JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in J Environ Qual 35:2273-2282 (2006)
DOI: 10.2134/jeq2006.0075
© 2006 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Agricola
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Related Collections
Right arrow Soil Hydrology
Right arrow Biogeochemical Processes
Right arrow Bioremediation and Biodegradation
Right arrow Organic Compounds
Right arrow Vadose Zone Processes and Chemical Transport

TECHNICAL REPORTS

Vadose Zone Processes and Chemical Transport

Occurrence and Rates of Terminal Electron-Accepting Processes and Recharge Processes in Petroleum Hydrocarbon-Contaminated Subsurface

Jani M. Salminena,*, Pekka J. Hänninenb, Jussi Leveinenb, Petri T. J. Lintinenb and Kirsten S. Jørgensena

a Finnish Environment Institute, P.O. Box 140, FI-00251 Helsinki, Finland
b Geological Survey of Finland, P.O. Box 96, FI-02151 ESPOO, Finland

* Corresponding author (jani.salminen{at}ymparisto.fi)

Received for publication February 20, 2006.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The occurrence and rates of terminal electron acceptor processes, and recharge processes in the unsaturated zone of a boreal site contaminated with petroleum hydrocarbons in the range C10 to C40 were examined. Soil microcosms were used to determine the rates of denitrification, iron (Fe) reduction, sulfate (SO4) reduction, and methanogenesis in two vertical soil profiles contaminated with oil, and in a noncontaminated reference sample. Furthermore, the abundances of the 16S rRNA genes belonging to Geobacteracaea in the samples were determined by real-time quantitative polymerase chain reaction (PCR). Analyses of ground water chemistry and soil gas composition were also performed together with continuous in situ monitoring of soil water and ground water chemistry. Several lines of evidence were obtained to demonstrate that both Fe reduction and methanogenesis played significant roles in the vertical profiles: Fe reduction rates up to 3.7 nmol h–1 g–1 were recorded and they correlated with the abundances of the Geobacteracaea 16S rRNA genes (range: 2.3 x 105 to 4.9 x 107 copies g–1). In the ground water, ferrous iron (Fe2+) concentration up to 55 mg L–1 was measured. Methane production rates up to 2.5 nmol h–1 g–1 were obtained together with methane content up to 15% (vol/vol) in the soil gas. The continuous monitoring of soil water and ground water chemistry, microcosm experiments, and soil gas monitoring together demonstrated that the high microbial activity in the unsaturated zone resulted in rapid removal of oxygen from the infiltrating recharge thus leaving the anaerobic microbial processes dominant below 1.5 m depth both in the unsaturated and the saturated zones of the subsurface.

Abbreviations: bgs, below ground surface • BTEX, benzene, toluene, ethylbenzene, and xylenes • DSMZ, Deutsche Sammlung von Mikroorganismen und Zellkulturen • FID, flame ionization detector • GC, gas chromatograph • HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid • PCR, polymerase chain reaction • rRNA, ribosomal ribonucleic acid


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE use of monitored natural attenuation requires evidence on the occurrence of processes that result in decrease of the mass and concentration of the contaminants in the subsurface (ASTM, 1998; USEPA, 1999). At petroleum hydrocarbon-contaminated sites undergoing natural attenuation, data on geochemical indicators are often provided to demonstrate the occurrence of microbiological processes (Borden et al., 1995). Such evidence is indirect and therefore the underlying microbiological, geochemical, hydrological, and hydrogeological processes need to be well understood to avoid misleading conclusions on the behavior and long-term evolution of a petroleum hydrocarbon plume. These processes have been extensively studied in the saturated zone of BTEX-contaminated aquifers (Chakraborty and Coates, 2004; Chapelle, 1992; Chapelle et al., 2002; McGuire et al., 2002; Vroblesky and Chapelle, 1994; Vroblesky et al., 1996). According to the current understanding, a succession of microbial terminal electron acceptor processes typically evolves in the saturated zone of a petroleum hydrocarbon-contaminated aquifer (Bouwer, 1992). This succession is strongly dependent on the availability of electron acceptors such as oxygen, nitrate (NO3), bioreducible ferric Fe oxides, and SO4 (Bekins et al., 2001; Chapelle et al., 2002; Vroblesky and Chapelle, 1994), which in turn is strongly influenced by recharge processes induced by precipitation (Vroblesky and Chapelle, 1994). Previous studies have indicated that Fe reduction, SO4 reduction, and methanogenesis are the most significant terminal electron-acceptor processes in the saturated zone of petroleum hydrocarbon-contaminated aquifers (Baedecker et al., 1993; Chapelle, 1992; Kleikemper et al., 2002, 2005). Furthermore, Anderson et al. (1998) reported that microorganisms belonging to the family Geobacteracaea were the most important Fe reducers in the Fe-reducing saturated zone of a petroleum hydrocarbon contaminated aquifer. Adding to that, Snoeyenbos-West et al. (2000) showed that Geobacter species rather than other known Fe reducers, namely Geothrix and Shewanella, became enriched in sandy aquifer sediments after a stimulation of Fe reduction with addition of electron donors. Consequently, the abundance of Geobacteracaea could be a useful indicator of Fe reduction in petroleum hydrocarbon-contaminated soils.

Unlike in the saturated zone of petroleum hydrocarbon-contaminated aquifers, the role of biogeochemical processes occurring in the unsaturated zones of petroleum hydrocarbon-contaminated aquifers has remained rather unstudied. The works of Kaufmann et al. (2004), Ostendorf and Kampbell (1991), and Pasteris et al. (2002) for example shed light to the aerobic respiration processes in this environment but the anaerobic processes remained to be studied in detail. Previous studies (Revesz et al., 1995; Salminen et al., 2004) have, however, shown that methanogenic conditions may prevail even in the unsaturated zone of aquifers contaminated with petroleum hydrocarbons. Furthermore, Bekins et al. (1999) used a most probable number technique for two aquifer sediment samples obtained from the unsaturated zone of a crude oil-contaminated aquifer to show that higher numbers of anaerobic microorganisms were found in the unsaturated zone than in the underlying saturated zone. They also speculated that the processes occurring in the oil-contaminated unsaturated zone can have a substantial influence on the composition of the recharge entering the aquifer. The above studies therefore highlight the need for detailed studies on the occurrence and rates of anaerobic terminal electron acceptor processes in the unsaturated zone of an aquifer contaminated with petroleum hydrocarbons in the range C10 to C40. They also show that there is a need to understand the interaction between microbiological and recharge processes and their joint implications on the ground water chemistry at sites where high concentrations of petroleum hydrocarbons are found in the unsaturated zone.

In the present work, we studied the occurrence and rates of denitrification, Fe reduction, SO4 reduction, and methanogenesis in the unsaturated and saturated zones of a boreal site contaminated with petroleum hydrocarbons in the range C10 to C40. Two vertical soil profiles extending from oxic top soil down to anoxic, methanogenic subsurface were examined. The rates of the terminal electron-acceptor processes were determined in microcosms at 8°C with and without an electron acceptor (NO3, Fe oxyhydroxide, or SO4) amendment. Iron reducers were further quantified using real-time quantitative PCR with Geobacteracaea-specific primers. We also explored the effect of recharge events on ground water redox potential. This was accomplished by continuous monitoring of soil water and ground water by using in situ sensors in the ground water and at various depths in the soil. Finally, we characterized the ground water geochemistry and soil gas composition at the site.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Site Description
The study site (Trollberget) is an abandoned dumpsite in southern Finland (59°53' N, 23°3' E). The soil and ground water at the site have been contaminated with light-weight fuel and lubrication oil (Salminen et al., 2004). The site was in operation from the 1960s until 1984. Since then, it is undergoing natural attenuation by biodegradation and recently, a high potential for anaerobic petroleum hydrocarbon degradation was recorded in soil samples obtained from the site (Salminen et al., 2004). The Trollberget site is located on the proximal slope of the Salpausselkä I end-moraine, the most extensive glaciofluvial formation in Finland. The stratigraphy of Trollberget site reflects the diverse deposition and erosion processes typical of ice-marginal formations. At Trollberget site, the maximum thickness of the Quaternary sediment cover is approximately 4 to 8 m and it consists of sediment units gently dipping to the northwest. The lower part of the strata comprises a coarse-grained matrix-supported diamicton. A diverse littoral unit several meters in thickness overlies the diamicton. The littoral deposit consists of coarse, medium, and fine sand and is typically horizontally laminated. In addition to natural sediments, dumped material comprising mostly of coarse sand overlies the primary littoral deposit. This material evidently originates from local deposits.

Soil Sampling and Sample Preparation
In this work, soil samples from three sampling locations were used and altogether 17 soil samples were studied. Two of the sampling points, namely G17 and G18, are located in the core of the plume (Fig. 1). Sampling point G4 is located in the clean area (Fig. 1). At points G17 and G18, vertical soil profile samples (n = 6, and n = 10, respectively) were taken with an engine-driven auger at a 0.5-m interval from 0.3 m below the ground surface (hereinafter bgs) down to 3.3 and 5.3 m bgs, respectively, in May 2001. Each sample thus comprised a 0.5-m thick section of the vertical soil profile. The clean reference sample (n = 1) was taken at 1 m below the ground surface (hereinafter bgs) at point G4. Immediately after withdrawing the samples from the ground, the soil from each 0.5-m section was mixed separately, and sieved through an 8-mm sieve and distributed to serum bottles to determine the rates of denitrification, Fe reduction, SO4 reduction, and methane production. All microcosms were brought to the laboratory at 4 to 10°C. The samples for DNA extraction (2 x 0.7 g of soil) were frozen at –70°C until further preparation. The rate of aerobic respiration, mineral oil concentration, and organic matter content of these 17 samples studied here are presented in Tuomi et al. (2004).


Figure 1
View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1. Location of the sampling points G4, G17, and G18, and the ground water monitoring wells MW2, MW3, MW8, MW9, MW11, DEMOGW4, DEMOGW6, and the piezometer DEMOGM20 at the Trollberget site. The core of the plume is indicated with dark gray color and the fringes of the plume with a light gray color. The arrows indicate the distances and directions to points G4, DEMOGW4, and DEMOGW6. Direction to the north (N) is indicated with a thick arrow.

 
Denitrification Rate in Microcosms
The rate of denitrification in the two vertical soil profile samples and in the clean sample was determined using the acetylene-inhibition method (Jørgensen et al., 2000). In the field, 10 g of soil was put into a 120-mL glass serum bottle that was then closed tightly with a butyl rubber septum and flushed with oxygen-free N2 for 4 min. In the laboratory, 10 mL of acetylene was added into the headspace of each serum bottle and the over pressure was released. Altogether, four microcosms were prepared of each sample. Two parallel microcosms were amended with 0.5 mL of anoxic 10 mM KNO3 and two parallel microcosms were left unamended. The anoxicity of the samples was confirmed by measuring the oxygen concentration by a gas chromatograph as previously described (Salminen et al., 2004). The samples were incubated at 8°C in the dark and the accumulation of N2O was measured daily using a gas chromatograph (Hewlett-Packard 5890 Series II, USA) equipped with a Porapak Q column and an EC-detector until the N2O production reached a plateau. The rate of denitrification was calculated by linear regression. In the calculation the period from Day 0 until the day that the N2O production reached a plateau was included.

Iron Reduction Rate in Microcosms
The Fe reduction rates were determined in the two vertical soil profile samples by measuring Fe2+ production in soil slurry microcosms using a modified ferrozine method (Stookey, 1970). In the field, 30 g of soil was placed into a 120-mL glass serum bottle and the headspace was flushed with oxygen-free N2 for 4 min. Altogether, four microcosms were prepared of each sample. Two parallel microcosms contained 0.5 mL of amorphous ferric oxyhydroxides prepared according to Lovley and Phillips (1986) while the other two parallel microcosms received no ferric oxyhydroxide amendment. In the laboratory, 60 mL of sterile, anoxic 0.9% sodium chloride (Cl–) was added into the serum bottles in an anaerobic glove box (Anaerobe Systems, Santa Clara, CA) under N2 85%, CO2 10%, and H2 5%. The samples were subsequently incubated without shaking at 8°C in the dark and analyzed for Fe2+ accumulation for 42 d at 3 to 11 d intervals: a 50- to 800-µL aliquot of the liquid phase in the microcosms was withdrawn through a 0.2-µm filter (Whatman Puradisc 25 AS, Arbor Technologies, Ann Arbor, MI) with a syringe and needle. Subsequently, 20 to 600 µL of the filtered aliquot was introduced into an pre-weighted volume of approximately 4 mL of ferrozine (0.5 g L–1 in 50 mM HEPES, pH 7) and analyzed for Fe2+ with a spectrophotometer (U-2000, Hitachi, Tokyo, Japan) at 562 nm (Stookey, 1970). The rate of Fe reduction was calculated by linear regression between Days 0 and 42.

Sulfate Reduction Rate in Microcosms
Sulfate reduction rates were determined in the two vertical soil profiles by measuring the disappearance of SO4 in soil slurry microcosms. Thirty grams of soil was placed into a 120-mL glass serum bottle and the headspace was flushed with oxygen-free N2 for 4 min. Altogether, four microcosms were prepared of each sample. In the laboratory, two parallel microcosms were amended with 60 mL of 0.9% NaCl. The other two parallel microcosms were amended with 60 mL of anoxic solution containing 0.5 mM Na2SO4 in 0.9% NaCl. Both amendments were done in an anaerobic glovebox (Anaerobe Systems, Santa Clara, CA) under N2 85%, CO2 10%, H2 5%. The samples were subsequently incubated without shaking at 8°C in the dark and analyzed for the disappearance of SO4 for 42 d at 4 to 9 d intervals: a 2-mL aliquot of the liquid phase in the microcosms was withdrawn through a 0.2-µm filter with a syringe and needle and analyzed for SO4 by an ion chromatograph (DX-100, Dionex, Sunnyville, CA) equipped with an AS-4 analytical column. The rate of SO4 reduction was calculated by linear regression between Days 0 and 42.

Methane Production Rate in Microcosms
Methane and CO2 production rates in anaerobic microcosms were determined as described in Salminen et al. (2004). In short, a 30-g aliquot of the soil sample was put into a 120-mL glass serum bottle and the headspace was flushed with oxygen-free N2 for 4 min. The microcosms were subsequently incubated at 8°C in the dark for 350 d and analyzed for CH4 and CO2 accumulation using a GC-FID (Hewlett-Packard 5890 Series II) and EasyQuant Carbon analyzer (Lammi, Finland), respectively. The rates of initial CH4 production were calculated by linear regression between Days 0 and 62. The rate of potential CH4 production was calculated by linear regression between Days 145 and 350.

DNA Extraction and Polymerase Chain Reaction
DNA from soil samples was directly extracted using a FastDNA SPIN Kit for soil (BIO 191, Q-biogene, Carlbad, CA) and a FastPrep Instrument (Savant Instruments, Holbrook, NY). From each of the 17 soil samples, duplicate DNA extractions were done, and the extracts were purified and quantified as previously described (Tuomi et al., 2004).

The control strain used for the quantification of the Geobacteracaea 16S rRNA genes in this study was Geobacter metallireducens supplied by DSMZ (no. 7210) and cultured according to the instructions provided by DSMZ. From the cultured cells, DNA was extracted and purified using a MasterPure Complete DNA Purification Kit (Epicentre Technologies, Madison, WI) and a Wizard DNA Clean-Up System (Promega, Madison, WI), respectively. From the DNA extract, 16S rDNA was subsequently amplified using primers Geobacteracaea-494f (Holmes et al., 2002) and Geo840R (Cummings et al., 2003) and a gradient thermal cycler (MJ Research, PTC100, MJ Research, Watertown, MA) with 30 cycles of 94°C, 60 s (initial cycle); 94°C, 45 s; 55°C, 45 s; 72°C, 60 s; with a final extension at 72°C for 6 min. The total volume of each PCR reaction was 50 µL and contained 0.2 µM of both primers, 200 µM of each dNTP, 0.2 U of Taq polymerase (MBI Fermentans, Vilnius, Lithuania, 1 U µL–1), 20 mM of NH4(SO4)2, 2.5 mM of MgCl2, and BSA, and 1 µL of template DNA. The PCR product was subsequently purified using Wizard SV Gel and PCR Clean-up System (Promega, Madison, WI) and the concentration of the DNA fragment was determined by a spectrophotometer (U-2000, Hitachi Ltd., Tokyo, Japan) at 260 nm. The number of the target copies was calculated based on the concentration and molecular weight of the DNA fragment. From the purified PCR product, a dilution series was prepared which contained 1.7 to 1.7 x 109 target copies per µL. This dilution series was used to make a standard curve for the quantification of Geobacteracaea 16S rRNA genes by real-time quantitative PCR. The standards together with ninefold dilutions of the DNA extracts were run in triplicates on a thermal cycler (ABI Prism 7000, Applied Biosystems, Foster City, CA) to quantify the number of the 16S rRNA genes belonging to Geobacteracaea in the subsurface samples. The total volume of the real-time PCR reaction was 20 µL and contained 10 µL of the 2x Quantitect SYBR Green mastermix (Qiagen GmbH, Hilden, Germany), 2.5 µM of each primer (494f and 840r) and 1 µL of the template DNA. The temperature program of the PCR run was 95°C, 15 min (first cycle only), 94°C, 30 s; 55°C, 30 s; 72°C, 60 s; 80°C, 30 s; with a final extension at 72°C for 6 min. Altogether, 40 cycles were run and the PCR products were run in a 1.5% (w/v) agarose gel in 0.5% TBE (150 V, 60 min) together with a molecular mass ruler (100 bp, MBI Fermentas, Vilnius, Lithuania).

Soil Physical Properties and Hydrological Measurements
To estimate the water infiltration rate in the unsaturated zone, and the temporal changes in soil water and ground water chemistry, several hydrological and geochemical parameters were continuously monitored both in the ground water and at different depths in the soil (Table 1). The sensors in the soil and a piezometer (designated DEMOGM20) were installed on 12.5.2004 as follows: a test pit was excavated 5 m to the south of point G17. The sensors were installed into the soil profile at given depths (Table 1) and soil samples from the test pit were taken for grain size distribution analysis. Thereafter, a piezometer was installed and the excavated soil was returned to the pit. The sensors used for the continuous monitoring of the ground water were then installed at given depths in the piezometer (Table 1). The piezometer had a screened section approximately 2 m in length extending from the saturated zone to the unsaturated zone (Fig. 2). Therefore, the water in the piezometer represented a mixture of ground water and infiltrating water from the unsaturated zone that had possibly seeped into the screened section. At the time of the installation, the lowest sensor in the piezometer (the pressure transducer) was 1.2 m below the ground water table. The sensors in the soil and in the piezometer were used for ground water monitoring between 12.5.2004 and 31.1.2005, and each parameter was measured automatically at 3 h intervals. The sensors in the piezometer were cleaned and recalibrated once a month.


View this table:
[in this window]
[in a new window]
  		Table 1. Parameters, equipment, and monitoring depths used in the continuous monitoring of the saturated and the unsaturated zones in the vicinity of point G17 at the Trollberget site. For the data collection, a CR10X (Campbell Scientific Ltd., Loughborough, UK) data logger was used.

 

Figure 2
View larger version (33K):
[in this window]
[in a new window]
  		Fig. 2. Schematic drawing of the soil sections at point G18 and in the vicinity of the point G17 at the Trollberget site. At point G18, the soil profile comprised following units: Unit A; fine sand (filling soil), Unit B; coarse sand (filling soil), Unit C; garbage-containing sand (filling soil), Unit D; layer with very low water conductivity, high oil and organic matter content (filling soil), Unit E; fine littoral sand, Unit F; coarse sand. A layer of perched ground water (P), approximately 0.2 m in thickness, above the unit D is indicated. The approximate level of the ground water is indicated with an arrow. In the vicinity of point G17, the soil profile comprised the following units: Unit A consisted of coarse sand (filling soil), unit B of medium to fine-grained littoral sand, and unit C of shell-rich, littoral, silty sand. The position of the soil monitoring sensors (black rectangles), piezometer (DEMOGM20), and the ground water monitoring sensors in the piezometer (black circles) installed in the vicinity of points G17 at the Trollberget site, and the highest and the lowest level of ground water (arrows) from 12.5.2004 to 31.1.2005 in the piezometer, are indicated.

 
The water infiltration rate through the unsaturated zone at G17 was also estimated based on the grain size distribution and bulk density of the aquifer material. For this purpose, grain size distribution in subsurface samples (n = 12) down to 5 m at points G17 and G18 was first determined by sieving (0.063–64 mm). Subsequently, a sedigraph (SediGraph 5100, Micromeritics, Norcross, GA) was used to further analyze the fraction that was <0.063 mm down to 0.2 µm. Additionally, bulk densities of five samples with a known volume were measured at the site. The water infiltration rate in the unsaturated zone was then estimated based on these data using the SOILPAR 2.00 software (Acutis and Donatelli, 2003; Rawls and Brakensiek, 1989).

Soil Gas Composition
The concentrations of O2, CO2, and CH4 were determined from soil gas monitoring well previously (Salminen et al., 2004) installed in different depths at points G17 and G18. The determination of O2 and CO2 was performed by using a MultiWarn II analyzer (Dräger, Lübeck, Germany). The samples for the determination of CH4 were collected as previously described (Salminen et al., 2004) and analyzed for CH4 by GC-FID. Soil gas composition was determined five times during 2001 to 2004.

Ground Water Chemistry
Ground water samples were obtained in November 2000 from the ground water monitoring wells MW2, MW3, MW8, MW9, and MW11 (Fig. 1). These monitoring wells are located in the core and in the downgradient fringe of the plume. In October 2004, more comprehensive characterization of the ground water chemistry was performed. Samples were taken also from the monitoring wells DEMOGW4, DEMOGW6, and in the piezometer (DEMOGM20) (Fig. 1). In 2000, samples were obtained by an inertial pump. In 2004, samples were obtained by a point sample bailer. The ground water samples were analyzed for NO3, SO4, and Cl– by using an ion chromatograph (DX-120, Dionex, Sunnyville, CA). For the determination of Fe2+ and Ca2+, the sample was filtrated (0.2 µm filter, Whatman Puradisc 25 AS, Arbor Technologies, Ann Arbor, MI). The Fe2+ content was immediately determined using the ferrozine method (described above) and a portable spectrophotometer (DR/2000, Hach Co., Loveland, CO) and the Ca2+ content was determined by using inductively coupled plasma–atomic emission spectrometry (ICP–AES) (IRIS Advanced Duo High Resolution, Thermo Jarrel-Ash Co, USA). In addition, to determine CH4 and HCO3– concentrations in the ground water at G17, 20 mL of ground water was injected into a 120-mL serum bottle containing 1 mL of 4 M H2SO4. In the laboratory, the samples were vigorously shaken and the concentrations of CO2 and CH4 in the head-space were determined using an EasyQuant Carbon analyzer and GC-FID, respectively, as described above. The concentration of petroleum hydrocarbons in the range C10 to C40 was determined according to a standard protocol (ISO 2000) using acetone–hexane extraction and subsequent gas chromatographic (GC-FID) analysis.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The Occurrence and Rates of the Microbial Terminal Electron Acceptor Processes
In the present study, 16 soil samples contaminated with C10 to C40 petroleum hydrocarbons and one clean reference sample were examined. The oil-contamined soils represented two vertical soil profiles from points G18 and G17 (Fig. 1). The vertical soil profiles extended from aerobic topsoils down to anoxic, methanogenic soil below depth of 2 m bgs (Fig. 3A and 3G). Both of the points G18 and G17 located in the core of the plume (Fig. 1) and had a petroleum hydrocarbon content up to 6 800 mg kg–1 dwt (Fig. 3B and 3H). However, the samples from the point G18 had significantly higher total organic matter content compared to that of samples from the point G17 (Fig. 3B and 3H).


Figure 3
View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Concentrations of O2 ({circ}), CO2 ({triangledown}), and CH4 (black circle) in the soil gas (panel A and G), concentrations of petroleum hydrocarbons in the range C10 to C40 (black circle) and organic matter ({triangledown}) (B and H; Tuomi et al., 2004), the rate of Fe reduction (C and I), the abundance of 16S rRNA genes belonging to Geobacteracaea ({nabla}; C and I), the rate of methane production (D and J) the rate of sulfate (SO4) reduction (E and K), and the rate of denitrification (F and L), as a function of depth in the vertical soil profiles obtained from point G18 (panels A-F) and G17 (panels G-L) at the Trollberget site. The rates of Fe reduction, SO4 reduction, and denitrification in unamended and amended soil microcosms are presented with white circles ({circ}) and black circles, respectively. The initial rates (Days 0–62) and potential rates (Days 145–350) of methane production in soil microcosms are presented with white circles and black circles, respectively. The gray color indicates perched ground water and ground water (A and G).

 
Iron reduction was recorded in the samples throughout the two vertical soil profiles (Fig. 3C and 3I). Similarly, 16S rRNA genes belonging to Geobacteracaea were detected throughout the soil profile (Fig. 3C and 3I), and a significant correlation (Pearson; 0.640, P < 0.008) was found between the Fe reduction rate and the abundance of the 16S rRNA genes belonging to Geobacteracaea. In the microcosms, a lag phase (5–7 d) in the production of Fe2+ was observed only in samples obtained from a depth of 0.3 to 0.8 m bgs at the two points G17 and G18 (data not shown). These data together demonstrate that Fe reduction was an important process throughout the vertical soil profiles extending from the unsaturated zone down to the saturated zone of the core of the plume. It is noteworthy that significant Fe reduction and up to 5 x 105 16S rRNA genes belonging to Geobacteracaea were detected also in the oxic layers: at 1 m depth at G18, for instance, the O2 concentration in the soil gas was approximately 10% (Fig. 3A) and the abundance of Geobacteracaea species as quantified by real time PCR was 10 times higher than that at 4 m depth bgs in the same profile. Recently, Lin et al. (2004) showed in a laboratory study that Geobacter sulfurreducens grows at O2 concentrations up to 10% suggesting that Geobacter species may well survive in subsurface with temporarily elevated O2 concentrations. Our work demonstrates that in addition to anoxic environments, Geobacteracaea do thrive in the oxic unsaturated zone of a petroleum hydrocarbon contaminated sandy aquifer. In the present work, the abundance of Geobacteracaea ranged from 2.3 x 105 to 4.9 x 107 copies g–1 soil and from 2.7 x 105 to 2.7 x 106 copies g–1 soil at G18 and G17, respectively (Fig. 3C and 3I). The abundance of the Geobacteracaea 16S rRNA genes in the samples that showed high Fe reduction rate in microcosms was in the same range as those recorded from other contaminated environments by Cummings et al. (2003) and Snoeyenbos-West et al. (2000). The microcosm experiments also indicate that Fe reduction at the Trollberget site was not limited by the lack of bioreducible Fe hydroxides; the unamended microcosms showed similar rates of Fe2+ production to those in the amended microcosms (Fig. 3C and 3I). The opposite has been reported at some petroleum hydrocarbon contaminated sites undergoing natural attenuation (Gieg et al., 1999), or in landfill leachate polluted aquifers (Albrechtsen and Christensen, 1994).

Methane production was recorded during the 62-d incubation in samples taken below 1 m bgs and between 0.8 and 1.8 m bgs at points G18 and G17, respectively. (Fig. 3D and 3J). During a prolonged incubation (350 d) under anaerobic conditions at 8°C, the rate of methane production significantly increased in most of the samples (Fig. 3D and 3J) and all the samples turned methanogenic except the topsoil sample taken at point G18. In the samples from point G17, methane production rate increased by a factor of 3 to 58 during the incubation. In spite of that, these rates of methane production were significantly lower than those measured for samples from point G18 (Fig. 3D and 3J). In many of the samples, the rate of CO2 production decreased during the incubation with concomitant increase in methane production (data not shown). This suggests that CO2 was being consumed by hydrogenotrophic, CO2–reducing microorganisms. The high soil gas methane concentrations measured in the unsaturated zones of the points G17 and G18 (Fig. 3A and 3G) support the microcosm data and demonstrate that methane production was a significant microbial process in the vertical soil profiles studied.

Sulfate reduction was recorded during the 40-d incubation in all the samples except the two samples taken 0.3 to 0.8 m bgs and 0.8 to 1.3 m bgs at point G18 (Fig. 3E and 3K). Sulfate reduction rates were higher in the amended microcosms (Fig. 3E and 3K) than in the unamended ones. This can be attributed to the quite low (3.8–41 mg SO42– kg–1 soil dwt) concentrations of SO4 measured initially in the unamended microcosms. This indicates that SO4 reduction was limited throughout the vertical soil profiles due to low availability of SO4.

Denitrification in the vertical soil profiles was shown to be seriously limited by the availability of NO3. In the unamended microcosms, no N2O was produced (Fig. 3F and 3L). However, addition of small amount (5 µmol) of NO3 resulted in the generation of N2O in all the samples (Fig. 3F and 3L). This generation of N2O was recorded in most of the samples without a lag phase. A lag phase (3 d) was observed only in the three samples obtained from the saturated zone at G17. This demonstrates that potential for denitrification existed in the soil samples studied.

In the vertical soil profiles examined here, the rates of the microbial terminal electron acceptor processes were primarily dependent on the amount of organic matter available: statistical analysis revealed significant correlation (Pearson) between organic matter content (Tuomi et al., 2004) and Fe reduction (unamended: 0.806, P < 0.0005; amended: 0.843, P < 0.0005), SO4 reduction (unamended: 0.743, P < 0.001; amended: 0.700, P < 0.003), and methanogenesis (initial: 0.628, P < 0.009; potential: 0.645, P < 0.007). The rates of terminal electron acceptor processes were significantly higher in the samples from the unsaturated zone than those measured in the samples from the saturated zone. Furthermore, in the three samples representing the saturated zone in this work (point G17; 1.8–3.3 m bgs), microbial activity decreased with increasing depth. In the vertical soil profile at point G17, Fe reduction, SO4 reduction, and methane production peaked in the unsaturated zone and the rates obtained in the deepest sample (2.8–3.3 m bgs) were lower or in the same order of magnitude that the rates obtained in the sample from the oxic zone (0.3–0.8 m bgs) in the same profile (Fig. 3I–L).

The rates of Fe reduction, SO4 reduction, and methanogenesis were significantly higher in the contaminated samples (Fig. 3) than in the reference sample obtained from point G4 at the clean area (Table 2). Similarly, the abundance of 16S rRNA genes belonging to Geobacteracaea were orders of magnitude lower in the clean sample (Table 2) than in the contaminated samples (Fig. 3). Yet, these sequences were found in the clean sample despite the prevailing highly oxic conditions in the clean soil. Moreover, no lag phase for Fe reduction was observed in this sample and the availability of bioreducible Fe seemed not to be a limiting factor for Fe reduction (Table 2). By contrast, both denitrification and SO4 reduction were strongly limited in the clean sample due to low concentrations of NO3 and SO4, respectively.


View this table:
[in this window]
[in a new window]
  		Table 2. Petroleum hydrocarbon concentration, organic matter content, rates of denitrification, Fe reduction, sulfate (SO4) reduction, and methanogenesis in microcosms, the abundance of 16S rRNA genes belonging to Geobacteracaea, and the concentrations of O2, CO2, and CH4 in soil gas at 1 m belowground surface at point G4 at the Trollberget site.

 
The highest rates of Fe reduction, SO4 reduction, and methanogenesis obtained in this work were significantly (by a factor of 5, 38, and 81, respectively) higher than those obtained by Ludvigsen et al. (1998) for landfill leachate-contaminated aquifer sediments at 10°C. By contrast, they obtained maximum rates of denitrification significantly (by a factor of 132) higher than those obtained here and the highest rates were recorded in locations with high NO3 concentrations. In this study, we cannot exclude the possibility that the presence of H2 (5%) and CO2 (10%) in the head-space of the Fe reduction and SO4 reduction microcosms may have resulted in an overestimation of the rates of these two processes by enhancing the activity of hydrogen-oxidizing Fe and SO4 reducers. The rates of these processes were, however, of the same order of magnitude when the head-space of microcosms containing soil taken from this site, was flushed with N2 (J. Salminen, unpublished data, 2001).

The Effect of Recharge Processes on Ground Water Redox Potential
The continuous in situ monitoring revealed that rapid recharge processes through the oil-contaminated unsaturated zone do occur at the Trollberget site. This was most clearly observed after the heavy rainfall events between 30 June to 2 July 2004 (Fig. 4A), which resulted in a sharp increase in the soil water content at 0.3 m bgs (Fig. 4B). Approximately 10 h later, an increase of soil water content was observed at 1.1 m bgs (Fig. 4C) and at the same time, the electric conductivity started to increase at that depth (Fig. 4D). Ten to 15 h later, the electric conductivity started to increase at 1.9 bgs (Fig. 4E). In 6 h from that, a sharp increase in the electrical conductivity and in the temperature of the ground water (Fig. 4G and 4H, respectively) was observed together with a slight but sudden decrease of the ground water pH (Fig. 4J) in the piezometer (DEMOGM20). This decrease in pH could be due to the introduction of organic acids, produced by fermentative bacteria, into the ground water. As expected, the hydraulic head in the piezometer had started to rise earlier (Fig. 4I). However, no changes were observed in electric conductivity in soil at 2.3 m bgs (Fig. 4F). Furthermore, the changes observed in the ground water elevation, temperature, electrical conductivity, and pH during these recharge events were not reflected in the ground water redox potential, which remained at –416 mV...–420 mV during this recharge event (Fig. 4K) and below –400 mV throughout the monitoring period (data not shown). The persistence of low redox potential in ground water points to oxygen being consumed in the unsaturated zone before its introduction into the ground water. This agrees with the previously (Tuomi et al., 2004) obtained rates for aerobic respiration in the vertical soil profile samples examined in this study. The oxygen consumption rates derived from the rates of aerobic respiration measured in microcosms exceeded the estimated water infiltration rate (2 x 10–5 m s–1) in the unsaturated zone at G17 by a factor of 3 to 15. It should be noted that oxidation of CH4 is not taken into account in these calculations. What we also want to highlight is that aerobic respiration initiated without a lag phase in the subsurface samples obtained even from the anoxic layers at G17 and G18 (P. Tuomi, unpublished data, 2001). This indicates that oxygen consumption by microorganisms capable of aerobic respiration is initiated as soon as oxygen appears in the anoxic subsurface.


Figure 4
View larger version (13K):
[in this window]
[in a new window]
  		Fig. 4. The daily rainfall at Tvärminne weather station (in Hanko, Finland) (A), and the water content at 0.3 m below ground surface (bgs) (B) and 1.1 m bgs (C), electric conductivity at 1.1 m bgs (D), at 1.9 bgs (E) and at 2.3 m bgs (F) in the vicinity of point G17 at the Trollberget site as measured continuously from 27.6. to 12.7.2004. Electrical conductivity (G), temperature (H), hydraulic head (I), pH (J) and redox-potential (K) in the ground water as measured continuously in the piezometer DEMOGM20 during 27 June to 12 July 2004 at the Trollberget site. The depths of each sensor in the soil and in the piezometer are indicated.

 
Although heavy rainfall events did not affect the ground water redox potential, the infiltration of O2 containing recharge is, however, bound to effect the geochemistry of the unsaturated zone by enhancing the (bio)chemical oxidation of, for instance, ammonium (NH4), Fe2+, and S. This results in the formation of NO3, ferric Fe oxyhydroxides and SO4, respectively, which can be subsequently used as electron acceptors in the oxidation of organic compounds. Furthermore, the aerobic oxidation of the high CH4 content found in the soil gas by methanotrophic microorganisms presumably results in equal consumption of oxygen.

The continuous monitoring of soil water content, electrical conductivity, and temperature also nicely illustrated the effect of subsurface stratification for the transportation processes in subsurface. The changes in the electrical conductivity during the recharge event during could be followed through the subsurface profile down to the ground water (Fig. 4). This allowed us to calculate that it takes approximately 1 d for the fresh recharge to infiltrate to the ground water through the approximately 1.9 m thick unsaturated zone at point G17. The effect of soil stratification on the variably saturated flow was, however, revealed by the lack of a swift response in the temperature and electric conductivity sensors at 2.3 m bgs during recharge events (Fig. 4F and data not shown). These sensors were right below the littoral silty sand layer (Fig. 1), which presumably had a substantially lower hydraulic conductivity that the overlying deposits. The screened section of the piezometer, however, penetrated this layer and therefore, the effect of the recharge could be observed in it.

In this work, relatively good agreement between the water permeability values obtained with different methods was achieved. The water infiltration rate of the unsaturated zone of the vertical soil profile at G17 was estimated to range from 1.8 x 10–5 to 2.7 x 10–5 m s–1 based on the bulk density and the grain-size distribution data analyzed with the SOILPAR 2.00 software. In comparison, the analysis of the changes observed in the soil water content, and electrical conductivity at G17 after heavy rainfall events gave a water infiltration rate estimate of 1 x 10–5 to 3 x 10–5 m s–1. Consequently, the continuous monitoring of soil water and ground water provides a reliable tool to monitor the effects of hydrological processes at sites undergoing natural attenuation.

Ground Water Chemistry
Although the focus of this work was on the unsaturated zone, ground water geochemistry in different parts of the site was determined to estimate how the occurrence of the microbial processes has affected the chemical composition of the ground water at the site. In the core of the plume, Fe2+ was measured in high concentrations together with methane (Table 3). By contrast, very low NO3 concentrations and low SO4 concentrations were recorded in the core of the plume. These findings agree very well with the high rates of Fe reduction and methane production and with the absence of denitrification, and low rates of SO4 reduction measured in the microcosms. The ground water geochemistry therefore seems to reflect the occurrence of various terminal electron-acceptor processes in the subsurface at Trollberget site. The concentrations of NO3, SO4, Fe2+, and methane had remained relatively unchanged from 2000 to 2004 in the core of the plume (Table 3). Similarly, high concentrations of methane were found both in 2000 and 2004 in the two vertical soil profiles studied. These results together indicate that the core of the plume is stable and methanogenic since at least 2000. The geochemical data also show that very high concentrations of HCO3– occurred in the core of the plume, making HCO3–, together with Ca2+, the dominating ions in the contaminated ground water (Table 3). High HCO3– content in ground water is a common feature at petroleum hydrocarbon sites (Azadpour-Keeley et al., 2001; Baedecker et al., 1993) and can result from various anaerobic degradation processes, including aceticlastic methanogenesis (Revesz et al., 1995). Comparison between the uncontaminated ground water in the upgradient wells (DEMOGW4 and DEMOGW6), and the wells in the core and the fringe of the plume shows a gradual change in the ground water geochemistry from the clean to the most contaminated area (Table 3). This together with the above data demonstrate that ground water geochemistry does reflect the occurrence of microbial terminal electron-acceptor processes at sites contaminated with petroleum hydrocarbons and undergoing natural attenuation.


View this table:
[in this window]
[in a new window]
  		Table 3. The concentrations of petroleum hydrocarbons in the range C10 to C40 (PHC), and nitrate (NO3), Fe2+, sulfate(SO4), methane, bicarbonate, Ca, and Cl– in the ground water monitoring wells located in the core (piezometer DEMOGM20, and wells MW2, MW3, and MW11) and fringe (wells MW8 and MW9) of the petroleum hydrocarbon plume, and outside the plume (DEMOGW4 and DEMOGW6) at the Trollberget site in October 2004. The values in parentheses indicate the concentrations measured in November 2000.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
To summarize, this work demonstrates that the unsaturated zone of a boreal aquifer contaminated with petroleum hydrocarbons in the range C10 to C40 exhibited a great diversity of microbial terminal electron acceptor processes that seemed to be occurring concomitantly within the 0.5 m long vertical sections of soil studied. This shows that the identification of distinguishable terminal electron-acceptor process zones in the unsaturated zone looks to be even more challenging than it is in the saturated zone (Chapelle et al., 2002). However, several lines of evidence demonstrated that Fe reduction and methanogenesis played significant roles in the vertical soil profile studied. In the microcosms, high rates of Fe reduction and methanogenesis were recorded. Furthermore, high abundances of Geobacteracaea 16S rRNA genes were detected throughout the vertical soil profiles extending from the oxic topsoil down to the anoxic saturated zone, and these abundances correlated with the Fe reduction rates measured in microcosms. Moreover, high concentrations of methane and Fe2+ were found in the soil gas and ground water, respectively, in the core of the plume. In this work, we also demonstrated that oxygen is rapidly removed from the recharge passing through the unsaturated zone contaminated with petroleum hydrocarbons in the range C10 to C40. Consequently, even rapid infiltration of recharge through the shallow, oil-contaminated unsaturated zone did not increase the redox potential of the highly reduced ground water at the boreal site studied here. This work highlights the importance of the microbial processes in the unsaturated zone of petroleum hydrocarbon contaminated aquifers. These processes do significantly affect the chemical composition of recharge and the availability of electron acceptors in the saturated zone at sites undergoing natural attenuation.


   ACKNOWLEDGMENTS
 
We would like to thank Ritva Väisänen, Jukka Rinkinen, Taija Pietilä, Sinikka Pahkala, Jaana Heiskanen, Heikki Jyllilä, Mikael Eklund, and Osmo Äikää for their technical assistance with the field work and the laboratory analyses, and Juha Reinikainen for his input in the geological characterization of Trollberget site. We also would like to thank Dr. Christina Lyra and Dr. Anna-Liisa Kivimäki for constructive comments on the manuscript. Finally, we gratefully acknowledge the financial support from the Academy of Finland (Grant 201359) and from the EU-Life Programme (LIFE 03 ENV/FIN/000250).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Agricola
Right arrow Articles by Salminen, J. M.
Right arrow Articles by Jørgensen, K. S.
Related Collections
Right arrow Soil Hydrology
Right arrow Biogeochemical Processes
Right arrow Bioremediation and Biodegradation
Right arrow Organic Compounds
Right arrow Vadose Zone Processes and Chemical Transport

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