Illinois
Fertilizer Conference Proceedings
January 25-27, 1993
| Home | 1993 Index | Search |
| Home | 1993 Index | Search |
R.M. Vanden Heuvel, R.L. Mulvaney, D.P. McKenna, and S.C. Schock1
Agricultural chemicals are a concern for the rural population since 90 percent of the rural population relies upon ground water for domestic use. With the threat of increasing NO3- levels in these supplies, the need for assessing the contribution from various sources, such as fertilizer N, soil organic matter, and animal wastes, becomes essential before remedial N management decisions can be made to reduce environmental contamination. Identification of these sources and their relative influences is often very difficult. Traditionally, field history of past N management practices has been the best indicator of contributing N sources. A non-traditional method, involving the use of natural nitrogen isotopes, has also been used but is controversial.
A two-year, interdisciplinary project was begun in March of 1990 to study and evaluate the validity of natural abundance 15N techniques (δ15N values) to identify and quantify sources of nitrate nitrogen (NO3--N) contaminating ground water. This is a cooperative effort between the University of Illinois Agronomy Department, Illinois State Geological Survey (ISGS), and Illinois State Water Survey (ISWS). The Illinois study is a companion study to a similar study currently being conducted on field lysimeters in North Dakota.
The desire to assess the contribution of fertilizer N to surface and ground water NO3- levels on a larger scale, particularly entire watersheds, has prompted research efforts relying on differences in natural abundance 15N levels of fertilizer and NO3- mineralized during soil incubations (Kohl et al., 1971). These techniques are based on the hypothesis that there is a consistent and measurable difference in the natural 15N concentration (δ15N) of NO3- derived from commercial fertilizers, manure, and soil organic matter. These differences are then used to identify and quantify contributing sources of NO3-. This approach has been criticized by a number of workers (Hauck et al., 1972; Edwards, 1973; Bremner and Tabatabai, 1973; Edwards, 1975; Meints et al., 1975; Broadbent et al., 1980) for several reasons, but most notably for the difficulty in establishing a single background δ15N value for NO3-- In general, the work performed by these authors has shown extensive spatial variability of δ15N values for soil-derived NO3-for small incubated samples taken from surface soils or soil cores. Despite the body of evidence that has accumulated against the approach, a number of investigations have apparently used the δ15N technique successfully to identify various sources of NO3- in ground water and estimate their relative contributions (Kreitler and Jones, 1975; Kreitler et al., 1978; Gormly and Spalding, 1979). In addition, more recent work has been performed that indicates the variability of δ15N values in soil is not a prohibitive factor in their use for studying N cycling in soils (Karamanos et al., 1981; Selles et al., 1986).
To resolve this apparent conflict, we are conducting four different studies
to evaluate the methodology. Only the Illinois well water study will be discussed
here.
The State of Illinois has funded a study to determine the occurrence of NO3- and pesticides in rural, private water supply wells. As part of that study, the State Geological Survey and the State Water Survey has selected for sampling, after an extensive screening process, 240 private wells in five areas. Forty-eight wells from each of five township-size areas were selected for the year-long sampling period. Subsamples from these water samples were analyzed for their variation in δ15N values. The study areas are located in Mason, Kankakee, Livingston, Piatt, and Effingham Counties and represent a diversity of hydrogeological conditions. Descriptions of the five areas are as follows:
Mason County Heavily irrigated, the top of the sand and gravel aquifer in the study area is within 5 feet of the surface. Most soils are Mollisols with about 85% of the area being planted to corn and soybeans. Most of the private wells are shallow.
Kankakee County The depth to the dolomite aquifer varies from 5 to 20 feet. Soils are mostly Mollisols and are cropped to corn and soybeans. Less than 10% of the area is irrigated. Well depths in the area are generally 40 to 125 feet deep.
Livingston County Depth to the sand and gravel aquifer is about 34 feet. Soils are mostly Mollisols under corn and soybeans with very little irrigation. Well depths are from 35 to 40 feet.
Piatt County The uppermost sand and gravel aquifer is found between 70 and 90 feet from the surface. Mollisols predominate here and are under corn and soybean production. Well depths range from 60 to 195 feet. No irrigation occurs in the study area.
Effingham County Soils here are exclusively Alfisols, mostly under non-irrigated corn and soybean production with about 30 % of the land under forest. The depth to the aquifer is greater than 50 feet. Well depth varies from 20 to 70 feet.
Wells were sampled biweekly far a one-year period. Characterization of the five study areas, including land use, agricultural practices, agricultural chemical usage, soils, hydrogeology, and hydrology were conducted by the ISGS and ISWS.
Information is available about such items as well proximity to barnyards or septic systems, well type, depth and construction, and past manure and N management practices. This and other pertinent information that can be related to N isotope data were collected by the ISGS and ISWS.
Calculation of δ15N Values
The 15N content of the NO3- in the water samples is expressed as:
δ15N = [%15N (sample) - %15N (standard) / %15N (standard)] X 1000
The range of δ15N values in water is generally about 0-22. According
to the literature on this methodology, as the value decreases, the level. of
NO3- originating from commercial fertilizer increases.
The NO3--N and δ15N (delta values) for all well samples is given by county in Table 1. Effingham county had the highest number of occurrences (concentrations > 0.01 ppm NO3--N) while Piatt county had the least. Three counties, Effingham, Kankakee, and Mason averaged more than 10 ppm NO3--N. Previous work has indicated that high NO3- concentration accompanied by NO3- δ15N values > 10 are indicative of manure as a contaminating source. Denitrification also creates high δ15N values but at the same time decreases NO3- concentration. Effingham and Livingston county averaged more than 10 δ units (15.1 and 18.8, respectively, Table 1). The high δ value for Effingham is most likely not an indication of manure contamination but rather an indication of denitrification considering that the soils in this county generally have a claypan in the soil profile. This is supported by the fact the δ values were as high for wells where there was no evidence of manure use in the surrounding area as for those where manure was present (17.4 vs. 17.9, respectively, Table 2). Moreover, where the δ15N data for Effingham were plotted vs. NO3--N concentration the δ15N value decreased with increasing NO3--N concentrations (figure 1). This trend is the opposite of what would be expected if manure were the source of NO3-. Where no manure was present, there was essentially no relationship between δ15N and NO3--N (figure 2). These data in particular indicate the confusion that can easily arise when trying to determine the source of NO3- based solely on δ15N values. There is no clear indication what the primary source of NO3- was for the Effingham samples.
Livingston county had relatively low NO3- levels (5.4 ppm) but had an average δ15N value of 18.8 (Table 1). Two sites from this county with previous evidence of manure, contributed to the extremely high δ15N average (30.0 units, Table 2). These sites were presumably contaminated from manure because of the significant number of animals within close vicinity of the well. Although the δ15N value is very high, and suggests manure contamination, there is no way to determine the contribution by denitrification to the high δ values. Isotopic data from sites with documented N management history should be considered only supplemental in nature. Such isotopic data without documented N management history should be considered highly speculative at best. Piatt county is not considered in Table 2 because of the low number of water samples with adequate N for analysis.
The correlation coefficients (r values) between NO3-
concentration and δ15N values for the
counties are given in figures 3, 4,
5, and 6.
All values were negative, and three of the four were relatively high (-0.47,
-0.32, and -0.39 for Kankakee, Livingston, and Mason counties, respectively).
These trends may represent either denitrification occurring in the groundwater
or leaching of commercial fertilizers (depleted 15N). Given the characteristics
of the soils in each county, denitrification was likely responsible for the
relatively high r values in Kankakee and Livingston counties. The sandy soils
in Mason county would be expected to allow leaching of fertilizer N. The use
of these figures and r values in this study did' not indicate any information
about contaminating sources of NO3- beyond what would
be gathered from general site characteristics. No relationship was found between
δ15N values and water well depth for this
study (figure 7). Previous studies by other
workers have used these types of relationships to draw firmer conclusions about
sources of NO3- and biological processes occurring in
groundwater.
Results from this study indicate that use of δ15N
techniques can easily lead to confusion as to contaminating sources of nitrate
as was observed in Effingham county comparing sites where manure was, and was
not, present near drinking wells. General soil and site characteristics are
at least as good an indicator of contaminating NO3- sources
as δ15N values. Other work we have conducted
indicates that extensive variability of natural, or background, soil δ15N values
can occur even in sandy soils. Such variability prevents clear identification
and quantification of the various possible sources of NO3-.
Table 1: Nitrate -N concentration and δ15N values for well water samples
Table 2: δ15N values for water samples when evidence of manure did, and did not, exist
Figure 3. 15N enrichment versus Nitrate concentration: Effingham County sites
Figure 4. 15N enrichment versus Nitrate concentration: Kankakee County sites
Figure 5. 15N enrichment versus Nitrate concentration: Livingston County sites
Figure 6. 15N enrichment versus Nitrate concentration: Mason County sites
Figure 7. Variation of 15N enrichment with well depth: all sites
Bremmer, J.M., and M. A. Tabatabai. 1973. Nitrogen-15 enrichment of soils and soil-derived nitrate. J. Environ. Quay. 2:363-265.
Broadbent, F.E., R. S.Rauschkolb, K. A. Lewis, and G. Y. Chang. 1980. Spatial variability of nitrogen-15 and total nitrogen in some virgin and cultivated soils. Soil Sci. Soc. Am. J. 44:524-527.
Edwards, A. P. 1973. Isotopic tracer techniques for identification of sources of nitrate pollution. J. Environ. Quay. 2:382-387.
Edwards, A. P. 1975. Isotope effects in relation to the interpretation of 15N/14N ratios in tracer studies. p. 455-468`. In, Isotope ratios as pollutant source and behavior indicators. International Atomic Energy Agency, Vienna.
Gormly, J. R.,, and R. F. Spald ng. 1979. Sources and concentrations of nitrate-nitrogen in ground water of the Central' Platte Region, Nebraska. Ground Water 17:291-301.
Hauck, R. D., W. V. Bartholomew, J. M. Bremmer, F. E. Broadbent, H. H. Cheng,
A. P. Edwards, D.R. Keeney, J. 0. Legg, S. R. Olsen, and L. K. Porter. 1972.
Use of variation in natural nitrogen isotope abundance for environmental studies:
a questionable approach. Science 177:4`53-454.
Kohl, D. H., G. B. Shearer, and B. Commoner. 1971. Fertilizer nitrogen: contribution to nitrate in surface water in a corn belt watershed. Science 174:1333-1334.
Karamanos, R. E., R. P. Voroney, and D. A. Rennie. 1981. Variation of natural N-15 abundance of Central Saskatchewan soils. Soil Sci. Soc. Am. J. 45:826-828.
Kreider, C. W., and D. G. Jones. 1975. Natural soil nitrate: the cause of the nitrate contamination of ground water in Runnels County, Texas. Ground Water 13:53-61.
Kreitler, C. W., S. E. Ragone, and B. G. Katy, 1978. Nls/N14 ratios of groundwater nitrate, Long Island, New York. Ground Water 16:404-409.
Meints, V. W., L. V. Boone, and L. T. Kurtz. 1975. Natural 15N abundance in soils, leaves, and grain as influenced by long term additions of fertilizer N at several rates. J. Environ. Qual. 4:486-490.
Montgomery, B. R., L. Prunty, and J. W. Bauder. 1987. Vacuum trough extractors for measuring drainage and nitrate flux through sandy soils. Soil Sci. Soc. Am. J. 51:271-276.
Selles, F., R. E. Karamanos, and R. G. Kachanoski. 1986. The spatial variability of nitrogen15 and its relation to the variability of other soil properties. Soil Sci. Soc. Am. J. 50:105-110.
