Nitrate contamination of groundwater is becoming a widespread problem in California. To evaluate the utility of stable nitrogen isotopes for identifying sources of nitrate contamination, nitrogen isotope ratios (?¹⁵N) were measured on nitrate extracted from core samples taken below natural, fertilizer, on-site sewage disposal (septic) and animal sources in the Sacramento and Salinas valleys. The mean ?¹⁵N value from natural sources was not significantly different from that of fertilizer sources. The mean ?¹⁵N value from animal sources was significantly different from that of septic sources and natural and/or fertilizer sources. Nitrogen isotope ratios tend to be site specific and should be measured below suspected sources in the subsurface and in groundwater.

[View Enlargement]

Soil samples were obtained by drilling 10 feet below the water-table surface.

Nitrate (NO₃) concentrations higher than the drinking water standard in groundwater are common in many parts of the world. Multiple sources of NO₃ contamination generally occur within the same watershed or groundwater basin. Sources include animal feedlots, horse corrals, dairy waste lagoons, manure applied to land, municipal sewage effluent, on-site sewage disposal systems, urban and agricultural fertilizer, natural soil organic matter and, in some cases, geologic sources. It is often difficult to ascertain which of these sources in a watershed may be contributing significant amounts of NO₃ to groundwater.

The nitrogen (N) isotope method of distinguishing sources is based on measuring the two stable isotopes of N (¹⁴N and ¹⁵N) in NO₃ of the sample. A common way of representing these ratios is delta ¹⁵N (?¹⁵N), which is the difference between the ratio ¹⁵N/¹⁴N of a sample and that of a standard (in the atmosphere) divided by the ratio of the standard, with the entire quantity multiplied by 1,000. This results in a ?¹⁵N value that typically varies less than several per mil (%₀) from that of the standard. The percentage of the two isotopes is nearly constant in the atmosphere at 0.366% ¹⁵N. However, because of the slight difference in atomic mass of the two isotopes, certain chemical and physical processes often preferentially utilize one isotope, causing a relative enrichment of that isotope in the product and a relative enrichment of the other isotope in the remaining reactants. Because of these isotopic fractionation processes, NO₃ from various N sources has been shown to have different N isotope ratios. This ratio in a sample depends on the series of reactions that formed the N compound and the composition of its precursors. If the sample is enriched in ¹⁵N relative to the standard, then the ?¹⁵N value is positive, and if the sample is depleted in ¹⁵N, then the ?¹⁵N value is negative. From literature sources, the ?¹⁵N values of NO₃ from soil organic nitrogen, fertilizer, animal waste and septic tank effluent range from approximately +2 to +8%₀, -3 to +2%₀, +9 to +25%₀ and +6 to +25%₀, respectively.

This study evaluates the use of N isotope ratios as a tool for identifying NO₃ contamination sources in groundwater for certain California conditions. The ?¹⁵N values in soil water and groundwater were measured in vertical profiles directly beneath various sources of NO₃ contamination (for example, fertilized fields, animal waste sites, septic systems and unfertilized, uncultivated land), from the source to the groundwater.

Sampling in Davis, Salinas Valley

Two study areas were chosen for field sampling, based on differing hydrogeologic conditions and ease of access for drilling. These areas were in the southern Sacramento Valley, where all sampling was performed in the vicinity of Davis, and in the Salinas Valley. Each study region contained sampling sites representing four different sources of NO₃: natural soil organic matter, inorganic fertilizer, animal feedlot/dairy and septic tank effluent. Soil and subsurface sampling was accomplished by drilling and sampling with an 8-inch-diameter, continuous-flight, hollow-stem auger and with a 2.5-inch-diameter California split spoon sampling ahead of the auger. Cores were taken continuously from the ground surface to the water table. Drilling continued into the water table to a depth of 10 feet below the water-table surface whenever possible. A water sample was collected from the bottom of each borehole upon reaching groundwater, using a bailer inside the auger. Subsamples of the core were taken approximately every 5 feet and were preserved for analysis. After the core was taken in the field, the sample was placed in a polyethylene zip-lock bag and placed in an ice chest filled with dry ice. The sample was then transferred to a freezer until it was processed.

Nitrogen isotope analyses

Nitrate in the soil samples was extracted with deionized water (1:5). Samples that were still cloudy after vacuum filtering were centrifuged and the decantant refiltered. A small amount of the thawed soil was taken to determine moisture content.

Steam distillation was used to convert the inorganic forms of N in the soil-water extracts to a stable form of ammonium salt needed for N isotope analysis. These samples were then sent to a laboratory for N isotope analysis. Standard deviation of 6 to 12 samples of reference standards was no larger than 0.8 ?¹⁵N units. Multiple extractions of field samples were generally within this same range. Thus we do not consider samples significantly different if they fall within ±1 ?¹⁵N unit.

NO₃ varies with location, depth

We drilled 26 boreholes during this study. Results are shown in terms of NO₃ concentrations and ?¹⁵N values versus depth for the Davis area (figs. 1 and 2) and the Salinas Valley area (figs. 3 and 4). Some gaps occur in the ?¹⁵N values due to insufficient NO₃ to conduct isotope analysis. Concentrations of ammonium, chloride and sulfate were also determined and used in the interpretation, but are not plotted here. In general, ammonium concentrations were negligibly small.

NO₃ concentrations vary considerably with location and depth owing to differing land uses, soil-water fluxes and geochemical processes (figs. 1 and 3). Most of the data represent nitrate in soil-water extracts expressed as ?g/g oven-dried soil.

As expected, NO₃ concentrations beneath the natural sources were low. NO₃ concentrations beneath the septic sources also tended to be low. This could be caused by the boreholes not intercepting the main discharge zones of the on-site sewage effluent, which can seep from the leach lines or pits in a spatially and temporally erratic fashion. Nevertheless, we are confident that our sampling captured enough NO₃ to register ?¹⁵N signatures of the various sources. The high NO₃ concentrations in shallow groundwater beneath the fertilizer, feedlot and septic sites in the Davis area and beneath the fertilizer and feedlot sites in the Salinas Valley could reflect both local and off-site sources.

[View Enlargement]

Fig. 1. NO₃ concentrations with depth for inorganic fertilizer, animal feedlot, septic tank (on-site sewage disposal) and natural soil organic matter sites in the Davis area (southern Sacramento Valley). Solid symbols represent balled groundwater samples. Soil-water extract concentrations are in ?g/g oven-dried soil, and groundwater concentrations are in ppm.

[View Enlargement]

Fig. 2. ?¹⁵N concentrations with depth for inorganic fertilizer, animal feedlot, septic tank (on-site sewage disposal) and natural soil organic matter sites in the Davis area (southern Sacramento Valley). Solid symbols represent bailed groundwater samples.

[View Enlargement]

Fig. 3. NO₃ concentrations with depth for inorganic fertilizer, animal feedlot, septic tank (on-site sewage disposal) and natural soil organic matter sites in the Sailnas Valley. Solid symbols represent bailed groundwater samples. Soil-water extract concentrations are in ?g/g oven-dried soil, and groundwater concentrations are in ppm.

[View Enlargement]

Fig. 4. ?¹⁵N concentrations with depth for inorganic fertilizer, animal feedlot, septic tank (on-site sewage disposal) and natural soil organic matter sites in the Salinas Valley. Solid symbols represent bailed groundwater samples.

Natural soil organic matter

sources

?¹⁵N values beneath the unfertilized, natural soil organic matter sites were very consistent among the borings in the unsaturated zone and varied between about zero and +5%₀ (figs. 2 and 4). The ?¹⁵N values in the groundwater samples for the Salinas site were all too high for a natural organic matter source and suggest a likely non-natural source in the up-gradient direction.

On-site sewage disposal

sources

?¹⁵N values from the Davis on-site sewage disposal (septic) site (fig. 2) showed an increasing trend with depth, but were well below the largest ratios expected for animal waste. This borehole was drilled between two leach lines spaced 25 feet apart. We suspect that water from the leach lines did not move far enough horizontally to intersect the borehole location until a substantial depth was reached; this would explain the low concentrations of NO₃ and low ?¹⁵N values in the shallow section. At the Salinas site (fig. 4), the borehole was drilled beside a deep vertical sewage disposal pit. Except for one very high ?¹⁵N value at the Salinas site, most of the ?¹⁵N values from the human waste disposal systems lie between +5 and +10%₀.

Animal waste sources

At the Davis animal waste site, a horse corral that had been in use for at least 30 years, the ?¹⁵N value was consistent at about +10%₀ throughout the vertical subsurface profile (fig. 2). The groundwater sample showed a slightly lower ?¹⁵N value than the soil water. At the Salinas Valley animal waste site, three boreholes were drilled in two of four abandoned dairy waste evaporation ponds and the animal feeding pens. The borehole with the most consistent ?¹⁵N values of the three boreholes had a nearly constant value of about +17%₀ in the upper 50 feet of the profile (fig. 4). Below 50 feet, the ?¹⁵N values decreased to around +6%₀; however, the other two boreholes at this site showed high ?¹⁵N values both deep and shallow. The spatial variability of ?¹⁵N values at this site can be attributed to subsurface heterogeneity and to the fact that the site contains several point sources of animal waste.

Fertilizer sources

Five boreholes were drilled at the Davis site and 10 at the Salinas site. Figures 2 and 4 give examples of ?¹⁵N values with depth for the two sites. The Davis site has been cropped with corn, alfalfa and sugar beets for many years. The Salinas site is in an irrigated agricultural field that has been in a mixed vegetable cropping system for several decades. Although most of the fertilizer applied was inorganic N, some compost containing animal waste was also applied occasionally at the Salinas site. The ?¹⁵N values for these sites were nearly constant with depth, consistent with the long-term use of fertilizer. The constant ?¹⁵N values with depth also indicate that denitrification in the subsurface must have been small. The cyclic high-low trend in NO₃ concentrations beneath the Salinas site (fig. 3) is due to geologic layering of the subsurface profile at this location, with the lower values tending to locate in the coarser-grained sediments. The lack of similar cyclicity in ?¹⁵N at the same borehole (fig. 4) attests to the independence of ?¹⁵N values with respect to concentration.

Comparisons of different

sources

Table 1 shows statistical comparisons of the differences in mean ?¹⁵N values for the various source types and locations. These statistics are based on all of the ?¹⁵N data, including those boreholes not shown in figures 1 through 4. The mean ?¹⁵N value for the agricultural (fertilizer) source was not significantly different from the mean of the natural background source at the 95% confidence level. The animal sources were significantly different from the septic sources and the fertilizer and natural sources.

In addition, we made statistical comparisons between mean ?¹⁵N values between areas for each source type. There was no significant difference in means for the natural, septic and animal waste sources between the Davis and Salinas areas, but means for the fertilizer sites were significantly different. The elevation of mean ?¹⁵N values at the Salinas fertilizer site over those at the Davis fertilizer site may be partly due to the occasional application of compost containing animal waste at the Salinas location and/or to a significant amount of nitrogen-fixing legumes in the cropping history of the Davis location.

Mixed sources of groundwater N

The ?¹⁵N data shown in figures 2 and 4 are chiefly from the unsaturated zone directly beneath known source types. In groundwater investigations of NO₃ contamination, however, much of the ?¹⁵N data will represent deeper groundwater pumped from production wells. If this deeper groundwater contains NO₃ contamination from more than one source, the ?¹⁵N values will reflect a mixture of those sources. Therefore it is typically not possible to “fingerprint” the source of NO₃ contamination by merely measuring ?¹⁵N in a few water wells. Clearly, determining the sources of NO₃ contamination in groundwater commonly requires careful hydrogeologic analysis of groundwater migration rates and flow paths as well as an exhaustive inventory of potential NO₃ sources. In this analysis, the ?¹⁵N technique is an important tool rather than “the solution.”

Even if an area contains only one source of NO₃ contamination that is theoretically distinguishable by ?¹⁵N techniques, it is possible that the ?¹⁵N values measured in groundwater or deep unsaturated zone will be measurably higher than in the shallow zone directly beneath the source. The reason is that denitrification, the process whereby NO₃ is converted to N₂O and N₂ gases, can raise the ?¹⁵N values above those of the original source. Denitrification is most likely when anaerobic conditions occur in a biologically active soil zone, which in turn most likely appears when the water table is within a few feet of land surface. When the water table is shallow, there is greater likelihood that anaerobic conditions of the saturated zone extend sufficiently into the biologically active soil zone to drive denitrification. Clearly, in addition to hydrogeologic analysis, application of the ?¹⁵N technique requires collecting data on ?¹⁵N values beneath the sources of concern to detect whether denitrification is occurring and to verify the ?¹⁵N signatures of the sources. Additional measurement of ?¹⁵N values beneath septic sources is particularly needed because of the dearth of such data in the literature.

In this study, little to no evidence of denitrification was found, presumably because the water table was 40 to more than 100 feet below land surface at all sites, well below the soil zone. That is, ?¹⁵N values tended not to increase with depth, except beneath the Davis septic site (fig. 2) and in one of the boreholes drilled at the Salinas septic site. At the Davis septic site, the borehole did not intersect the plume until a depth of approximately 25 feet, which explains the increase in ?¹⁵N value with depth. At the Salinas Valley septic tank site, the ?¹⁵N value increased with depth and the NO₃ concentrations decreased with depth, which could be indicative of denitrification. In addition, a few locally elevated ?¹⁵N values were found at various sites, such as the deepest sample beneath the Salinas Valley fertilizer site (fig. 4). However, ?¹⁵N values from the other boreholes and from deeper groundwater verify that these elevated values are isolated anomalies, apparently caused by highly localized processes.

[View Enlargement]

TABLE 1. Statistical comparison of ?¹⁵N data for Davis Area and Salinas Valley

Isotope ratios as indicators

Nitrogen in the subsurface zone existed primarily as nitrate, except for one septic and one dairy lagoon site in the Salinas Valley where significant amounts of ammonium existed. The ?¹⁵N values determined from the various sites fall within the range of values reported in the literature, but with lower variances. The mean ?¹⁵N value for the natural background sites was about +2.6%₀ in both the Salinas Valley and Davis study areas. Mean ?¹⁵N value for the fertilized sites was slightly greater than for the natural sites with +4.4%₀ in the Salinas Valley and slightly lower at +1.6%₀ in the Davis area. The animal waste sites had a mean value of +10%₀ at Davis and +14%₀ in Salinas Valley. The on-site sewage disposal sites had mean ?¹⁵N values of about +7.3%₀ at Davis and +8.7%₀ at Salinas Valley. It is clear from our data that the ?¹⁵N methodology cannot be used to distinguish between fertilizer and natural soil organic matter sources. However, in the sites that we examined, the ?¹⁵N method can be used to distinguish between animal sources and fertilizer and/or soil organic matter sources. In addition, our data show that there is a significant difference between mean ?¹⁵N of the animal and septic sources. These results exemplify the necessity of measuring the specific ?¹⁵N of suspected sources of NO₃ contamination for the particular area of interest. Simply using values from the literature may not allow clear separation of source types.

The ?¹⁵N values at most sites were fairly consistent with depth from the surface to the water table. There is no evidence from our data that denitrification is a significant process at any of the sites, with the possible exception of the Salinas Valley septic tank site. Thus, except for one site in the eight main test sites, the results demonstrate that measuring the ?¹⁵N value immediately below the NO₃ source can be an accurate indicator of the fingerprint of that source and that, under the conditions prevailing at these sites, the fingerprint will not change much during NO₃ transport to groundwater. This is a very important conclusion for use of the N isotope technique to indicate sources of NO₃ in groundwater. Nevertheless, users of the ?¹⁵N approach should be aware of the potential for mixing of ?¹⁵N from multiple sources and of denitrification under some circumstances. Careful hydro-geologic characterization as well as sampling of both the unsaturated and saturated zones beneath potential sources are therefore typically required for successful application of the ?¹⁵N approach.

Return to top