Lovley 1991

Lovley DR. 1991. Dissimilatory Fe(III) and Mn(IV) reduction. Microbiological Reviews 55: 259-287.

This author comprehensively reviews microbe-mediated reduction of iron and manganese in soils and sediments, in a long and detailed paper. Dissimilatory reduction is distinct from assimilatory reduction, in which metal ions are reduced when they are incorporated into cellular macromolecules such as enzymes and cofactors. Dissimilatory reduction is a process that ends with the accumulation of reduced metal outside the cell, and is responsible for the majority of iron and manganese reduction in sediments. While it has been observed in aerobic environments, such reduction occurs mainly in anaerobic conditions. Fe(III) in particular is most often reduced when it is the final electron acceptor in the anaerobic oxidation of organic molecules.

Microbes capable of dissimilatory metal reduction can be categorized in a number of ways. This author presents 5 categories, though I think there is considerable overlap between them, as in cases where one cell is able to reduce Fe(III) and metabolize a range of carbon sources. In the first category, reducing fermenting bacteria, the amount of Fe(III) reduced during metabolism is far less than the stoichiometry of the redox couple would suggest. This implies that Fe(III) is a minor electron acceptor during fermentation reactions, rather than the electron acceptor of choice or necessity for these organisms. Sulfur-oxidizing species in contrast do appear to reduce quantities of iron in line with stoichiometric predictions, but do not seem to gain significant energy from these reactions which occur under aerobic conditions on elemental sulfur. Hydrogen-oxidizing reducers appear to be abundant in anaerobic sediments, and seem to have high affinity for hydrogen gas; where Fe(III) is being reduced, hydrogen concentrations are low, and when hydrogen is added, Fe(III) reduction increases. These organisms need other material, such as simple organic molecules, to grow, as this reaction provides energy but little else. Organic-acid oxidizing reducers and aromatic-oxidizers are probably in many cases the same cells. As these molecules are sometimes the result of fermentation metabolisms, one possible food chain in anaerobic environments is fermentation, with some Fe(III) reduction, followed by greater levels of Fe(III) reduction linked to the decomposition of smaller organic molecules such as acetate. Such an ecological pairing is probably widespread, given the known abundance and diversity of fermenting species and the probable abundance of more aggressively iron-reducing species, many of which are probably Archaea rather than Bacteria.

There are three competing models for how Fe(III) is reduced in natural environments. The first is the enzymatic model, in which microbial cells employ membrane-bound or intracellular enzymes to transfer electrons to Fe(III) ions during the process of anaerobic oxidation of organic matter. The second is termed the redox model, and posits the majority of Fe(III) reduction is driven by equilibrium thermodynamics, with the relative levels of Fe(III) and Fe(II) in sediments controlled by abiotic factors such as temperature and pH. The third model is termed the direct-reduction model, in which some organic molecules react directly with Fe(III), without the intervention of cells or enzymes.

At first glance, the observation that Fe(III) reduction rates fall when microbes are removed from sediments supports the enzymatic model, but in fact all three models rely on microbial metabolisms. In the redox model, competing electron acceptors such as oxygen and nitrate are consumed by microbes, lowering their concentrations to levels too low to influence the transitions between Fe(III) and Fe(II). And the direct model relies on microbes releasing key organic molecules known to reduce Fe(III) in vitro. However, several other lines of evidence, such as the lack of spontaneous shifts in iron valency ratios in stored sediments, the large changes in pH associated with widely-used extraction methods, and the general rarity of rapidly-acting direct-reducing molecules all lead to the conclusion that while the other mechanisms may contribute some iron reducing activity in some situations, the overwhelming majority of Fe(III) reduction to Fe(II) occurring is driven by microbes and their enzymes.

This author spends nearly as much time discussing Mn(IV) reduction as Fe(III) reduction, but I have little interest in Mn chemistry at this time. However, in the discussion of the competing models of metal reduction, mention is made that Fe(II) in solution may reduce Mn(IV), removing accumulated Fe(II) from iron reduction and abiotically returning the iron to Fe(III) while generating Mn(II). This abiotic back-reaction closes the loop on cycling Fe(III).

Iron reduction is postulated as one of the first, if not the first, globally important metabolic pathway, with early Archaea using Fe(III) as their final electron acceptor in a generally reducing environment lacking free oxygen and nitrate. If Fe(III), a non-soluble, precipitating substance is the primary oxidizing agent, the redox environment of the Earth 2 billion years ago was upside-down compared to today: the surface and atmosphere was reducing, while buried and water-saturated sediments were oxidizing.

Freshwater swamps differ from aquatic and marine sediments in a number of ways, including a generally high organic content and plenty of sulfates. Under these conditions, both iron and manganese may cycle rapidly between reduced and oxidized forms, further reinforcing the idea that iron may cycle in a closed or nearly-closed loop between Fe(III) and Fe(II) in wet soils. Soils that periodically dry and become oxic, such as rice paddies, may also employ molecular oxygen in this cycling, with the formation of amorphous Fe(III) oxides during the dry season, and reduction of this iron during the wet season. These reactions would restrict methane production, at least during the early part of the wet season, because Fe(III) reduction diverts electrons away from methane production.

The physical structure of iron oxides in the environment has a large effect on rates of Fe(III) reduction and populations of microbes. More strongly crystalline forms, such as goethite and hematite, are not readily reduced, while amorphous forms are consumed rapidly. Presumably it is amorphous forms that accumulate when Fe(II) is oxidized to Fe(III) and precipitates from solution as an oxide, leading to a labile pool of iron distinct from the highly resistant pool of crystalline mineral iron.

When it is not back-oxidized rapidly, Fe(II) and Mn(II) accumulate in solution, and are subject to water movements, potentially removing them from the site of production. Dissolved, reduced metal can be problematic, as these ions are readily oxidized by atmospheric oxygen if water flows contact the atmosphere, and will precipitate as rusty powder in drinking water and irrigation infrastructure. This also suggests that Fe(II) will not accumulate in natural systems over long periods, and that measuring the amount of Fe(II) in a sample is not a good indicator of iron-reducing microbial populations outside of microcosm experiments.

Iron oxides tend to provide strong adsorption surfaces for many other substances, including phosphate and heavy metals. The release to solution and movement of these substances can be a major concern when iron oxides are reduced.

Banded iron formations in some sediments and rocks, in which magnetite is deposited, appear to be the result of dissimilatory iron reduction. Magnetite is a mixed-valence iron oxide that behaves magnetically; small, characteristically-shaped crystals of it are produced intracellularly by magnetotactic bacteria, but much larger amounts are produced by a range of iron-reducing bacteria. The associated organic matter, masses and crystal structures of banded iron formations strongly suggest ancient dissimilatory iron reduction coupled to the decomposition of organic matter.

There are a range of factors controlling iron and manganese reduction in natural environments. Metal reduction is decreased in the presence of alternate electron acceptors, especially oxygen and nitrate. Oxygen is a thermodynamically favourable electron acceptor compared to either Fe(III) or Mn(IV), and nitrate appears to inhibit Fe(III) reduction by lowering electron availability below necessary levels. From a biological perspective, aerobic bacteria can usually outcompete iron reducers, many of whom are obligate anaerobes and are killed or inhibited by the presence of oxygen.

As previously described, the form of metal available in the environment also has a major effect on metal reduction rates, with more strongly crystalline forms most resistant to chemical alteration by microorganisms. “Poorly-crystalline” forms, presumably including amorphous metal oxides, are the major source of oxidized metals for reducing cells. Thus, extraction and measurement procedures that involve only the least crystalline forms of Fe(III) oxides are a good measure of the iron available for reduction by microbes. The widely-used oxalate extraction method normally does not extract significant quantities of Fe(III) oxides bound in strongly crystalline forms, except when “catalytic quantities” of Fe(II) are also available, as when weakly crystalline mixed-valence minerals such as magnetite are present. In such systems, oxalate extraction will gather a larger fraction of the total dithionate-citrate extractable iron, even though much of the iron so extracted will be in a form not actually available to soil microbes.

This paper is an enormous review of iron metabolisms in soils and sediments. It did not actually answer my current question, about the cycling of Fe(III) in soils and how comparisons of amorphous (oxalate-extractable) and total (dithionate-extractable) can be used to support inferences about long-term redox status in soils. However, I do think I now have a better grasp of general soil iron and manganese conditions and chemical transformations.