| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Land, Air, and Water Resources, Univ. of California, Davis, CA 95616
b U.S. Geological Survey, 6000 J Street, Placer Hall, Sacramento, CA 95819
* Corresponding author (pjhernes{at}ucdavis.edu).
Received for publication April 10, 2007.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
4 to 8 mmol total THM/mol DOC) across the management areas, indicating the dependence of reactivity on DOC composition. The results indicate that land management strongly controls the quantity and quality of DOC to reach shallow ground water and hence should be considered when managing ground water resources and in any efforts to mitigate contamination of ground water with carbon-based contaminants, such as pesticides and pharmaceuticals.
Abbreviations: DBP, disinfection byproduct • Delta, Sacramento-San Joaquin River Delta • DO, dissolved oxygen • DOC, dissolved organic carbon • MCL, maximum contaminant level • OC, organic carbon • SOM, soil organic matter • STHMFP, specific trihalomethane formation potential • THM, trihalomethane • TOC, total organic carbon • TTHM, total trihalomethane • TTHMFP, total trihalomethane formation potential
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Many of the world's alluvial, unconsolidated aquifer basins are home to some of the most productive agricultural regions, which often include increasingly intensive animal feeding operations that heavily depend on the availability of large amounts of forage (UNESCO, 2006). Like in the San Joaquin Valley, many domestic and public drinking water wells in such basins are located in close proximity to animal agriculture and other agricultural activities that are potential sources of organic carbon, but also of organic contaminants (e.g., pesticides). In the USA, most (public) drinking water treatment facilities use chlorination as a means of disinfecting water for potable consumption and prevention of microbial growth during distribution. However, chlorine reacts with DOC to produce potentially harmful byproducts such as trihalomethanes (THMs), the amount of which is regulated in finished drinking water (USEPA, 1998). Thus, when ground water is a primary source for drinking water, it is important to understand the longevity of high concentrations of DBP precursors in ground water as a function of land use. The persistence of DOC in ground water and the associated DBP precursors depends on factors that enhance or inhibit degradation, including inherent bioavailability, the microbial population density and diversity, sorptive partitioning to the soils and subsoils, and the dissolved oxygen (DO) concentrations of that water (Chiou, 1987; Hornsby, 2003).
In surface waters, bioavailability studies highlight both the reactive and persistent pools of DOC (e.g., Stepanauskas et al., 2005). In ground water systems that are essentially closed to the atmosphere, the reactive or bioavailable pool represents the potential to consume all or part of the dissolved oxygen through degradation and subsequently provide a food source for denitrifying bacteria, whereas the persistent pool represents potential long-term water quality effects. The amount of DOC that is bioavailable is dependent both on DOC composition as well as microbial activity and community composition and associated factors. In many cases, microbial processing has been demonstrated to convert bioavailable DOC to refractory forms (Ogawa et al., 2001; Tranvik, 1993; Ziegler and Fogel, 2003). The bioavailability of carbon loaded to ground water may indicate how long organic compounds will persist in an aquifer and may highlight differences in the microbial community in the aquifer and overlying soils.
More than 30% of the water used in California during an average year comes from ground water (CDWR, 2005), and most of the organic carbon (OC) in typical ground water is in the form of DOC (Aiken and Kuniansky, 2002). Because of health concerns associated with DBPs, the CALFED Bay-Delta program has proposed a goal of 3 mg L–1 total organic carbon (TOC) (CALFED, 2000) for drinking water sources in the California Delta region. DOC concentrations in uncontaminated ground water are typically around 1 mg L–1; however, in some cases DOC concentrations can range to >15 mg L–1 (e.g., Christensen et al., 1998; Gron et al., 1996; Leenheer et al., 1974), as influenced by differences in the organic material in aquifer material and overlying soils (Gron et al., 1996; Wassenaar et al., 1991). Elevated ground water DOC usually results from anthropogenic activity, however, and the development of high DOC plumes has been documented under landfills and application of reclaimed wastewater (Christensen et al., 1998; Jensen et al., 1999; Repert et al., 2006). These studies concluded that the elevated DOC resulted in significant ground water quality impairment. In one case, even after the DOC source was eliminated, natural attenuation of DOC was estimated to take decades (Repert et al., 2006).
The potential influence of agricultural activities on water quality is significant. Within concentrated animal feeding operations including large dairies, the application of liquid manure to forage crops provides for wastewater disposal, fertilization, and irrigation. Land application of agricultural manure, which occurred on nearly 92,000 km2 throughout the USA in 2002 (USDA, 2006), has been shown to affect ground water, and in some cases surface waters, through ground water recharge (Bacchus and Barile, 2005). The objectives of this study were (i) to establish the impact of dairy-derived DOC on ground water quality, (ii) to determine whether the concentration of DOC in the ground water is related to different dairy management units, and (iii) to determine the extent to which differences in water quality parameters (DOC bioavailability, specific THM formation potential [STHMFP]) are related to management units.
|
Materials and Methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
The two dairies sampled during the 3-yr duration of this study are located approximately 40 km apart in the San Joaquin Valley of California. The area is characterized by alluvial deposits of the Tuolumne, Stanislaus, and San Joaquin Rivers, which are primarily composed of granitic material from the Sierra Nevada mountain range located to the east of the San Joaquin Valley. Most of the wells sampled are located on one dairy (primary dairy), which has low slopes (0–3%) and loamy sand to sandy soils. This dairy is shown in Fig. 1 and has been described by Harter et al. (2002) and Singleton et al. (2007), who denoted it MCD (Merced County Dairy). Two wells are located on the second dairy, which has 0 to 1% slopes and loamy sand soils (Harter et al., 2002).
|
|
|
Previous work demonstrated that the source area of water obtained from the monitoring wells forms a narrow, elongated area immediately upgradient of the well spanning from 150 m to a few hundred meters (Harter et al., 2002). This distance is typically less than the upgradient extent of adjacent forage crop fields. The source area length was determined by accounting for recharge rates to ground water, the long-term average ground water discharge rate, and the length of the screened interval in each well (Harter et al., 2002). Wells were classified based on the management unit in the source area immediately upgradient from the well. This ties DOC composition in well water with the type of land use dominating the source area specific to that well (Harter et al., 2002). We therefore refer to the monitoring wells by their dominant source area: corrals and freestalls (corral wells), manure-irrigated fields (manured field wells), and liquid manure storage ponds (pond wells). Sampling on the dairy also included tile drain outlets (tile drains) and a domestic supply well that draws water from a deeper, semi-confined aquifer.
Potential influences on background DOC concentrations in ground water include the DOC content of rain, snow in the Sierra Nevada Mountains, local surface waters, and the soil organic matter (SOM) content in recharge areas. Major local rivers have similar concentrations to rainfall in their upper reaches (1–2.5 mg L–1 DOC). As the rivers flow into the San Joaquin Valley, DOC concentrations increase to values between 2 and 5 mg L–1 DOC (Kratzer and Shelton, 1998). The local reaches of the San Joaquin reflect strong agricultural influence from return flows, with median DOC values of 6.5 mg L–1 DOC (Saleh et al., 2003). As surface or irrigation water passes through the vadose zone en route to the aquifer, soil DOC can be incorporated. Land use that significantly elevates SOM concentration can contribute to high soil DOC concentrations (2–30 mg L–1 DOC globally [Michalzik et al., 2001]). Local SOM concentrations vary over an order of magnitude, with variation partially due to differences in historical land use (R. Meyer, personal communication, 2006).
Both dairies are typical of those found in the northern San Joaquin Valley of California. Corrals (for exercise) and freestalls (for feeding) contain the cattle, and recycled wastewater removes manure via flush lanes adjacent to freestalls. Mechanical screens filter the sludge to separate liquid wastewater from large solids. The liquid is transferred to settling and storage ponds, which remain open to the atmosphere. Throughout the spring and summer, pond water is periodically removed from the ponds, diluted with irrigation water, and applied to forage crop fields.
The primary dairy (Fig. 1) is downgradient from agricultural land with similar manure management techniques. Eleven shallow monitoring wells were sampled at this site, including four corral wells (samples 1, 11, 13, and 14), two pond wells (samples 7 and 10), and five manured field wells (samples 5, 6, 8, 12, and 15). These wells are located over an 1.3 km2 area and are screened in the upper aquifer up to a depth of 7 m. A domestic supply well (sample 16) is screened in the deeper, semi-confined aquifer. Tile drains, which maintain water levels by draining multiple land management types, were sampled at the point of discharge to a local canal (samples 9 and 21). Raw wastewater from the holding pond located upgradient from well number 10 was sampled (sample 18). At this site, a tile-drain system removes ground water and discharges into nearby irrigation canals, which may contribute to flow in the San Joaquin River.
The area surrounding the second dairy is dominated by orchards that do not receive manure applications. A well located on the upgradient (southeast) side of this dairy serves as a background monitoring well, referred to in this study as the "upgradient" well (sample 20). A second monitoring well at this site, screened in the upper unconfined aquifer as at the first site, is located immediately downgradient from a wastewater pond (sample 19).
Sampling Protocol
Well water was sampled using a submersible pump, with five or more well volumes removed before collection. Water quality parameters, including temperature, pH, dissolved oxygen, and electrical conductivity, were monitored during removal and sampling with a YSI 556 Multi-Probe System (YSI, Yellow Springs, OH). Sampling took place on 29 May 2003, 7 and 8 Mar. 2005, 7 June 2005, 4 Sept. 2005, and 26 Mar. 2006. On site, all samples were passed through a 0.45-µm filter, collected into bottles triple-rinsed with well water, and kept in coolers on ice for 2 to 10 h.
Samples collected before September 2005 were frozen until analysis. Samples collected in September 2005 were analyzed within 24 h of sampling. Samples collected in March 2006 were acidified to a pH of <2 and refrigerated until analysis, a method shown effective for preservation of DOC on the order of weeks (Tupas et al., 1994). Freezing for long periods has been shown to have minimal effect on the DOC content of previously filtered samples, giving comparable results to samples preserved by acidification and storage at 4°C (Tupas et al., 1994). Pond water samples were frozen before filtration. Once filtered, samples were acidified and refrigerated until analysis.
Samples for determining STHMFP were collected once in March 2005 using amber bottles and were refrigerated at 4°C until analysis (<5 d). STHMFP is the molar ratio of total THM formation potential (TTHMFP) to DOC and indicates the potential yield of THM per mole of DOC, the compound class most commonly formed during chlorination (Saleh et al., 2003). Individual THM concentrations (µg L–1) are summed to give TTHMFP.
Samples for determining DOC bioavailability were collected once on 26 Mar. 2006. These samples were collected with zero head space and kept in coolers on ice for 2 to 20 h and transferred to a 4°C refrigerator. DOC bioavailability experiments were initiated within 48 h of sampling. Unfiltered water was collected at all wells for DOC bioavailability analyses. This water was used to inoculate the DOC bioavailability experiments (described below).
Laboratory Analyses
Total DOC values on filtered samples were determined by high-temperature combustion using a Shimadzu TOC-VCSH analyzer. Samples were prepared and analyzed automatically according to procedures described in the TOC-VCSH User Manual (Shimadzu Corporation, 2001). In accordance with this methodology, a five-point calibration curve (0, 0.1, 1.0, 10.0, and 100.0 mg L–1) with 99% or better correlation was used to determine the DOC concentration within samples and blanks. Two blanks of ultraviolet light–oxidized water (NANOpure, Dubuque, IA) were analyzed as samples with every batch, and the average value subtracted from the calculated DOC value of the samples (Benner and Strom, 1993). The TOC-VCSH automatically acidifies samples prior to DOC measurement using high-temperature (680°C) catalytic combustion. Samples are automatically analyzed three or more times to minimize standard deviation within replicates; the average of these values is reported. Standard deviations for these replicate injections were 
2.15% of the measured concentrations.
STHMFP analyses of DOC were conducted using the method described by Fram et al. (2002), a modified version of the U.S. Environmental Protection Agency (EPA) standard method 502.2 (1995). Briefly, water samples were dosed with sufficient Cl2 (derived from NaOCl) to result in a 1 mg L–1 residual, incubated at 25°C for 7 d, and quenched using sodium sulfite (Na2SO3). The chlorine dose was determined stoichiometrically from measured ammonia, sulfide, and DOC in the water sample (Fram et al., 2002). After quenching, samples were analyzed by purge and trap (Archon purge and trap with a Tekmar 3100 concentrator [Teledyne Tekmar, Mason, OH]) coupled to a gas chromatograph (HP 5890; Hewlett-Packard Company, Palo Alto, CA) equipped with a 30-m megabore DB-VRX column and an electron-capture detector. Concentrations of four THM compounds were quantified, including chloroform (CHCl3), bromodichloromethane (CHCl2Br), dibromochloromethane (CHClBr2), and bromoform (CHBr3). Seven-point standard curves were constructed with Supelco certified THM standard mixtures and used to quantify THM concentrations, which were summed as total THM (TTHM). The bromine in these compounds is derived from bromide already present in the ground water samples.
DOC bioavailability experiments (21 d) were conducted for a subset of wells, including a manured field well (sample 12), two corral wells (samples 13 and 14), two pond wells (samples 7 and 19), and the upgradient well (sample 20), using a modified version of the USEPA 5-d Biological Oxygen Demand standard method 5210 B (1986). For these experiments, filtered water samples were spiked with nutrients; sodium phosphate (NaH2PO4·H2O), ammonium chloride (NH4Cl), and sodium nitrate (NaNO3) were added to filtered well water for concentrations of 1.5, 7.5, and 7.5 mM, respectively. These amounts were selected to provide excess phosphate, ammonium, and nitrate to ensure that carbon was the limiting nutrient in the system. Samples were then inoculated with 10% unfiltered water from each respective well to provide a natural assemblage of microbes and shaken to ensure oxygen saturation and homogenization. The control sample for these experiments consisted of the nutrient suite added to carbon-free water and contained 0.03 mg L–1 DOC.
All DOC bioavailability samples were prepared in triplicate, wrapped in aluminum foil to exclude light, and placed in a water bath maintained at 20 ± 1°C. Samples were removed and acidified after 0, 2, 7, 14, and 21 d. DOC content was analyzed as described previously. DOC differences were calculated between 0 and 21 d and corrected for DOC depletion in the control, and the resulting value is reported here as DOC bioavailability (as percent of initial DOC). Standard deviations for the triplicates from the 0- and 21-d samples used in the bioavailability calculation were 1.5% of the measured concentration.
Statistical Analyses
Descriptive statistical analysis was conducted for DOC using two approaches. The first approach used a two-step process similar to that used by Harter et al. (2002). First, the average DOC was computed for each well across all time steps. Each average was then treated as an individual sample in a one-way ANOVA (Davis, 1986) for the effect of management units. Individual sampling data were not considered for the ANOVA in this approach because well sampling data collected over time were not statistically independent of each other, as discussed below. In the second approach, ANOVA was performed for all data points rather than for 3-yr averages.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
The concentration of DOC in the monitoring wells showed a statistically significant (p < 0.05; see Table 2 ) dependence on the associated management area. Pond wells had the highest DOC concentration (mean ± SD, 27 ± 13 mg L–1) among the wells (Fig. 2 ; Table 1). The second highest DOC concentrations occurred in wells recharged by corrals (17 ± 9 mg L–1), which receive fresh OC input, followed by manured field wells (8 ± 3 mg L–1), both elevated above local background levels as measured in the upgradient well (2 ± 1 mg L–1) (Table 1). Tile drains collect the shallowest ground water over an area of approximately one square mile, including the entire primary dairy facility. Tile drain samples exhibited higher DOC concentrations (10 ± 2 mg L–1) than manured field wells but lower DOC than corral wells.
|
|
Dissolved Oxygen Concentrations
Hypoxic conditions (typically defined as <3 mg L–1 DO) were observed during September 2005 and March 2006 in multiple manured field, corral, and pond wells (Table 3
). During September, both tile drains were hypoxic, and two corral wells (samples 11 and 14) and a pond well (sample 7) were anoxic (<0.5 mg L–1 DO). Dissolved oxygen in any given well frequently varied by a factor of two between sampling periods and was nearly ninefold higher in the anoxic pond well (sample 7) when sampled in March 2006. Overall, the concentration of DO was lower on average in September 2005 (1.50 mg L–1) than in the following spring (2.68 mg L–1 in March 2006) (Table 3). Dissolved oxygen and DOC in this study were inversely correlated (Fig. 3
; R2 = 0.45, y = 14.641x–0.9053) in March (R2 = 0.49) and September (R2 = 0.58).
|
|
|
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Ground water DOC concentrations observed in this study are also much greater than DOC concentrations measured in other ground water basins in California, including the northern San Francisco Bay area (15 wells within detection limits, median 0.45 mg L–1 DOC [Kulongski et al., 2006]), the San Diego Drainages (14 wells within detection limits, median 0.8 mg L–1 DOC [Wright et al., 2005]), and the San Jacinto Basins (26 wells within detection limits, median of 0.4 mg L–1 DOC near the coast, inland median 3.4 mg L–1 DOC [Hamlin et al., 2002]). All of the basin samples in these cases were retrieved in domestic and public supply wells, which are typically screened at depths of significantly greater than 15 m, much greater than the monitoring wells described here (7 m) but comparable to the domestic well.
The contrast between concentrations found in ground water beneath the dairy and those reported previously accentuates the impact of land use on the shallow-most ground water and the potential for exceptionally high DOC concentrations in the vicinity of large confined animal feeding operations. The significant temporal variability in DOC concentrations measured in this study indicate a highly dynamic system that seems to be capable of responding rapidly to changes in DOC loading and other conditions at the surface.
DOC Relationships to Land Use
Trends observed in DOC concentrations reflect the influence of land management units. The high concentration of DOC in pond wells (range, 13.7–54.7 mg L–1; mean, 27.5 mg L–1) likely reflects underground leakage previously recognized at this site (Harter et al., 2002; Singleton et al., 2007) and common in wastewater ponds (Ham and DeSutter, 1999). The second highest DOC concentrations (range, 7.9–38.6 mg L–1; mean, 16.9 mg L–1) occurred in wells recharged by corrals, which receive fresh OC input, followed by manured field wells (range, 4.4–15.2 mg L–1; mean, 7.6 mg L–1), which are elevated above local background levels and reflect infiltration from irrigation with diluted pond water (Table 1). The intermediate concentrations observed in tile drain samples (range, 5.3–12.4 mg L–1; mean, 10.0 mg L–1) likely derive from the fact that tile drains underlie the entire dairy farm operation and therefore contain DOC from manured fields and corrals.
Infiltration rates and carbon loading are likely the controlling factors on the concentrations of DOC beneath the dairy. Corrals are areas of concentrated OC loading but lack the constant presence of a fluid (such as a pond) to facilitate infiltration. Infiltration is also intermittent under fields, mainly by way of irrigation and rain events, as compared with beneath ponds where infiltration is potentially continuous when the pond is in use. Infiltration rates are high under the ponds (0.8 m yr–1 or greater) compared with the fields (0.45 m yr–1) (Harter et al., 2002; Singleton et al., 2007). DOC concentrations are highest in ground water influenced by ponds with associated high DOC concentration and infiltration rates. DOC concentrations decrease in ground water influenced by fields and corrals, as might be expected under lower DOC concentration and lower infiltration rates. Forage crops receive nonmanured irrigation water during some months, and the lower concentrations of field wells may be reflective of this decreased OC loading as compared with other management units.
In portions of the dairy with high DOC loading to ground water, anoxia may play a role in the elevated concentrations measured in this study—part of a positive feedback system in which high DOC loads drive consumption of DO, which in turn leads to slower rates of DOC degradation. Given the average depths of our wells (7 m), typical recharge rates (0.6 m yr–1), and a specific yield of approximately 15%, average ground water age and associated DOC in these samples is approximately 1 yr (range, 0–2 yr). However, when exposed to oxygenated conditions, our bioavailability experiments indicate that as much as 10% of the DOC can be degraded within 21 d. The high DOC concentrations and low DO levels associated with the pond wells (Tables 1 and 2) are consistent with the findings of Harter et al. (2002) and Singleton et al. (2007) that the pond wells are intercepting an extended anoxic plume originating from recharge at the bottom of the liquid manure pond.
Anoxic degradation is slower than aerobic degradation because microbial degradation of OC proceeds in part through anaerobic fermentation (a slower process than aerobic decomposition), organic compounds are less completely degraded, and biomolecules produced by microbial activity may be less prone to further oxidation (Anderson, 1995). In a persistent anoxic plume resulting from wastewater application, Repert et al. (2006) found the highest DOC concentrations in the core of the plume and predicted an extensive recovery time due to decreased microbial degradation rates under anoxia. Similar trends of high DOC concentrations within anoxic contaminant plumes under persistent carbon loading have been observed in many situations (Cozzarelli et al., 2001; Wersin et al., 2001). The inverse correlation of DO and DOC (Fig. 2) demonstrates the relationship between high carbon loading and DO consumption. The high DOC concentration and recharge rates of the dairy may, through anoxia, promote the retention of high DOC levels in the associated ground waters.
THM Formation Potential
The current maximum contaminant level (MCL) for THMs as established by the EPA is 80 µg L–1. Total THM formation potential values exceeded this MCL in every well measured, including the upgradient (140 µg L–1) and domestic wells (202 µg L–1). Perhaps more significantly to regional water quality, the tile drain measurement (761 µg L–1) was nearly an order of magnitude higher than the MCL. Tile drain waters come primarily from manured fields and discharge directly into surface waters that eventually feed into the San Joaquin River.
Typically, higher TTHMFP values are associated with higher DOC values (Fleck et al., 2004; Thurman, 1985), a reflection of the fact that natural samples contain a wider range of DOC concentrations (in this study nearly three orders of magnitude) than the associated STHMFP values (two-fold variation). The STHMFP range observed in this ground water study is relatively similar to values in the literature (Bergamaschi et al., 1999; CDWR, 1994) for surface water. The elevated STHMFP yields from dairy-influenced wells over the upgradient well demonstrates the importance of the manure management practices on DOC composition and hence STHMFP and water quality issues. Although aromatic compounds are widely believed to be the primary precursors for disinfectant THMs (e.g., Minear and Amy, 1996), other studies demonstrate that aromaticity alone cannot fully predict THM precursor content (Fujii et al., 1998). In theory, any unsaturated compounds are susceptible to halogenation, and it has been shown that carboxylic acids, ketones, amides, and select amino acids in addition to aromatic compounds are all precursors to DBPs (e.g., Kanokkantapong et al., 2006).
Although the precise precursors to THMs are unknown, the relationship of STHMFP to land use and associated DOC compositions is apparent. In this study, DOC from manured field well samples had the highest propensity to form THMs (Fig. 4), and, significantly, manured fields also occupy the greatest land area. Thus, the greatest potential impact of dairy-derived DOC on water quality seems to derive from the application of manure to fields. Because field recharge is associated with dairy waste that has undergone the most processing, the results suggest that the most persistent or diagenetically altered OC compounds are more likely to yield THMs. Thus, lower STHMFP values derived from pond wells may simply reflect a larger proportion of labile compounds that have not been degraded, due in part to the anoxic conditions beneath the ponds. In any case, our results indicate that best management practices aimed at minimizing the loading of these DBP precursors to ground water might best be focused on the application of manure to fields.
Bioavailable DOC
The variability in ground water DOC bioavailability values (3–10%) is consistent with previous studies of dairy farm ground water that showed variability in steroid, nitrate, and salinity concentrations (Harter et al., 2002; Kolodziej et al., 2004). Our ground water bioavailability values were comparable to previous measurements in surface waters, such as Ozark mountain streams in Arkansas that registered relative bioavailability ranging from 0.7 to 13.2% (Ziegler and Brisco, 2004). Within the Central Valley, average bioavailability at 13 surface water stations in the Delta over the course of a year was 10% (Stepanauskas et al., 2005). Although DOC bioavailability is dependent partly on the capabilities of the microbial population to use organic compounds within the DOC, the most significant factor is likely the chemical composition of the DOC. Bioavailable DOC from surface water can result from direct injection of fresh organic material via in situ algal production or leaching of DOC from wetlands and riparian vegetation. In contrast, ground water DOC undergoes selective partitioning between the dissolved phase as well as soils and subsoils, which can result in significant alteration of the DOC composition (Hernes et al., 2007). The range of bioavailability measured in this study indicates that individual wells vary both in terms of the DOC composition of the recharge water and the sorptive properties of the mineral surfaces along the hydrologic flowpaths feeding the wells.
DOC bioavailability can also be recast as biological oxygen demand using stoichiometries of respiration. Redfield ratios for respiration of organic matter predict the consumption of 138 O2 molecules per 106 organic carbon molecules (Redfield et al., 1963), and using these stoichiometries we calculated that respiration of 2 to 3 mg L–1 of DOC is sufficient to drive ground water to anoxia in this system. In wells in which bioavailability was measured, corral well 13 contained bioavailable DOC values (3.9 mg L–1) that exceeded the 2 to 3 mg L–1 needed to drive ground water to anoxia, pond well 7 contained enough bioavailable carbon (1.8 mg L–1) to drive the ground water to hypoxia, and the remaining four wells ranged from 0.1 to 0.8 mg L–1 bioavailable DOC. There was no clear statistical relationship between bioavailable carbon and measured DO in these wells, which is perhaps not surprising given the time and opportunity that DOC in these waters has had to degrade and undergo sorptive partitioning since recharge.
In addition to DO consumption, DOC bioavailability has relevance to the denitrification processes. Denitrification only occurs under anoxic conditions but has the additional requirement of an OC source that can be used by denitrifying bacteria to fuel their metabolism. Although denitrification in ground water at the depths of our monitoring wells (7 m) is not readily apparent, ground water from greater depths on these dairies has undergone considerable denitrification (Singleton et al., 2007). Using estimated stoichiometries of five OC molecules per four nitrates (with the oxygen content of organic matter as the primary uncertainty) (Galloway, 2003), the bioavailable DOC measured in this study seems to be sufficient for the denitrification of 0.6 to 16 mg L–1 of total nitrate. Denitrification and DOC degradation occur over much longer time scales than the 21 d in our bioavailability experiments, and clearly the total DOC and bioavailable DOC measured in shallow ground water demonstrate the potential to fuel a significant amount of denitrification.
Long-Term Impacts
This study demonstrates that land management strongly controls the quantity and quality of DOC to reach shallow ground water. The fact that 
90% of the DOC in these wells was not bioavailable indicates that DOC emanating from dairies could persist for years, if not decades. The direct impact on water quality is the potential delivery of high concentrations of DBP precursors to domestic water supplies. However, persistence of high DOC loads also has important implications for the degradation of other high-profile contaminants likely to be present in agricultural systems, such as pesticides, pathogens, and pharmaceuticals. Hypoxic, carbon-saturated ground waters such as these can be considered impaired in their ability to degrade these carbon-based compounds (Cozzarelli et al., 2001; Wersin et al., 2001). Steroid hormones have been shown to degrade more slowly under anaerobic conditions (Ying et al., 2003), a factor that may allow them to move further in an aquifer as an indirect consequence of high DOC loading. Two chemically distinct pesticides (mecoprop and phenoxic acid) were demonstrated to be more persistent in hypoxic conditions, and suggested that the redox conditions occurring in a pesticide's flow path are important to assessing the environmental risks of that chemical (Vink and van der Zee, 1997). Conversely, high DOC loading may contribute to the mitigation of high nitrate concentrations by fueling denitrification. This research indicates that long-term monitoring of DOC loading to ground water from agricultural operations such as dairy farms and its relationship to land use could be an important element for assessing the full impacts to water quality. Consistent with other ground water legacy issues in California, the true impact of this elevated DOC on ground water resources may not be known for decades.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
), specific trihalomethane formation potential (STHMFP
), and DOC bioavailability
(as percent consumed over 21 d).
