OUP user menu

Microbial communities in acid mine drainage

Brett J. Baker, Jillian F. Banfield
DOI: http://dx.doi.org/10.1016/S0168-6496(03)00028-X 139-152 First published online: 1 May 2003


The dissolution of sulfide minerals such as pyrite (FeS2), arsenopyrite (FeAsS), chalcopyrite (CuFeS2), sphalerite (ZnS), and marcasite (FeS2) yields hot, sulfuric acid-rich solutions that contain high concentrations of toxic metals. In locations where access of oxidants to sulfide mineral surfaces is increased by mining, the resulting acid mine drainage (AMD) may contaminate surrounding ecosystems. Communities of autotrophic and heterotrophic archaea and bacteria catalyze iron and sulfur oxidation, thus may ultimately determine the rate of release of metals and sulfur to the environment. AMD communities contain fewer prokaryotic lineages than many other environments. However, it is notable that at least two archaeal and eight bacterial divisions have representatives able to thrive under the extreme conditions typical of AMD. AMD communities are characterized by a very limited number of distinct species, probably due to the small number of metabolically beneficial reactions available. The metabolisms that underpin these communities include organoheterotrophy and autotrophic iron and sulfur oxidation. Other metabolic activity is based on anaerobic sulfur oxidation and ferric iron reduction. Evidence for physiological synergy in iron, sulfur, and carbon flow in these communities is reviewed. The microbial and geochemical simplicity of these systems makes them ideal targets for quantitative, genomic-based analyses of microbial ecology and evolution and community function.

  • Acid mine drainage
  • Pyrite dissolution
  • Iron oxidation
  • Sulfide mineral
  • Sulfur oxidation
  • Geomicrobiology

1 Introduction

Acidic, metal-rich fluids are formed by chemical weathering of metal sulfide-rich rocks. These acid rock drainage (ARD) solutions are hot because metal sulfide oxidation reactions are highly exothermic. The predominant metal sulfide mineral in most rocks is pyrite (FeS2). Pyrite-rich deposits are often mined for metals such as Au, Ag, Cu, Zn, and Pb, which are typically present as impurities in pyrite or occur in sulfide minerals such as chalcopyrite (CuFeS2), sphalerite (ZnS), and galena (PbS). Mining increases the surface area of sulfide ores exposed to air and water, thus, increases rates of acid generation. Regions where rocks have low buffering capacity generate highly acidic toxic solutions that are referred to as acid mine drainage (AMD).

Despite the extreme acidity, heat, and high concentrations of sulfate and toxic metals, a diverse range of microorganisms populate AMD environments. These organisms can form a chemoautotrophically-based biosphere in the subsurface, ultimately sustained by electron donors derived from sulfide minerals, CO2, O2, and N2 derived from air, and phosphate liberated by water–rock interaction. Microbial activity increases the rate of AMD formation and may be responsible for the bulk of AMD generated [1].

Microbe–mineral interactions are of importance because AMD is a very widespread environmental problem. The organisms can be used in ore processing and are a source of novel biomolecules (especially enzymes) for industrial processes.

DNA-based studies of organisms populating mining environments have provided insights into the diversity of acidophilic, metal-tolerant species. Here, we review the importance of archaeal and bacterial lineages, and integrate microbiological, geochemical, mineralogical, and molecular information necessary for quantitative descriptions of the ecology of AMD. Eukaryotes (protists, fungi, and yeasts) are abundant and important in some parts of acid systems. However, due to the paucity of data on eukaryotes in AMD, our review focuses primarily on the prokaryotic components of these communities. We show that the prokaryotic richness of acidophilic communities is low compared to other extremophile and non-extremophile assemblages, yet the species are broadly distributed across the tree of life. Because of their biological and geochemical simplicity, AMD environments have potential as model systems for analysis of biogeochemical interactions and feedbacks and microbial community structure and function.

2 Dissolution of sulfide minerals

Many factors impact AMD generation. In part, rates of dissolution reactions are determined by fluid chemistry and flow, mineral and rock type, and temperature. The rate of supply of oxidant to the mineral surface influences the rate at which pyrite dissolves. The typical oxidants are oxygen and ferric iron. The concentration of oxygen in groundwater is very small compared to the large requirement for O2 in the overall reaction: Embedded Image 1 Thus, the predominant source of oxygen in rapidly oxidizing systems is air. In fact, to create typical AMD, each packet of solution must be re-oxygenated hundreds to thousands of times along its flow path [2].

Geochemical studies have established that oxygen is a less effective sulfide oxidant than ferric iron. Thus, the dominant pathway for pyrite dissolution involves oxidation of ferrous iron by oxygen: Embedded Image 2 followed by reduction of ferric iron by sulfide: Embedded Image 3 Note that the sum of reactions in Eqs. 2 and 3, required to describe the sustainable process, yields the reaction in Eq. 1. Ferrous iron oxidation by O2 at low pH is slow, thus the rate of the reaction in Eq. 2 may limit the rate of AMD generation. However, iron-oxidizing prokaryotes catalyze ferrous iron oxidation, thus can determine the rate of pyrite dissolution [1,3]. The feedback between metabolic activity and mineral dissolution can drive the pH down to values <2, thus selecting for community members optimized for life in acid.

Physiological experiments have shown that AMD microorganisms can impact rates of sulfur oxidation during dissolution of pyrite, arsenopyrite, chalcopyrite, marcasite, and sphalerite [4,5]. Oxidation of sulfide ions to sulfate occurs via a series of intermediate sulfur-bearing compounds. Because protons are generated in the subset of reactions that add oxygen to the reduced sulfur species, microbial utilization of sulfide and intermediate sulfur compounds can significantly affect acidification, as well as pyrite dissolution rates.

There is debate about the details of the mechanisms by which microbe–sulfide mineral interactions occur. Pyrite oxidation has been proposed to proceed via indirect and direct pathways [1]. Increase in the rate of sulfide mineral dissolution due to oxidation of aqueous ferric iron is typically referred to as an ‘indirect’ mechanism. Oxidation of intermediate sulfur species can occur on the mineral surface or in solution. In contrast, ‘direct’ catalysis is used to describe the possibility of a direct enzymatic interaction with ions bound to the pyrite surface [6].

Enzymatic oxidation via a direct mechanism requires that the cells are either in close proximity or attached to the solid surface. Larsson et al. [7] showed that close proximity of the archaeon Acidianus brierleyi to pyrite was necessary for optimal growth and oxidation rates. Cell-sized pits, often observed on pyrite surface after reaction with Acidithiobacillus ferrooxidans [8], have contributed to speculation about a ‘direct’ enzymatic pathway.

Experiments were conducted by Edwards et al. [9] to determine whether cell attachment was necessary to generate cell-sized pits. High-resolution scanning electron microscopy (SEM) was used to characterize dissolution patterns on sulfides by Ferroplasma acidarmanus (at 37°C) and Acidithiobacillus ferrooxidans (at 25°C). Elliptical pits (1–2 μm) formed on pyrite surfaces in the presence of these species. However, they also formed in abiotic experiments with just Fe3+, indicating that the presence of cells was not required for pit formation. Cell-sized pits were not observed on marcasite or arsenopyrite reacted with A. ferrooxidans, or on marcasite surfaces reacted with F. acidarmanus. However, rounded F. acidarmanus cells were found within round, deep pits (<0.5 μm) in arsenopyrite surfaces, clearly indicating enhanced dissolution in proximity to cells.

Fowler et al. [10] used constant ferric and ferrous ion concentrations in continuous flow reactors, with and without Acidithiobacillus ferrooxidans, to determine if that organism enhances leaching of pyrite by a mechanism other than increasing Fe3+ in the reactor. Dissolution rates with cells were faster, implying the presence of leaching enzyme(s) or a localized accumulation of Fe3+ at the cell surface.

Edwards et al. [11] conducted surface colonization experiments with pyrite in situ at Iron Mountain, CA, USA and in the laboratory using enrichments from the site. They found that bacteria tended to orient parallel pyrite surface steps in {110} and {100} orientations. Edwards et al. [5] also found that attachment of Acidithiobacillus caldus to pyrite, marcasite, and arsenopyrite was non-random. No cells attached to quartz inclusions within pyrite, indicating that attachment is mineral specific [12]. These observations suggest that cells are not simply using Fe2+ in solution, but exhibit chemotaxis, probably to gradients in [Fe2+]aq outward from dissolving surfaces.

In addition to ‘indirect’ and ‘direct’ approaches to sulfide mineral dissolution, cells may make soluble organic molecules to shuttle electrons from insoluble elemental sulfur on pyrite surfaces to cell-associated enzymes [13]. It is also possible that oxygen and hydroxyl radical species are important in some sulfur transformations [2]. Alternatively, Sand et al. [6] proposed that cells attach to the mineral surface and accumulate Fe3+ in their outer membranes, causing local enhanced reactivity via a non-enzymatic direct mechanism.

Microorganisms accelerate iron oxidation rates at low pH by five orders of magnitude [14], thus can increase rates of pyrite dissolution by the reaction in Eq. 3. Numerous microbial pyrite dissolution rates have been reported. Typical experimental values range between 10−5 and 10−7 mol m−2 s−1 for microbial and ferric iron experiments compared to 10−6 to 10−9 mol m−2 s−1 for inorganic oxygen-mediated rates [11].

The degree to which microorganisms enhance sulfide mineral dissolution is determined by the number of iron-oxidizing cells present and the level of activity of the cells. These parameters are not incorporated in most reported microbial rates. Edwards et al. [11] determined the increase in release of iron from pyrite in two enrichments containing Leptospirillum ferriphilum to be 5×10−18 mol Fe cell−1 s−1, comparable to the Thiobacillus ferrooxidans iron oxidation rate of 6×10−18 mol Fe cell−1 s−1[15]. Using cell-normalized rates, it is possible to estimate the magnitude of the microbial impact in a natural environment if information about numbers of iron-oxidizing microorganisms is available. Using this approach, Edwards et al. [5] suggested that microbial activity accounts for about 75% of the AMD produced.

3 Diversity

It is essential to define the microbial population in order to analyze an ecological system. Currently, the primary method of determining which organisms comprise a microbial community involves determining the sequence of 16S ribosomal RNA genes (16S rRNA) from environmental samples. This approach provides information about species richness as well as the evolutionary relationships between lineages. Microbial species composition can then be correlated with environmental data to determine how communities are shaped by geochemical factors.

Any AMD system has many microbial niches. Variations in temperature, ionic strength, and pH yield communities characterized by different mixtures of organisms, but all habitats are restricted in membership to a few species (Table 1) [16,17]. This can likely be attributed to the limited number of energy-deriving reactions available in AMD environments. The number of substrates multiplied by the number of geochemically distinct habitats might be loosely related to the total number of AMD-adapted species. This would predict that the total number of AMD lineages would be small, given that AMD solution chemistry is strongly controlled by pyrite dissolution (see details below). Microbes from 11 putative divisions have been detected in AMD environments (Fig. 1).

View this table:

List of prokaryotic diversity of acid mine drainage communities relative to other phylogenetically well-characterized environments

EnvironmentNumber of putative divisionsNumber of genera and novel lineagesaReference
Arid southwestern US soil7+25[18]
Wisconsin agricultural soil17+50[19]
Grass pasture rhizospheres22145[20]
Contaminated aquifer2556[21]
Marine bacterioplanktonic community937[22]
High-temperature petroleum reservoirs1928[23]
Mid-Atlantic Hydrothermal Vent9+19[24]
Yellowstone N.P. hot spring2832[25]
Antarctic sea ice and water442[26]
Deep subsurface paleosol9+42[27]
Iron Mountain Mine, CA 199846[16]
Total of all studied AMD sites1116compiled here
  • ‘+’ indicates that clones of unresolved phylogenetic positions were recovered.

  • aNovel lineages were counted subjectively by phylogeny provided in paper, each deeply branched cluster was counted.


Phylogeny of prokaryotic 16S rRNA genes from acid mine drainage and bioleaching sites (in bold) with reference lineages. * indicate lineages known to contain facultative anaerobes and φ appear to be obligate aerobes. The lineages in blue have no cultivated members to date. Lineages in red are not known to be able to utilize sulfur, only iron oxidizers. Putative divisions are shown near each of their branchings, the β-, γ-, δ-, α-proteobacteria are shown. Tree generated using neighbor joining method in ARB software package. Bar represents 0.1 changes per site or 10% difference in sequence, IM=Iron Mountain.

The AMD system most intensively studied by culture-independent molecular methods is within the Richmond Mine at Iron Mountain in northern California. The methods used include 16S rDNA clone library analyses and cell imaging using multiple fluorescent probes designed to bind to the ribosomal RNA with varying levels of specificity. Pyrite-associated communities, subaqueous slime streamers, and subaerial biofilms in pH 0.5–1.4, 27–50°C solutions have been monitored and community responses to seasonal variations have been determined. Results confirm that a handful of prokaryotic taxa (often less than five groups distinct at the genera level) make up the communities in any specific microenvironment [17,28]. Low diversity has also been noted using cultivation-based approaches [29,30].

3.1 Bacteria

Among the bacterial lines of descent are divisions within the proteobacteria, nitrospira, firmicutes, and acidobacteria. The most extensively studied group (but possibly the least relevant under AMD-generating conditions, see below) are the γ-proteobacteria, specifically Acidithiobacillus spp. (formerly T. ferrooxidans, Thiobacillus caldus [31]) and Thiobacillus spp.

Two β-proteobacterial groups have been detected to date. Among this group are Thiomonas sp. (strains Ynys1 and Ynys3 [32]) and subsequently an isolate designated NO-16 from a Norwegian mine [30].

There are six species of heterotrophic α-proteobacteria of the genus Acidiphilum in pure culture. Acidiphilum cryptum (strain JF-5) was originally isolated from acidic (pH 3.0) iron-rich sediment from a lake associated with a coal mine in eastern Germany [33]. Another group of α-proteobacteria have been detected in clone libraries constructed from pH<1.0 Iron Mountain AMD environments. However, these live in near neutral pH habitats inside acidophilic protists [34].

Recently, a novel group of δ-proteobacteria clones were recovered from Iron Mountain (Fig. 1) [16,17]. However, because this group has defined cultivation, their metabolic characteristics remain uncertain.

Leptospirillum ferrooxidans’-group organisms are commonly detected in AMD and bioleaching systems. To date, Leptospirillum isolates and environmentally-derived clones cluster within one of three phylogenetically distinct groups (I, II, and III), as designated by Bond et al. [16]. Two species have been named: L. ferrooxidans [35] (group I) and L. ferriphilum (group II) [36]. To date, group III Leptospirillum has only been detected via clone library analysis of Iron Mountain microbial communities [16,17].

Three distinct groups of AMD organisms fall within the firmicutes division. These include an Actinobacteria group containing Acidimicrobium ferrooxidans and Ferromicrobium acidophilus (Fig. 1) and the polyphyletic (two lineages) low G+C Gram-positive Sulfobacillus. Sulfobacillus thermosulfidooxidans VKLM and Sulfobacillus disulfidooxidans SD-11 belong to the original Sulfobacillus spp. group, which is related to the Alicyclobacillus lineage (Fig. 1). Another large bacterial group named Sulfobacillus, represented by some S. thermosulfidooxidans and Sulfobacillus acidophilus, is distinct from the group represented by ‘Sulfobacillus’ and separate from the Alicyclobacillus lineage. We suggest that in the future organisms within the Alicyclobacillus (including the type strain VKLM) be formally renamed to Alicyclobacillus based on phylogeny (Fig. 1).

The first isolate of the Acidobacteria division, Acidobacterium capsulatum, was recovered from an AMD site by Kishimoto et al. [37]. Closely related environmental clones have been reported from Iron Mountain (Fig. 1) [12,17].

3.2 Archaea

Archaeal lineages reported from AMD environments are restricted to the Thermoplasmatales and the Sulfolobales. A number of clones phylogenetically divergent from Thermoplasma spp. have been detected in clone libraries (Fig. 2) [16,17]. Further surveys and isolation efforts are needed to resolve the tentative groupings within this lineage, currently referred to as the ‘alphabet plasmas’ (Fig. 2).


Phylogeny of AMD-associated archaea within the Thermoplasmatales. Putatively proposed group names (the ‘alphabet plasmas’) shown on right and nodes for each corresponding group marked with diamonds. Clones from Iron Mountain site are shown in bold type. Tree generated via maximum likelihood method in ARB software package, outgroup not shown. Bar represents 0.1 changes per site, or 10%. Baker clones from [17].

One member of the Sulfolobales, Metallosphaera prunae, has been detected in AMD environments. Two other Sulfolobales genera, Acidianus and Sulfolobus, have only been found in geothermal acidic environments, thus are not considered further here.

3.3 Eucarya

There have been a few reports of eukaryotes in AMD environments [3841]. Ciliates belonging to Cinetochilium genus and an amoeba related to Vahlkampfia sp. within the lineage Heterolobosea, and three flagellates (Eutreptial spp.) were isolated from an AMD site and shown to be able to graze on mineral-oxidizing acidophilic bacteria [42]. Lutz et al. [38] recovered clones related to Vahlkampfia sp. from Iron Mountain. At Iron Mountain fungal filaments provide structure to subaqueous biofilms termed ‘slime streamers’ (Figs. 3 and 4). Recently, Amaral Zettler et al. [43] reported a diverse community eukarya (including alga) present in the pH 2 Rio Tinto River, Spain. In this instance, eukaryotes comprised at least 60% of the total biomass. It remains to be seen if subsurface non-phototrophic protistan communities growing in more acidic environments (such as encountered at Iron Mountain) are as diverse. Metabolically active protists are associated with Richmond Mine biofilms (pH 0.5–2.0; Fig. 3).


Fluorescent in situ hybridization (FISH) biofilm samples from Iron Mountain Mine, CA. All images are the result of probes labeled with Cy3 colored with 4′,6-diamidino-2-phenylindole (DAPI) DNA stain for background cells. A is Sul228 specific for Sulfobacillus spp. [16], B Arc915 for archaea, C demonstrates archaea cells attached to fungal filaments, and D is eukarya-specific probe (Euk502). Note the nuclei of the protist are DAPI stained and many present in the sample are targeted by the Euk502 probe.


High-resolution TEM images of Richmond Mine, CA environmental samples. A, C: A drift snottite. B: A slump subaerial slime. The size of the bar in B and C is 200 nm. Coccoid structures in A are less than 1 μm in diameter.

4 Evolution of acidophilic metal-tolerant organisms

It is widely believed that early Earth environments were relatively anoxic prior to the appearance of bacterial oxygenic photosynthesis. Given this, and the low abundance of sulfates in the early Earth record, it is likely that ARD environments were rare in the Archaean. Once the concentration of O2 increased, new metabolic options, such as oxidation of iron and sulfur coupled to reduction of O2, may have stimulated diversification within existing lineages, explaining the broad distribution of acidophiles over the tree of life (Figs. 1 and 2).

ARD and, more recently, AMD environments exist as isolated point sources that are exposed then removed via erosion within a few thousand to a few millions of years. Lateral gene transfer (LGT) is a mechanism by which some AMD survival genes could be introduced to create new acid- and metal-tolerant lineages. Recent analyses of LGT provide evidence for exchange of genes between extremely acidophilic trading partners [45]. However, there is no evidence to suggest that AMD organisms evolve from non-extremophiles when local acidic environments appear. In fact, organisms that colonize AMD environments possess habitat-specific genes (e.g., involved in metal resistance) [44] whose phylogeny matches the 16S rRNA gene phylogeny and modern inhabitants of AMD environments have very close relatives that are widely geographically distributed. Organisms are apparently introduced to newly created extremely acidophilic habitats from distant sites via as yet poorly understood dispersal mechanisms. The survival of these lineages over geologic timescales depends upon this.

5 Metabolic options and pathways

In regions of AMD systems exposed to sunlight, photosynthesis is an important source of energy and fixed carbon [46]. However, in subsurface ARD and AMD systems, inputs of externally-derived fixed carbon and nitrogen are minimal. The primary metabolic groups detected in AMD-generating regions are lithoautotrophs that oxidize Fe2+ and S released by pyrite dissolution, organoheterotrophs that utilize carbon produced by lithoautotrophs, lithoheterotrophic Fe and S oxidizers, and anaerobes that couple oxidation of sulfur or organic carbon to Fe3+ reduction. A subset of these organisms must produce all fixed carbon and nitrogen required by the community.

In addition to S, intermediate sulfur and sulfoxy oxidation products can serve as electron donors. Electrons from these compounds are passed either to O2 or Fe3+ (generated by oxidation of Fe2+ by O2) because these are typically the only suitable electron acceptors available in significant concentrations.

Utilization of Fe3+, and possibly sulfate, as electron acceptors is possible under microaerophilic and anoxic conditions. O2 consumption by microbes and pyrite dissolution can induce some typically aerobic species such as A. cryptum, Sulfobacillus, and possibly F. acidarmanus and related ‘alphabet plasmas’ to grow anaerobically. To date, no obligate anaerobes have been cultivated from an AMD habitat. However, uncharacterized d-proteobacteria phylogenetically closely related to anaerobic species have been detected [16,17].

The acidophilic lineages documented to date have diverse evolutionary histories. If, as suggested above, most arose after the divergence of the three domains, the enzymes involved in the key transformations may be diverse. Our ability to test this hypothesis is limited due to the restricted amount of biochemical and genome sequence data currently available. Work of Blake and coworkers [47] suggests that the critical process of iron oxidation involves unrelated enzymes in most lineages examined to date.

5.1 Iron oxidation

The autotrophic, facultative anaerobic Fe oxidizer Acidithiobacillus ferrooxidans (formerly T. ferrooxidans) is readily cultivated and has been the focus of extensive genetic, genomic, and physiological studies [4,10,65]. Consequently, it is the only organism for which there are biochemical models of the electron transport chain for iron oxidation.

All characterized L. ferrooxidans and L. ferriphilum isolates are iron oxidizers [48,49,36,50,29]. Although no members of Leptospirillum group III (Fig. 1) have been isolated yet, it is likely that these are also iron oxidizers. S. acidophilus, S. thermosulfidooxidans, and Acidimicrobium ferrooxidans strains TH3 and ICP are also iron oxidizers [5153].

A few archaea associated with AMD have been shown to oxidize iron. Ferroplasma acidiphilum is an obligate autotroph able to oxidize ferrous iron as a sole energy source [54]. F. acidarmanus can oxidize ferrous iron and can grow heterotrophically [55]. A crenarchaeota, Metallosphaera sedula, can oxidize iron and sulfur at high temperatures (optimal growth at 74°C and pH 2.0). Thus, it is a candidate for high through-put industrial bioleaching [56].

5.2 Iron reduction

AMD solutions are iron rich because ferric and ferrous iron are very soluble at low pH (<2.5). In some cases, Fe3+ concentrations in AMD may exceed oxygen concentrations by several orders of magnitude [17]. Thus, Fe3+ may be widely used as an electron acceptor in microbial metabolism.

Johnson and McGinness [57] showed that the ability to reduce soluble Fe3+ is widespread among heterotrophic acidophiles. Moderately thermophilic species S. thermosulfidooxidans (strain TH1), S. acidophilus (strains ALV, THWX, and YTF1), and Acidimicrobium ferrooxidans (TH3) use ferric iron as the sole electron acceptor when grown heterotrophically under oxygen-limited conditions (with yeast extract and glycerol [57]). Thiobacillus acidophilus and a number of pure cultures of mesophilic heterotrophs are capable of mixotrophic Fe3+ reduction [58]. A. cryptum can couple Fe3+ reduction to oxidation of a variety of substrates including H2 and glucose. For example, Kusel et al. [59] demonstrated that A. cryptum strain JF-5 couples oxidation of glucose to reduction of either O2 or soluble Fe3+.

Some microorganisms can utilize Fe3+ even if it is not in solution. S. acidophilus is capable of anaerobic dissolution of iron hydroxide, jarosite, and goethite [51]. Bridge and Johnson [57] demonstrated that A. cryptum strain SJH is capable of reductive dissolution of a wide range of ferric iron minerals (akageneite, goethite, jarosite, natrojarosite, magnetite, and amorphous ferric hydroxide). They reported that contact was necessary for dissolution to occur and that rates were faster at lower pH (2.0 vs. 2.8–3.0).

Anoxic conditions are not required for Fe3+ reduction by mixed cultures of acidophiles [58]. Recently, Kusel et al. [59] demonstrated that under oxic conditions JF-5 reduced soluble Fe3+ and schwertmannite (a Fe3+-sulfate mineral) in sediment microcosms at pH 3.

In addition to coupling Fe3+ reduction to oxidation of organic carbon or hydrogen, some organisms can use Fe3+ to oxidize sulfur compounds. Bridge and Johnson [57] showed that S. acidophilus strains THWX, ALV, and TH1 use ferric iron to oxidize tetrathionate anaerobically.

5.3 Sulfur oxidation

Most members of AMD communities that oxidize sulfur also can fix CO2. An exception is A. cryptum, which may utilize sulfur oxidation to some extent but is normally considered to be an organoheterotroph [60]. T. acidophilus can oxidize sulfur, iron, and organic carbon and is capable of autotrophic growth [61]. Autotrophic sulfur oxidation by pure cultures of Acidithiobacillus ferrooxidans [62], Thiobacillus thiooxidans [63], Sulfobacillus (more correctly Alicyclobacillus) disulfidooxidans SD-11 [64], T. albertis [65], S. acidophilus [66] has been demonstrated. A. ferrooxidans also can grow under anoxic conditions using ferric iron as the electron acceptor and S° the electron donor [62].

Several autotrophic sulfur oxidizers can also oxidize organic carbon and/or ferrous iron. For example, S. disulfidooxidans is capable of mixotrophic S oxidation [64], but does not oxidize iron. However, SB37, a recent S. disulfidooxidans isolate of close phylogenetic association, is able to oxidize Fe2+ (Baker, unpublished data). Fourteen strains of Gram-positive sulfide mineral-oxidizing bacteria, including S. acidophilus (one being ALV) and S. thermosulfidooxidans (including TH1), can grow autotrophically on ferrous iron and S°, mixotrophically in the presence of yeast extract, and heterotrophically on just yeast extract [66].

A variety of sulfur compounds with oxidation states intermediate between 2− (e.g., sulfide in sphalerite) and 6+ (sulfate) form during metal sulfide oxidation (e.g., polysulfide: Sn2−, elemental sulfur: S°, thiosulfate: S2O32−). Some of these can be utilized by AMD microorganisms. Elemental sulfur and tetrathionate (S4O62−) are thought to be especially biologically important because they are relatively stable in low pH solutions. T. acidophilus grows autotrophically on trithionate (S3O62−), elemental sulfur, and tetrathionate [66,67]. As noted above, some strains of S. acidophilus couple autotrophic oxidation of tetrathionate to reduction of Fe3+[51,66]. It is interesting to note that none of the strains described as Thiobacilli that were adapted to higher temperatures (∼50°C) was capable of tetrathionate oxidation [66]. This is likely due to a decrease in bioavailability because of faster reaction rates.

5.4 Other metabolic requirements

Most subsurface AMD sites receive relatively minimal inputs of fixed carbon and nitrogen from external sources. The Thiobacilli (including Acidithiobacillus ferrooxidans) are probably the dominant group responsible for CO2 fixation in lower temperature (<30°C), higher pH (>2) environments. In lower pH, higher temperature communities, autotrophic taxa include Leptospirillum spp., Ferroplasma spp., Sulfobacillus spp., Ferromicrobium spp., and Acidimicrobium spp. At high temperatures (∼65–95°C) archaea such as Metallosphaera spp. may be the key sources of fixed carbon for the AMD biome. Although it is known that the Acidithiobacillus ferrooxidans pathway involves a bc1 and an NADH-Q oxidoreductase complex functioning in reverse cycle for CO2 fixation [68], little is known about the pathways used by other acidophilic autotrophs.

Fixation of N2 in largely aerobic and microaerophilic AMD environments is potentially problematic due to inhibition of nitrogenase by O2. Acidithiobacillus ferrooxidans, a moderate acidophile, may overcome this problem by using tetrathionate as an electron donor and ferric iron (rather than O2) as an electron acceptor when fixing nitrogen [69]. The fixation of N2 in very low pH environments is enigmatic, because no one has observed it directly. L. ferriphilum has been shown to possess nifH genes [70]. However, attempts to demonstrate active nitrogenase in pure cultures of L. ferriphilum were unsuccessful. Determination of which organisms are responsible for N2 fixation in the full range of AMD environments is an important goal for further work.

6 Abundance, community structure, and physical/chemical regimes

Studies over the last decade indicate microbial communities in acidic environments are dominated by the Thiobacilli. It now appears this is an artifact of the culturing and sampling methods used, at least in some cases. Schrenk et al. [71] and Edwards et al. [72] used fluorescent in situ hybridization (FISH) probes to quantify the relative abundances of Acidithiobacillus ferrooxidans within the Richmond Mine at Iron Mountain (pH 0.3–0.7, 30–50°C). They reported that A. ferrooxidans were nearly undetectable. Subsequent studies demonstrated that A. ferrooxidans occur in relatively low numbers in pyrite-dominated bioreactor systems [73]. The failure of A. ferrooxidans to thrive in more extreme AMD environments is attributed to their mesophilic growth optima (26°C) and moderately acidophilic nature (pH 1.3–4.5). Druschel et al. [17] did not obtain any A. ferrooxidans clones in six libraries constructed from six pH<1.0 AMD communities associated with the Richmond Mine. However, they demonstrated that an A. ferrooxidans strain was abundant in an oxidized, pH 1.4 pool.

Schrenk et al. [71] used FISH probes to show that A. ferrooxidans cells are abundant in environments with pH>1.3 and temperatures <30°C at the Richmond Mine. In this case, the higher pH, lower temperature environments occur outside the ore body, thus away from sites of active pyrite dissolution. However, pH 2–4, <30°C environments are encountered in many AMD systems early in weathering, when rocks are less enriched in pyrite, or when the buffering capacity is higher.

Leptospirillum group isolates have been obtained from AMD environments characterized by a wide range of temperatures and pHs. They are as abundant, or more abundant, in bioleaching systems than A. ferrooxidans over much of the A. ferrooxidans growth range. Sand et al. [74] note that the growth rate of A. ferrooxidans only outcompetes that of L. ferrooxidans (group I) below 14°C. They also note that in their bioreactors high ratios of Fe3+ to Fe2+ appear to be less inhibitory to Leptospirillum than to A. ferrooxidans.

L. ferrooxidans DMZ2705 (group I) is reported to grow in the pH range of 1.3–4.0, with an optimal pH range of 1.6–2.0 (compared to 1.4–1.8 for L. ferriphilum) [36], and an upper temperature limit around 55°C [75,35]. Goebel and Stackerbrandt [29] isolated strain Lf30-A (group II, thus L. ferriphilum) from a continuous bioreactor (pH 1–2 and 35–40°C) and showed that the isolate was capable of growth at 28, 37, and 45°C. Coram and Rawlings [36] noted that L. ferriphilum-dominated commercial biooxidation tanks operated at 40°C, and that some strains were able to grow at 45°C (no L. ferrooxidans strains were capable of growth at 45°C). Golovacheva et al. [50] characterized an isolate from an iron-containing hydrothermal spring (pH 2.0, 45°C) and named it ‘Leptospirillum thermoferrooxidans’ (also see [35]), based on greater G+C content and higher growth temperature (45–60°C) than previously described ‘L. ferrooxidans’ strains. The status of isolates described as ‘L. thermoferrooxidans’ is uncertain because these strains are not available for molecular characterization.

Oligonucleotide probe-based studies within the Richmond deposit indicate that Leptospirillum strains often dominate microbial communities growing at temperatures up to 50°C in solutions with pH values as low as 0.5 [16,17,71,72]. Leptospirillum (group III) were shown to comprise the majority of bacteria in subaerial slimes on the surface of a ‘slump’ of fine-grained pyrite ore [16] and, at times, to dominate subaqueous environments otherwise populated by L. ferriphilum [17]. L. ferriphilum and Leptospirillum group III primarily reside in lower pH microenvironments in the mine, while L. ferrooxidans occurs in higher pH environments (>1) [17].

There are several other groups of bacteria whose abundance and distribution have been studied in AMD. F. acidophilus has a pH range of 1.3–4.8 and is adapted to temperatures from <20 to 40°C [75]. Environmental clones closely related to F. acidophilus were recovered from the Richmond Mine AMD system (43°C) [16,17]. Ferromicrobium spp. were shown to be minor community members via probe-based studies [76]. Acidimicrobium ferrooxidans has been cultivated at temperatures between 34 and 57°C [53] and has been reported from a surprisingly diverse range of environments [77].

Sulfobacillus spp. have a broad range of physical growth regimes. Some isolates are capable of growth up to 65°C [78]. Recently, Yahya et al. [79] described Sulfobacillus-like strains (L-15 and Riv-14) isolated from geothermal sites. The strains are effective pyrite oxidizers at pH<1, suggesting that thermophilic species may be encountered in AMD systems. The FISH image in Fig. 3 shows Sulfobacilli rods within a Richmond Mine biofilm growing at pH∼0.7. Sulfobacillus clones have been recovered from pH 0.7–0.9, 35–43°C [16] and have been shown to comprise up to ∼6–8% of microbial communities (in pH 0.8–1.4, 27–32°C environments [17]).

Six members of the genus Acidiphilum are adapted to temperatures ranging from 17 to 45°C and pH values from 1.5 to 6.0 [75]. Strain SJH has a pH optimum of ∼3.0 [57]. One isolate was retrieved from a continuous bioreactor [29] and a clone was identified from the pH<1.0 solutions at Richmond Mine [12], though later studies at this site have not detected this group [16,17]. Peccia et al. [80] used a 16S rRNA probe to determine the abundance of Acidiphilum in sediments from an active AMD site in Colorado and in laboratory reactors. They found that the Acidiphilum outnumbered Acidithiobacillus ferrooxidans in their mixed culture bioreactors and that Acidiphilum and A. ferrooxidans populations were approximately equal in the sediment samples. In general, Acidiphilum sp. tends to occupy lower temperature, higher pH microenvironments. Under anoxic conditions they may contribute to iron cycling by redissolving ferric iron-based minerals precipitated when the pH is increased by mixing with groundwater and in streams.

The metabolic role of the δ-proteobacteria has not been elucidated because no members have been isolated. As δ-proteobacteria are often anaerobic sulfate or metal reducers, Bond et al. [16] suggested that anaerobic microenvironments may exist within the slime samples from which the clones were recovered. Attempts to amplify the highly conserved dissimilatory sulfite reductase (Dsr) gene were unsuccessful, and therefore it has not been possible to confirm these bacteria as sulfate reducers (Baker and Banfield, unpublished data).

Acidobacteria spp. populate relatively moderate AMD environments (20–37°C and pH 3.0–6.0). Recent analyses suggest that this group is limited to the higher pH environments (∼pH 1.4) at the Richmond Mine site [17].

Application of culture-independent approaches led to recognition that archaea are important and abundant in AMD environments. Iron-oxidizing archaea of the genus Ferroplasma have been isolated from a bioleaching plant [54] and the Richmond Mine [55]. Ferroplasma lack a cell wall and are bounded only by a cytoplasmic membrane. However, like other Thermoplasma and Sulfolobus spp., cells may possess an S-layer (Fig. 4) [81,82,76]. F. acidarmanus, isolated from the Richmond Mine, was shown to be the dominant organism in some habitats (see Fig. 3), comprising as much as 85% of the population in pH∼0.5, ∼40°C, high ionic strength habitats [55]. F. acidarmanus has a lower pH growth range than F. acidiphilum (0–2.5 vs. 1.3–2.2).

Clones within the Thermoplasmales (e.g., Thermoplasma acidiphilum and Thermoplasma volcanium), but distinct from Ferroplasma spp., have been detected in clone libraries created from samples collected from many very acidic geothermal hot spring environments [83]. T. acidophilum and T. volcanium have moderately thermophilic (45–67°C) growth ranges [81,82] and are adapted to pH ranges typical of AMD. T. acidophilum has been isolated from a coal refuse pile [81].

The metabolic role of the ‘alphabet plasmas’ within the order ‘Thermoplasmatales’ (Fig. 2) [16,17] is yet to be determined. Their environmental distribution indicates adaptation to high biomass, metal-rich, pH 0.5–1.4, 30–50°C habitats close to the air–biofilm interface [17]. Within the Richmond Mine system, all archaea are low abundance or absent from in pH 1.3–2.4 habitats [17,71]. Metallosphaera spp. (Sulfolobales order) can impact pyrite dissolution via catalysis of ferrous iron oxidation [84] and has been isolated from an acidic uranium mine [85] and a bioleaching reactor [56].

7 Synergistic interactions involving iron, sulfur, and carbon oxidizers

As indicated by the above discussion, community membership varies with pH and temperature, mineralogy, as well as with metal concentrations. However, in most environments (particularly in the subsurface) some species, or combination of species, within communities probably carry out iron oxidation, sulfur oxidation, organic carbon oxidation, carbon fixation, nitrogen fixation, extracellular polymeric slime production, as well as iron and sulfur reduction. It is also apparent from the above review that microbial communities in AMD systems tend to contain few distinct taxa (though these taxa may be remarkably phylogenetically diverse). Fig. 5 illustrates a simplified AMD system and provides examples of organisms that may control flow of iron, sulfur, nitrogen, carbon, and energy in different temperature regimes.


Potential iron, sulfur, and carbon cycling based on known metabolic capabilities (1, 2, 3, and 4) associated with AMD members. Crystalline pyrite (Fe2S) is in yellow at the bottom and green is representing AMD solution. Elemental sulfur is shown at the pyrite–water interface as a possible inhibitor of surface dissolution. The overall oxidation of pyrite is shown at the bottom, with Fe3+ indicated as the primary oxidant. Intermediate sulfur compounds are indicated as follows: S2O32− being thiosulfate and S4O62− is tetrathionate. C30H60O30N6P indicates organic carbon compounds.

Interactions between members of microbial consortia are probably critical in optimization of AMD microbial community activity. For example, an important symbiosis exists between heterotrophic and certain autotrophic species: autotrophs may depend on coexisting heterotrophs to remove organic compounds that are toxic to them. Heterotrophic acidophiles are able to utilize organic materials produced by acidophilic autotrophs [86,56]. The culture filtrate from the autotroph Acidithiobacillus ferrooxidans contained sufficient organic matter to support heterotrophic growth of S. thermosulfidooxidans (strain TH1) [87]. Clark and Norris [53] conducted mixed culture experiments with Acidimicrobium ferrooxidans and both S. acidophilus and S. thermosulfidooxidans. They found iron oxidation by mixed cultures of Sulfobacillus spp. and Acidithiobacillus ferrooxidans was more extensive than in pure cultures containing any isolate. They attributed this greater leaching in mixed cultures to mixotrophic growth of the Sulfobacillus removing organic byproducts of A. ferrooxidans. This may explain the close correlation in populations of A. ferrooxidans/T. thiooxidans and Acidiphilum in environments studied by Peccia et al. [80].

Raman spectroscopy has been used to show that abundant S° forms on the surfaces of arsenopyrite [88] and galena [89] and pyrite [90] during oxidative dissolution. It has been hypothesized that microbial removal of elemental sulfur (S°) and other compounds is critical for continued access of the Fe3+ oxidant to pyrite surface sulfur-bearing groups. Dopson and Lindstrom [91] examined the effect of defined mixed cultures of the sulfur oxidizers A. caldus and S. thermosulfidooxidans (capable of Fe and S oxidation) on arsenopyrite dissolution. By following concentrations of released iron, tetrathionate, and sulfur in the cultures they found that A. caldus grows on surface sulfur compounds, resulting in an increase in the arsenopyrite dissolution rate. They also suggested that A. caldus produces surface-active agents to mobilize the extremely hydrophobic S°. Subsequently, Hu et al. [90] determined that elemental sulfur accumulations on galena significantly decrease dissolution rates. Because removal of S° from mineral surfaces is insignificant in abiotic controls, microbial oxidation of sulfur is likely to be a key determinant of acid generation rates (because all protons are released in conversion of sulfur to sulfate).

8 Outlook

Perhaps the most exciting aspect of AMD communities is that they provide the opportunity to study largely self-contained biomes (independent of sunlight and most other life). Culture-independent methods have provided a detailed understanding of the full diversity and phylogeny of organisms populating AMD systems. Progress in understanding the biochemistry of acidophiles has been accelerated due to availability of genome sequences for a few isolates (e.g., F. acidarmanus and Acidithiobacillus ferrooxidans). However, we know essentially nothing about the subset of organisms that have not been isolated.

An important approach for analysis of uncultivated organisms is the construction and sequencing of large insert genomic libraries from environmental DNA [92]. Cultivation-independent sampling of genomes of acidophiles will reveal enzymes potentially involved in the major metabolic pathways and provide insights into pH homeostasis and metal resistance mechanisms. The low species richness of AMD communities may permit us to largely reconstruct genomes of populations from whole community libraries. Related gene expression studies should enable development of AMD ecosystem models that resolve the contributions of individual species at the molecular (pathway) level, describe and quantify flows of materials and energy, and identify symbioses and competitive strategies. Thus, through integration of geochemical and biological information, a comprehensive model for AMD production may be feasible. Studies of AMD systems could have broader significance. The remarkable simplicity of these habitats may permit a more fundamental understanding of how microbial communities work than is possible through study of more complex ecosystems.


We are grateful to Jenn Macalady and Philip Hugenholtz for critically reviewing this manuscript and discussion, and Ian Lo for preparation of FISH slides. We also thank Joe Coggliati (Stouffer Corporation), Rudy Carver (IT Corporation), Don Dodds (North Pacific Research), Ted Arman (Iron Mountain Mines), Richard Sugarek (EPA), and others for access to the field site and assistance with field work. This work was funded by the NSF LExEn and EGB programs and the US Department of Energy Microbial Genomics Program.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
  53. [53].
  54. [54].
  55. [55].
  56. [56].
  57. [57].
  58. [58].
  59. [59].
  60. [60].
  61. [61].
  62. [62].
  63. [63].
  64. [64].
  65. [65].
  66. [66].
  67. [67].
  68. [68].
  69. [69].
  70. [70].
  71. [71].
  72. [72].
  73. [73].
  74. [74].
  75. [75].
  76. [76].
  77. [77].
  78. [78].
  79. [79].
  80. [80].
  81. [81].
  82. [82].
  83. [83].
  84. [84].
  85. [85].
  86. [86].
  87. [87].
  88. [88].
  89. [89].
  90. [90].
  91. [91].
  92. [92].
View Abstract