Ecohydrological Factors Affecting Nitrate Concentrations in a Phreatic

Apr 17, 2008 - Most groundwater NO3− appears to be depleted relative to Cl− in rainfall concentrated by evapotranspiration, indicating net N losse...
0 downloads 0 Views 7MB Size
Environ. Sci. Technol. 2008, 42, 3531–3537

Ecohydrological Factors Affecting Nitrate Concentrations in a Phreatic Desert Aquifer in Northwestern China J O H N B G A T E S , * ,† J O H N K A R L B Ö H L K E , ‡ AND W. MIKE EDMUNDS† Oxford University Centre for Water Research, Dyson Perrins Building, South Parks Road, Oxford OX1-3QY, U.K., and U.S. Geological Survey, 431 National Center, Reston, Virginia 20192

Received October 1, 2007. Revised manuscript received February 19, 2008. Accepted February 25, 2008.

Aerobic conditions in desert aquifers commonly allow high nitrate (NO3-) concentrations in recharge to persist for long periods of time, an important consideration for N-cycling and water quality. In this study, stable isotopes of NO3- (δ15NNO3 and δ18ONO3) were used to trace NO3- cycling processes which affect concentrations in groundwater and unsaturated zone moisture in the arid Badain Jaran Desert in northwestern China. Most groundwater NO3- appears to be depleted relative to Cl- in rainfall concentrated by evapotranspiration, indicating net N losses. Unsaturated zone NO3- is generally higher than groundwater NO3- in terms of both concentration (up to 15 476 µM, corresponding to 3.6 mg NO3--N per kg sediment) and ratios with Cl-. Isotopic data indicate that the NO3derives primarily from nitrification, with a minor direct contribution of atmospheric NO3- inferred for some samples, particularly in the unsaturated zone. Localized denitrification in the saturated zone is suggested by isotopic and geochemical indicators in some areas. Anthropogenic inputs appear to be minimal, and variability is attributed to environmental factors. In comparison to other arid regions, the sparseness of vegetation in the study area appears to play an important role in moderating unsaturated zone NO3- accumulation by allowing solute flushing and deterring extensive N2 fixation.

Introduction Aerobic conditions are common in arid region groundwaters owing to the limited occurrence of organic matter. The presence of dissolved oxygen deters microbial denitrification and NO3- can persist for thousands of years in the saturated zone (1–3). Because of these conditions, natural concentrations of NO3- in uncontaminated groundwater and unsaturated zone moisture in (semi)arid regions worldwide span at least 4 orders of magnitude. High concentrations of natural nitrate in groundwater and unsaturated zone moisture have been recorded widely from the Sahel region, Kalahari Desert, southwest and high plains U.S., central Australia and elsewhere (2, 4–10). Ratios of NO3-/Cl- are variable and may indicate varying influence of net N loss or N fixation in desert soils in addition to atmospheric deposition of N and Cl (11–13). Transmission of NO3- into groundwater is an * Corresponding author e-mail: [email protected]. † Oxford University Centre for Water Research. ‡ U.S. Geological Survey. 10.1021/es702478d CCC: $40.75

Published on Web 04/17/2008

 2008 American Chemical Society

important consideration in terms of global terrestrial Ncycling as well as for establishing baseline water quality conditions above which anthropogenic effects can be measured. Furthermore, human health risks associated with elevated NO3- intake may represent a limiting factor on freshwater potability in some water resource-scarce arid areas. This paper presents a case study from the region of the Badain Jaran Desert, a remote, sandy desert in arid northwestern China (Figure 1). Most of the area is uninhabited except for subsistence farmers and is approximately 100 km from any intensive agriculture; hence, it is an attractive field area for the study of natural geochemical cycling. The aims of the investigation were to infer the key processes of NO3cycling in groundwater and the unsaturated zone under largely pristine conditions using water chemistry and NO3isotopes (δ15NNO3 and δ18ONO3). The isotopic composition of NO3- (δ15NNO3 and δ18ONO3) provides insight into addition and removal of NO3- in groundwater systems because (1) some chemical and biological processes which alter nitrogen phase or concentration cause isotopic fractionation and (2) different NO3- sources often have distinguishable isotopic signatures in both N and O (e.g., refs 14–19). For example, the distinctively high δ18O and δ17O of atmospheric NO3 deposition may be preserved in some hyperarid soil NO3 but largely disappears where biologic cycling is important (20–23). Regional and local variations in soil N cycling processes resulting in gaseous loss (as NH3, N2, N2O) may be related to spatial variations in the δ15N values of plants, soils and recharging NO3- (13, 24). The Badain Jaran is a sandy desert of approximately 50 000 km2, bordered by palaeolacustrine plains of the Heihe River Basin to the west and southwest, including the Gurinai area which is relatively well vegetated with grasses and trees. Mountainous terrain borders the desert toward the southeast (Yabulai Mountains) and the south (Heishantou Mountains). To the east and north the sandy Badain Jaran transitions into the rocky Gobi Desert which extends into Mongolia. The regional climate is arid (89 mm/yr precipitation) and strongly continental with extreme temperature variance between summer (up to 40 °C) and winter (daytime temperatures below -10 °C). Aerobic groundwater of pH 7–8 and mixed hydrochemical facies is present at shallow depths in low-lying areas between sand dunes, and discharges to form lakes in many locations. Unconsolidated aeolian sands comprise the unsaturated zone and form an unconfined Quaternary aquifer that includes interbedded lacustrine deposits. The saturated thickness is ∼60 m in the southeastern Badain Jaran. Isotopic and geochemical studies indicate that the Yabulai Mountain front is the primary recharge zone for the area and that some recharge is occurring under the modern climate (25) with regional flow from the southeast to northwest. Diffuse recharge through the dunes is about 1 mm/yr based on chloride mass balance in the unsaturated zone (26). Recharge to the Gurinai grassland is thought to derive primarily from direct infiltration and river discharge, but may also include some lateral flow from beneath the dune field. A deeper aquifer (>100 m) under artesian conditions is used for irrigation and municipal water supply in the region. While the hydrogeology of this formation remains poorly characterized, no indication of interaction with shallow groundwater is apparent from hydrochemical and isotopic studies (25). VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3531

FIGURE 1. Map of isotope sample locations.

Experimental Section Groundwater samples were obtained from wells and springs in July 2006. Most of the samples are from shallow farmstead wells along the margin between the Badain Jaran dune field and the Yabulai Mountains, along with the areas of Xugue and Gurinai (Figure 1). Two additional wells are screened in the deeper strata, one each in Xugue (W37) and Gurinai (W43). Three of the desert’s numerous oasis lakes (Huhejaran, Yindertu, and Xugue) were sampled, as well as one spring. Shallow wells are typically lined with rock or brick and have saturated open intervals of 1–2 m. Most are located in lowlying areas with water tables less than 2 m below land surface. Shallow wells were purged with a submersible pump for at least 20 min before sample collection, and all deep wells had been pumping continuously for hours or more prior to sampling. Samples were filtered and stored in HDPE bottles at low temperature. Dissolved oxygen, pH, and water temperature were measured in the field. Unsaturated zone cores were collected from the southeastern section of the dune field by hand augur (26). Pore water was extracted by elutriation (50 g sediment to 30 mL deionized water), and a subset of these samples were selected for NO3- isotope analysis. Some samples apparently were affected by microbial degradation of nitrate during storage, and these have been excluded from isotopic analysis based on measured changes in concentration. Major ion concentrations in groundwater and unsaturated-zone moisture extracts were measured using ion chromatography (Dionex DX-500), some of which were previously reported (25). Isotopic analyses of N and O in NO3- were conducted at the U.S. Geological Survey, Reston, VA, by bacterial reduction of NO3- to N2O using Pseudomonas aureofaciens, purging and trapping the N2O, then releasing the N2O through a gas chromatograph into a ThermoFinnigan Delta Plus isotope-ratio mass spectrometer (27–29). The data were calibrated by analyzing NO3- isotopic reference materials with the samples and normalizing to the following reported values (30, 31): for USGS32, δ15N ) +180‰ and δ18O ) +26.7‰; for USGS34, δ15N ) -1.8‰ and δ18O ) –27.9‰; for IAEA-N3, δ15N ) +4.7 and δ18O ) +25.6‰; for USGS35, δ18O ) +57.5‰. The average standard deviation of 3532

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008

normalized data from two to five aliquots analyzed in different batches was approximately ( 0.3‰ for δ15N and δ18O.

Results and Discussion Groundwater nitrate concentrations in isotope samples range from nondetect (