Hydrogen-Based Microbial Ecosystems in the Earth

(1) report finding hydrogen gas (H) of possible geochemical origin, and they propose that this H2 supports lithotrophic methano-genic bacteria that are physiologically active beneath the Columbia River plateau. Methanogenic bacteria are ubiquitous in the biosphere's anaerobic habitats (for example , in soils and sediments), and the ability to use H2 as an electron donor for carbon dioxide reduction to methane is almost universal among methanogens (2). In order for methanogens to be linked to pho-tosynthesis, H2 is usually produced by an anaerobic microbial food chain responsible for the decay of photosynthetically produced plant materials. But H2 production is also commonly associated with geothermal activity. Furthermore, a variety of habitats where geothermal H2 is emitted have been shown to support methanogenic bacteria (2, 3). These previously described microorganisms do precisely what was postulated for the microbial community beneath the Columbia River plateau: They grow in an-aerobic habitats at the expense of abiogenic H2. Thus, as a strictly physiological phenomenon , the subject of Stevens and Mc-Kinley's report is not unique. There are, however, three ecological aspects of the work that merit attention: (i) The proposed H2 source for methanogenic life was neither biogenic (from an anaerobic food chain) nor geothermal; (ii) C isotopic ratios suggested that methanogenesis was occurring in situ, within the basaltic subsur-face deposits; and (iii) lithotrophy (regardless of its aerobic or anaerobic basis) has not been previously reported in subsurface environments. Given the diversity of microbial biogeochemical reactions and efforts by scientists to describe them (4), it is important to place new discoveries within the scholastic context of microbial ecology. Stevens and McKinley (1) suggest that H2 generated from rock weathering supports an autotrophic microbial community in deep anaerobic basaltic aquifers in the western United States. Data about microbiological populations, hydrogen concentrations, and carbon isotopes, as well as a laboratory experiment showing H2 production from a basalt-water interaction, are provided as evidence for this conclusion. However, each of these lines of evidence has an alternative interpretation. The strongest evidence for the proposed abiological H2 production [figure 4 of the report (1)] demonstrates that H2 accumulates when some basalts are incubated in buffered water, but these experiments were conducted in phosphate buffer at pH 6, whereas the pH of the ground waters at this site is alkaline (pH = 7.5 to 9.9). It seems likely that H2 production could be favored by artificially lowering the pH, and so …