by KGS and the sampling dates were from 1990 through 1998. Agencies contributing to the work include the five ground-water management districts and the Kansas Department of Agriculture (KDA).

Distribution by well type shows 472 irrigation wells, 206 domestic wells, 40 monitoring wells, and 29 public water supply wells. Additional information was obtained from water well completion forms (WWC-5s) on file at KGS, Water Information Management and Analysis System (WIMAS) database (Wilson, 1998) and field notes.

Because some of the data were collected for specific site studies where nitrate contamination was observed or suspected, there is some bias in the overall evaluation. However, there are a number of samples from more regional sampling events, such as those performed by the KDA chemigation program.

Sample Collection and Chemical Analyses

Samples were collected from wells after the well was purged for 3 to 5 well volumes and/or the temperature, pH, and specific conductance had stabilized for three consecutive readings at 3 to 5 minute intervals. Notes concerning land use around the well and farmstead were recorded to use in the evaluation of sources of nitrate in the water samples.

Samples for nitrate analysis were stored on ice in 250-ml polyethylene bottles with 2 ml HCl as preservative. Samples were transferred to a refrigerator upon return to KGS. Nitrate analyses were performed by the KGS Analytical Services section within one month of sample time. The method used was the UV screening method developed by Hathaway (1990).

Samples for nitrogen-15 isotope analysis were collected in 125-ml polyethylene bottles. These samples were stored on ice until return to the Survey, where they were frozen. Samples were shipped frozen by next day delivery to the isotope lab at the University of Virginia, Charlottesville. Samples were stored frozen until thawed and analyzed on GC Mass Spectrophotometer.

Statistical Methods

SAS statistical package for UNIX systems (SAS Institute, 1996, v. 6.2) was used to evaluate the data for this study. Because of the lack of normality in the data set, non-parametric methods were used. The specific tests used were Spearman rho rank correlation test, the Kruskal-Wallis test (analysis of variance equivalent) and the Mann-Whitney t-test equivalent. All statistical tests were performed at the 90 percent level of confidence (α = 0.10). Tests with values greater than the α = 0.10 are considered not significant.

Minimum values of 0.02 mg/L nitrate-N indicate concentrations below the detection limit of the nitrate analysis. Because only 1% of the data set (8 samples) had values below the detection limit, the simple substitution method discussed by Helsel and Hirsch (1992) was used. Substitution of 0.02 mg/L for the non-detects was used in calculation of the mean and median values reported in Table 1 in the Results and Discussion section and in all other statistical analyses.

Results and Discussion

Distribution of Sampling Points and Nitrate-N Concentrations

The distribution of sampling points for this study is shown in Figure 1 for irrigation wells, and Figure 2 for non-irrigation wells, which include domestic, public water supply, and monitoring wells. The data set was divided into irrigation and non-irrigation wells because of the difference in depth of the wells in some parts of the state and also because in some parts of the state the wells sampled were predominantly either irrigation or non-irrigation type. Evaluation of these figures indicates the areas where detailed site studies were undertaken (indicated by high data density) and the more regional areal surveys.

Figure 1. Distribution of nitrate-nitrogen in groundwater from irrigation wells in Kansas. A larger version of this figure is available as an Acrobat PDF file.

Figure 2. Distribution of nitrate-nitrogen in groundwater from non-irrigation wells (domestic, stock, monitoring, and public-water supply) in Kansas. A larger version of this figure is available as an Acrobat PDF file.

Figures 1 and 2 also show the geographic distribution of nitrate-N concentrations throughout the state. As can be seen from these two figures nitrate concentrations in most of the samples are greater than the background level of 3 mg/L. This indicates that much of Kansas' ground water has been impacted by anthropogenic sources of nitrate.

Results of chemical analyses and other physical data for the sites are presented in Appendix A. Table 1 is a summary of simple statistics for parameters evaluated in this study. The mean, median, and range of the sample values (minimum and maximum) are reported. Data were grouped a number of ways to ascertain possible relationships between nitrate-N concentrations and different independent variables. Table 1 includes the nitrate-N values for different groupings of the data. The principal groups are depth of well, well type, age of irrigation wells, and the three geographic areas that were evaluated in this assessment.

Nitrate-N concentrations in samples from irrigation wells ranged from less than 0.02 to 43.3 mg/L, with a mean of 6.5 and median of 4.8 mg/L. Concentrations from domestic wells ranged from less than 0.02 to 77.4 with a mean of 8.7 and a median of 5.7 mg/L. The range of nitrate-N values for public water supply was 0.2 to 58.4 mg/L with a mean of 8.6 and median of 6.3 mg/L. The range of nitrate-N values for monitoring wells was less than 0.02 to 20.1 with a mean of 3.9 and a median of 1.8 mg/L.

Although the reported means suggest the possibility of discernable differences in concentrations among the groups, the statistical tests used, which are based on the median values (see Methods section), indicate that the variation between the well types is not statistically significant. Well type is one of the factors that will be reevaluated as more data become available.

Variables that did have statistically significant relations with nitrate-N concentrations are discussed in the section: Factors Affecting the Occurrence of Nitrate in Ground Water.

Table 1. Summary of Simple Statistics.

Potential Sources of Nitrate-Nitrogen

Figure 3 illustrates the various sources and destinations for nitrogen throughout the nitrogen cycle. The major sources of nitrate are from surface activities including application of fertilizer, human and animal wastes and plant decay. The major destinations for nitrogen in different forms are incorporation by soil microorganisms as food, storage as organic nitrogen in plants or dead microorganisms that have not yet decayed, adsorption of ammonium ion to clay particles, volatilization of ammonia, denitrification (conversion of nitrate to nitrogen gases), and leaching of nitrate to ground water.

Figure 3. Nitrogen cycle illustrating sources and sinks for different forms of nitrogen. Figure illustrates that many different processes can lead to the formation of nitrate.

Two major sources of anthropogenic nitrogen for the formation of nitrate are animal waste and fertilizer. Animal waste sources of nitrogen include feedlots, manure applied as fertilizer, and septic systems. Roughly 30 percent of the population of Kansas disposes of its waste through use of septic systems. Table 2 lists an estimate of the potential volume of nitrogen available from various animals commonly raised in Kansas and also from septic systems.

Table 2. Potential nitrogen loads from animal wastes in Kansas. Population statistics from 1990 U. S. Census; Animal Statistics from KDA 1997 Farm Facts.

Nitrate in ground water is frequently associated with agriculture, the largest industry in Kansas. Plants have basic needs: water, nitrogen, and other nutrients. Although there are other sources of nitrogen for crops, nitrogen is commonly applied as chemical fertilizer. Figure 4 shows the quantity of nitrogen-based fertilizers sold from 1945-1997, providing an estimate of the quantity used. The striking upward trend indicates the potential for a continuous and growing source of nitrate available for leaching from the unused portion of the fertilizer. The figure also shows the cumulative water rights issued mainly for irrigation wells since 1945. The increase in the number of water rights reflects the increase in irrigated farming. Generally, more fertilizer is applied to irrigated cropland compared with non-irrigated cropland. The extra irrigation water applied increases the potential for leaching of water and nitrate.

Figure 4. Relationship between the increase in irrigated farming (as illustrated by the increase in irrigation water rights) and the quantity of nitrogen fertilizer sold.

Factors Affecting the Occurrence of Nitrate in Ground Water

Many studies have identified a variety of factors that affect the occurrence of nitrate in ground water. Some of these factors include depth of well, well construction, depth to water, soil type, precipitation, land use, farming practices (such as irrigated versus dryland farming, timing of fertilizer application, use of manure versus chemical fertilizer), and many others (Hallberg, 1989; Kolpin and others, 1994; Spalding and Exner, 1993). The next few sections describe the findings from this preliminary assessment concerning some of the factors that are related to the occurrence of nitrate in Kansas' ground water.

Geographic Area

Across the state of Kansas, characteristics such as geology, precipitation, soils and land use vary widely. For example, mean annual precipitation varies from 16 inches/year in western Kansas to over 40 inches/year in southeastern Kansas. Consideration of the above factors may be important in evaluating the occurrence of nitrate in ground water. One way to encompass the influence of these factors is to evaluate the occurrence of nitrate in ground water by dividing the state into geographic sub-units.

Figure 5 shows the subdivisions of the state used in this assessment. Areas 1 and 2 overlie the High Plains aquifer and the great majority of wells sampled in these areas tap the High Plains aquifer. More specifically, Area 1 overlies the Ogallala portion of the High Plains aquifer in western Kansas, and Area 2 overlies the Great Bend Prairie and Equus Beds portions of the aquifer in south-central Kansas. The aquifer in Area 2 is generally shallower and soils are sandier compared with Area 1. Wells in Area 3 are typically screened in relatively shallow and narrow alluvial aquifers in the Solomon and portions of the Kansas River basins. No samples were collected in southeastern Kansas during the period of study.

Figure 5. Geographical subdivisions used in evaluation of occurrence of nitrate in ground water. Area 1 is representative of western High Plains, primarily the Ogallala aquifer. Area 2 represents the south-central High Plains aquifer. Area 3 represents the alluvial aquifers of the Solomon and portions of the Kansas River basins.

Figure 6 illustrates the variation of nitrate concentration in ground-water samples in the three areas. Nitrate-N concentrations were generally higher in Area 2 (south-central Kansas) compared with Area 1 (western Kansas). This is attributed primarily to the shallower aquifer (shallower wells and shallower depth to water) and sandier soils in Area 2. Also, the greater amount of precipitation may enhance the amount of nitrate leached to ground water. It is also possible that the presence of caliche zones in the unsaturated zone in western Kansas may permit decreases in nitrate concentration by either chemical or biological reduction (Herbel and Spalding, 1993). Future work will examine these questions as well as potential differences in land use (such as fertilizer and irrigation rates), occurrence of feedlots, and other potential point sources in the two Areas.

Figure 6. Box plots of ground-water nitrate-N concentrations in different areas of the state. Values of 0.02 mg/L are proxy values for nitrate-N values below detection limit.

Concentrations in Area 3 tended to be higher than those in the High Plains regions (Table 1). These higher concentrations are probably related to the limited aquifer extent (in both depth and width), the permeable sediments, the higher rate of precipitation, and the land uses along the river corridors. For example, flood irrigation is more common in this area than in the western parts of the state. Elevated concentrations may also be related to the length of time that potentially polluting activities have occurred along the river corridors. The majority of the samples from public water supply wells came from Area 3, which may explain the relatively high concentrations observed in these wells (Table 1). Future work will examine these variables in more detail.

Depth of Well

Many studies on nitrate contamination of ground water have found a negative correlation between the concentration of nitrate and the depth of well, indicating that shallower wells are more likely to be contaminated than deeper wells (Spalding, 1988). In this assessment shallower wells tended to show more nitrate contamination than deeper wells. This result was expected, as the source of nitrate is normally at the land surface. As indicated previously, shallower well depths in Area 2 compared with Area 1 are as least partly responsible the higher concentrations in Area 2. Future work will examine other relations within the individual regions such as depth to water and vadose zone characteristics.

Figure 7 shows the distribution of nitrate for the 536 samples with reliable depth measurements. A higher percentage of shallower wells (≤ the median depth of 108 ft; Table 1) have concentrations greater than 10 mg/L. The graph also illustrates that the majority of the samples overall have nitrate-N greater than 3 mg/L, which indicates that nitrate contamination may be more widespread than previously suspected.

Figure 7. Nitrate-N ranges from wells less than and greater than the median depth of 108 ft. Number of samples in each group shown above bar.

Age of Well

Age of well may be an indicator of at least two factors that may affect nitrate concentrations in ground water: 1) well construction and 2) length of time potentially polluting activities may have occurred near a well. Both of these factors would favor higher concentrations in older wells.

Poor well construction is frequently cited as a possible pathway for movement of nitrate from land surface to ground water (Kolpin and others, 1994; Steiken and others, 1988). In Kansas, legislation was passed in 1975 which required more stringent well construction methods, primarily to mitigate movement of surface water to ground water via the annular space of wells. The 1975 legislation required that the annular space of wells be grouted from land surface to a minimum of 20 ft or to a minimum of five ft into the first clay or shale layer if one is present, whichever is greater.

Wells installed prior to 1975 frequently have a gravel packed well bore from land surface to the bottom of the well. The presence of this gravel may permit rapid flow of surface water to ground water. In addition, the gravel provides a highly permeable zone that may permit flow of perched water from clay or caliche zones to ground water. Multiple well screens, common in irrigation wells in some areas, may permit mixing of different zones of ground water.

Because of the 1975 legislation, the irrigation well data, based on water right issuance date, were divided into pre- and post-1975 categories to see if a statistically significant difference in nitrate-N concentrations exists. The age division based on water right issuance date assumes that the well was installed near the time that the water right was issued, which may not be true for all wells. Preliminary analysis shows that for irrigation wells, the older (pre-1975) wells do tend to have higher concentrations than newer wells. As Figure 8 shows, a higher percentage of newer (≥ 1975) wells have concentrations less than or equal to the background concentration of 3 mg/L. Further analysis of the distribution of age of water rights throughout the state may provide further insight as to the potential relationship that length of time farming in a given area has with the observed nitrate concentrations.

Figure 8. Comparison of age of irrigation wells with observed nitrate-N concentrations. Number of wells in each group shown above bars.

Variation of Nitrate Concentrations with Time

Increasing nitrate-N concentration in ground water over time is a growing concern. One way to evaluate possible increases is to compare data sampled from the same well at different time periods. Figure 9 shows a comparison of samples collected from the same irrigation wells in both the 1970's and the 1990's. The diagonal line in the figure represents the line of no change in concentration. If the values were the same in both time periods the point would fall on the line. The figure shows that many of the points fall above the line suggesting that the nitrate values have increased from the 1970's to the 1990's. This graph echoes Figure 4 in that increasing nitrate in ground water appears to follow the trend of increased irrigated farming in Kansas.

Figure 9. Comparison of nitrate-N concentrations from the same wells in the 1970's and 1990's. Values below line indicate decreased nitrate concentration from the 1970's to 1990's. Values above the line indicate increased nitrate concentration from 1970's to 1990's. Maximum Contaminant Level for nitrate-N in drinking water shown (10 mg/L). Most of the samples in graph show an increase in nitrate-N with time (Townsend and Young, 1999).

One potential problem with this graph is that these are single samples collected at different times from the same well. There is no continuous record available to help determine if there was a fluctuation due to rainfall, drought, season, change in land use practices, possibilities of spills etc. However, the utility of the figure is that as a snapshot it indicates that at least at these locations nitrate-N concentration has generally increased over time.

Nitrogen-15 Isotope Method for Identification of Nitrate Sources

Nitrogen-15 isotope values are used to determine the sources for nitrate-nitrogen in water (Townsend et al., 1994; Kreitler, 1975; Kreitler, 1979; Gormly and Spalding, 1979). The method uses the relation between nitrogen-14 (¹⁴N), which is common in air, and nitrogen-15 (¹⁵N), which is present at approximately 0.3% in air, as an indicator of sources of nitrogen for the measured nitrogen in water. The following equation shows the ratio used in the method:

In this equation, the standard is air; the sample is the water to be tested. The ‰ sign means parts per thousand or "per mil" as researchers using isotopic methods commonly report the units.

Figure 10 shows the ranges of nitrogen-15 observed for different sources of nitrogen and processes discussed previously. Heaton (1986) compiled the data from various studies. The figure shows that fertilizer ranges generally from δ¹⁵N of -2 to +8 ‰ while animal waste ranges from δ¹⁵N of +9 to greater than +20 ‰. Denitrification, the process of breaking down nitrate by bacteria to form nitrogen gases, also results in a δ¹⁵N range of +10 to greater than +20 ‰ in ground water. The concentration of observed nitrate-N is a strong indicator of whether animal waste or denitrification is the cause of the observed high δ¹⁵N value. As indicated in Figure 10, animal waste sources generally result in relatively high nitrate-N concentrations and denitrification processes generally result in relatively low nitrateN concentrations.

Figure 10. Range of δ¹⁵N values from selected references (after Heaton, 1986).

Figure 11 shows the distribution of nitrate-N and δ¹⁵N values for the samples collected in Kansas. The majority of wells shown in the figure have nitrate-N values greater than 3 mg/L (background level shown by gray dashed line). Most of the non-irrigation wells, particularly those with high nitrate-N concentrations, fall within the animal and human waste area of the graph (above +10 ‰). Many of these samples were collected from site study areas where elevated nitrate-N concentrations occurred.

Figure 11. Graph shows ranges of δ¹⁵N and nitrate-N (mg/L) values from areas in Kansas. Possible sources of nitrate are indicated: fertilizer and animal waste. Denitrification is illustrated as a sink for nitrate removal. Background and drinking water limits for nitrate are also indicated.

The majority of irrigation wells with nitrate-N above 10 mg/L plot within the fertilizer source range (+8 ‰ or less). Samples from wells that have nitrate-N less than 3 mg/L but an enriched (high positive value) δ¹⁵N value may indicate that denitrification of nitrogen by bacteria has occurred. The areas in the state where this has occurred, particularly in Harvey County, generally have reducing water chemistry. Samples with concentrations between 3 and 10 mg/L showed either one or the other of the sources or mixed/combined sources.

In summary, nitrogen-15 values indicate that 29% of all samples indicate a fertilizer source, 22% of all samples indicate mixed sources, 42% of all samples indicate an animal waste source, and 7% of all samples indicate denitrification processes have occurred. Nearly all the samples with δ¹⁵N values in the animal waste range are all from domestic and public water supply wells with high nitrate-N concentrations. Also, wells with the highest nitrate-N concentrations were non-irrigation wells with animal waste sources, which is indicative of point-source contamination typically from either septic systems or confined animal operations. Nitrogen-15 results indicate those irrigation wells with nitrate-N values greater than 10 mg/L typically have a fertilizer source for nitrate.

Conclusions

This preliminary assessment of nitrate in Kansas ground water indicates that much of the ground water has indications of nitrate present above the background level of 3 mg/L. The presence of measurable nitrate indicates that human activities are having an impact on the overall ground-water quality in the state.

South-central Kansas has higher mean and median nitrate-N concentrations than western Kansas, at least in part because of the sandier soils and shallower aquifer in south-central Kansas. Sampled wells in north-central and northeastern Kansas wells are sited in alluvium and are generally shallower than those in other parts of the state. Nitrate values from this area are higher than other parts of the state partly due the to permeable alluvium, higher precipitation, and to the possibility of a higher concentration of point sources such as septic systems or small farmstead feedlots located near wells, and differences in farming practices.

As expected, shallower wells in general had higher nitrate-N concentrations than deeper wells. Also, shallower wells had a greater percentage of nitrate-N above 3 mg/L, suggesting that the shallower portions of the aquifers within the state are more prone to contamination than the deeper portions of the aquifers.

Age of well may be an indicator of both well-construction practices used and length of time farming has occurred at a site. Older wells tend to have gravel pack to the surface and no grout across clay or perching zones, thus permitting contaminated surface and/or perched water to migrate down the well bore to ground water. Irrigation wells also may have multiple screen intervals that can permit mixing of different quality of water. In this assessment, older (pre-1975) irrigation wells generally had higher concentrations than newer wells and a higher percentage of the older irrigation wells had nitrate-N concentrations above 3 mg/L, suggesting that age of well may be an indicator of areas with potential problems.

Older irrigation well water right numbers give an indication of how long farming has occurred in different parts of the state. Further analysis of the distribution of age of water rights throughout the state may provide further insight as to the potential relationship that length of time farming in a given area has with the observed nitrate concentrations. Future work will also examine other relations within the individual regions, which might be indicators of potential contamination.

Nitrogen-15 values from non-irrigation wells typically indicate an animal waste source for nitrate: either septic systems or farmstead feedlots. Approximately 44% of the non-irrigation wells tested for nitrogen-15 had values in the animal waste range. This suggests that point source contamination may be more common, especially for shallower wells, than was previously thought. Nitrogen-15 from irrigation wells suggests a fertilizer source for nitrate, which may be a broader indication of nonpoint source contamination in areas where crop production dominates.

Because this is a preliminary assessment only a limited number of factors were considered. Future work will examine the effects of differences in precipitation and unsaturated zone characteristics as well as potential differences in land use (such as fertilizer and irrigation rates), occurrence of feedlots, and other potential point sources on the distribution of nitrate in Kansas ground water.

Recommendations

To prevent against further nitrate contamination from anthropogenic sources, proper management of all nitrate sources is essential. Many recommendations for fertilizer and irrigation water management are available from Agricultural Extension offices. Some of these recommendations include: split application of fertilizer, preplanting soil testing to determine nitrogen load, accounting for nitrate content in irrigation water as part of the fertilizer used during the growing season, and evaluating the efficiency of water application rates to prevent excess leaching of applied fertilizer.

To reduce the risks for point-source contamination, wells must be designed and installed properly. Also recommended are: proper plugging of abandoned wells, storing and mixing chemicals and fertilizers away from wells, and locating areas of livestock confinement away from wells (and vice versa).

References

Conover, W.J., 1980, Practical Nonparametric Statistics. 2nd Edition: John Wiley and Sons, New York, 493 p.

Gormly, J. R. and Spalding, R. F., 1979. Sources and concentrations of nitrate-nitrogen in ground water of the central Platte region, Nebraska: Ground Water, 17:291-301.

Hallberg, G. R., 1989, Nitrate in ground water in the United States. in Follet, R.F. (ed.), Nitrogen Management and Ground Water Protection: Developments in Agricultural and Managed-forest Ecology 21, Elsevier, New York, p. 35-74.

Heaton, T. H. E., 1986, Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: a review: Chemical Geology, 59:87-102.

Helsel, D. R., and Hirsch, R. M., 1992, Statistical methods in water resources: Studies in Environmental Science 49, Elsevier, New York, p. 357-376.

Herbel, M. J. and Spalding, R. F., 1993, Vados zone fertilizer-derived nitrate and δ¹⁵N extracts: Ground Water, 31:3:376-382.

Kansas Department of Agriculture, 1997, Kansas Farm Facts: Topeka, KS., 118 p.

Kolpin, D.W., Burkart, M.R., and Thurman, E.M., 1994, Herbicides and nitrate in near-surface aquifers in the midcontinental United States, 1991: U.S. Geological Survey, Water-Supply Paper 2413. [available online]

Kreitier, C. W., 1975, Determining the source of nitrate in groundwater by nitrogen isotope studies: Texas Bureau of Economic Geology, Report Investigations, No. 83. 57 p.

Kreitler, C. W. and Jones, D. C., 1975, Natural soil nitrate: the cause of the nitrate contamination of groundwater in Runnels County, Texas: Ground Water, 13:53-61.

Kreitler, C. W., 1979, Nitrogen-isotope ratio studies of soils and groundwater nitrate from alluvial fan aquifers in Texas: Journal Hydrology, 42: 147-170.

Madison, R. J., and Brunett, J. O., 1985, Overview of the occurrence of nitrate in ground water of the United States; in, National Water Summary 1984: U.S. Geological Survey, Water-Supply Paper 2275, p. 93-105.

SAS Institute., 1996, Master index to SAS system documentation, Version 6.12: SAS Institute, Gary, N.C.

Siegel, S., 1956, Nonparametric Statistics for the Behavioral Sciences: McGraw-Hill Book Co., New York. 312 p.

Spalding, M. E., 1984, Implication of temporal variations and vertical stratification of groundwater nitrate-nitrogen in the Hall County special use area: Nebraska Water Resources Research Institute, v. 372906, 44 p.

Spalding, R. F. and Exner, M. E., 1993. Occurrence of nitrate in groundwater--a review: Journal of Environmental Quality, v. 22, p. 392-402.

Steichen, J. Koelliker, J. Grosh, D. Heiman, A. Yearout, R. and Robbins, V., 1988, Contamination of farmstead wells by pesticides, volatile organics, and inorganic chemicals in Kansas: Ground Water Monitoring Review, v. 8, no. 3, p. 153-160.

Townsend, M. A., Macko, S. A., Young, D. P., and Sleezer, D. O., 1994, Natural ¹⁵N isotopic signatures in ground water: a cautionary note on interpretation: Kansas Geological Survey, Open-file Report 94-29, 24 p.

Townsend, M. A. and Young, D. P., 1999, Nitrate in Kansas ground water: Kansas Geological Survey, Public Information Circular 14, 4 p. [available online]

U.S. Census Bureau., 1999, 1990 Population Census Data; http://www.census.gov/.

Wilson, B. B., 1998, Water Information Management and Analysis System (WIMAS), version 4, for ArcView--Users Manual: Kansas Department of Agriculture, Division of Water Resources, 35 p.

Appendix A

Chemical and Physical Data for Kansas Ground-Water Samples, 1990-1998