Environ. Sci. Technol. 2009, 43, 6548–6552
Does Iron Cycling Trigger Generation of Acidity in Groundwaters of Western Australia? S T E F A N P E I F F E R , * ,† C A R O L Y N O L D H A M , ‡ U R S U L A S A L M O N , ‡ A D A M L I L L I C R A P , ⊥,§ ¨ SEL| AND KIRSTEN KU Department of Hydrology, University of Bayreuth, Universita¨tsstrasse 30, 95440 Bayreuth, Germany, Centre for Ecohydrology, School of Environmental Systems Engineering, University of Western Australia, Crawley, Western Australia 6009, Australia, Department of Agriculture and Food, Albany, Western Australia 6330, Australia, and Limnology Research Group, Institute of Ecology, Friedrich Schiller University of Jena, Germany
Received February 4, 2009. Revised manuscript received May 22, 2009. Accepted July 8, 2009.
In large areas of Western Australia, acidic groundwaters occur with pH values distinctly lower than 3, generation of which has been attributed to the oxidation of Fe(II). Incubation experiments performed with sediments from playas receiving acid groundwater demonstrated occurrence of reductive dissolution of ferric iron minerals at rates [670 nmol (g reactive iron)-1 h-1] similar to those observed in sediments of acidic mining lakes (AML), indicating that the pH was established through an acidity-driven iron cycle in analogy to processes occurring in AML systems. The low pH values observed in acidic groundwaters and AML, however, can only be achieved if the anion corresponding to Fe(II) is that of a strong acid. In AML, sulfate is derived from pyrite oxidation. Because this process is reported not to occur in the acidic groundwater systems of Western Australia, we have derived a conceptual model according to which sulfate is generated upon reaction of weathering-derived alkalinity with gypsum to form calcite, which is abundant in these areas. The model proposes that part of the alkalinity generated during weathering is stored as calcite in the landscape, which leads to spatial separation of acidity and alkalinity.
Introduction Acidic groundwaters and hyper-saline lakes with pH values distinctly lower than 3 can be found in large areas of South and Western Australia (1-3). In the Western Australian (WA) wheat belt, acidic groundwaters (AGW) extending over an estimated 100000 km2 cause severe impacts on infrastructure and agriculture (4) as well as creating challenges for mineral exploration and extraction (5). Formation of AGW is related to the extremely long weathering history of the continent (4), which has been geologically inactive since the Miocene (6). Several mechanisms for the origin of the acidity have * Corresponding author e-mail:
[email protected]. † University of Bayreuth. ‡ School of Environmental Systems Engineering, University of Western Australia. ⊥ Centre for Ecohydrology, University of Western Australia. § Department of Agriculture and Food. | Friedrich Schiller University of Jena. 6548
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 17, 2009
been proposed such as pyrite oxidation, nitrification, and evaporation (7). The abundance of ferric minerals in discharge areas of AGW has led to the conclusion that oxidation of ferrous to ferric iron and the subsequent hydrolysis reaction, termed ferrolysis, plays a prominent role in acidification of the groundwater (1, 2, 7). Ferrolysis, however, is only able to establish such low pH waters if the anion corresponding to Fe2+ such as SO42- does not contribute to alkalinity, a fact that has not been considered in previous discussions. It has been demonstrated that ferrolysis is also responsible for the acidification of mining lakes that receive Fe(II)- and SO42--containing groundwater (8). These waters rapidly oxidize upon contact with oxygenated lake water, resulting in acidic mining lakes (AML). AML are characterized by a remarkably stable pH that reflects equilibrium between Fe3+ and the ferric oxyhydroxy sulfate mineral schwertmannite [Fe8O8(OH)xSOy · nH2O, where 8 - x ) y and 1.0 < y < 1.75] (9) buffering the pH at values between 2.6 and 3.3 (10). The stability of the acidic conditions is caused by an aciditydriven iron cycle in the sediments of the lakes (11). Although schwertmannite is microbially reduced under acidic conditions by acidophilic bacteria, the alkalinity derived from reductive dissolution is completely consumed by the reoxidation of Fe(II) (11). Thus, the pH does not increase in the lake water or in the top 5 cm of the sediment. To test the analogy between acidification processes in AML and AGW-lake systems, we have studied the biogeochemistry of iron at two different playas of the WA acidic groundwater belt. From these observations, we have derived a conceptual model that contributes to the fundamental understanding of the evolution of groundwaters in an intensively weathered environment and will support water management in such areas. Moreover, it will add to our understanding of the formation of an environment that is regarded to be an analogue to conditions on planet Mars (12).
Study Sites The area containing acidic groundwater extends across the southwest of Australia. The percentage of acidic groundwater bores increases in the broader, drier paleo-drainage channels extending to the east of southern WA (Figure 1). The groundwaters typically discharge into playas that contain sulfate of marine origin. As a consequence, sulfate salts characteristic of arid sulfate-rich environments such as alunite or gypsum (14) are abundant (15). We collected two samples from Lake Gilmore (S 32°36.539, E 121° 33.680, alt 240 m) at the same locations described in an earlier study (2) and one sample from an unnamed lake which will be called Green Lake (S 33° 03.376 S, E 121° 40.578 E, alt 230 m) on September 23 and 26, 2006, i. e., at the end of the southern Australian rainy season. Lake Gilmore contained no water at this time, and the pH in Green Lake was 2.6. Sediments were sampled at the capillary fringe at the shore after digging until the groundwater surface was reached (not deeper than 30 cm). At Lake Gilmore, the sites corresponded approximately to sites 3 and 5 in ref 2. Sediments were placed into plastic bags. Pore water was sampled into polyethylene vessels without further treatment. The samples were stored in a cooler filled with ice until the end of the sampling campaign (max. 4 d). pH, total, and ferrous iron were determined directly in the field and compared to laboratory measurements performed after the field trip. Materials and Methods. Release of Fe(II) from the sediments was studied by closed vessel incubation experi10.1021/es9001086 CCC: $40.75
2009 American Chemical Society
Published on Web 07/24/2009
FIGURE 1. Percentage of groundwater bores that are acidic in the southwest of Western Australia, based on data presented in ref 13.
TABLE 1. Fe and SO42- Concentrations of Sediment Pore Waters and Solid-Phase Contents of Dithionite Extractable (Fed) and HCl Extractable (FeHCl) Fea
Lake Gilmore (LG-1) Lake Gilmore (LG-2) Green Lake (GL) a
pH
Fe(II) [mmol L-1]
Fetot [mmol L-1]
SO42- [mmol L-1]
Fed (g kg-1)
FeHCl (g kg-1)
3.0 3.4 3.7
0.035 0.006 0.150
0.035 0.021 1.2
36 34 51
73 2.0 24
0.78 0.09 7.2
Samples taken at the shore of two playas.
ments (16) using the same methodology as described in ref 17. In brief, the samples were incubated at room temperature (about 23 °C) with artificial groundwater that was prepared according to the chemical composition measured at site LG-2 (172.9 g of NaCl, 20.92 g of MgSO4 · 7H2O, 1.546 g of KCl, 5.979 g of Na2SO4, 47.00 g of MgCl2 · 6H2O, and 0.9347 g of CaCl2; the pH was adjusted to pH 2.8 with H2SO4/NaOH). Fe(II) and total iron was determined using the phenanthroline method (18). Sulfate was measured using ion chromatography by a laboratory experienced with high salinity samples. The solid phases were analyzed with respect to its mineral composition using X-ray diffraction (XRD). In order to characterize the reactivity of iron minerals, chemical extractions were performed with HCl to account for reactive Fe(III) (c ) 1 mol L-1) (19) and sodium dithionite-citrate-bicarbonate to quantify total ferric (hydr)oxide concentration (20). Fe(III)-reducing microbial community was characterized by polymerase chain reaction (PCR) amplification and DNA extraction with primer sets specific for phylogenetically known Fe(III) reducers (21). DNA was extracted from the sediment using the MOBIO Power Soil DNA extraction kit according to manufacturer’s instructions. Aliquots of DNA were PCR-amplified using bacteria domain-specific (GM3, GM4) (22) and 16S rRNA gene primers specific for Acidiphilium (Acido594F, Acido1150R) (23), Geobacter (GM3, 825R) (23), Geothrix (Gx182F, Gx472R) (24), and Shewanella
(Shw783F, Shw1245R) (24) as previously described (21). Anaeromyxobacter-specific PCR was performed according to an earlier described method (Ab112F, Ab227R) (25). Geochemical modeling was performed using the computer code PHREEQC (26).
Results and Discussion Biogeochemical Processes at the AGW-Playa Interface. The groundwater was acidic with pH values
