34S and 18O values of sulfate at the two sites showing only marginal sulfate reduction throughout the 120- to 320-m-deep sediment column.
Drilling and sampling procedures during Leg 201 were adapted to provide optimal samples for microbiology and geochemistry. Advanced piston coring was applied as deep as possible and sample recovery on deck was accomplished as quickly as possible. Subsamples for cultivation or activity experiments were taken from whole-round cores, and microbiological samples were prepared in a cold room using aseptic and anaerobic techniques. Intermittent warming of heat-sensitive microorganisms from the cold seabed was monitored and minimized by the new procedures. Contamination of subseafloor samples from the surface world was monitored continuously by the use of a dissolved tracer compound in the drilling fluid and fluorescent tracer particles of bacterial size deployed at the tip of the piston corer. The combined experience from Leg 201 showed that subsurface sediment samples can indeed be routinely obtained without significant contamination but that this requires special procedures and continuous contamination control.
Sediments recovered ranged from calcareous nannofossil oozes in the eastern equatorial Pacific to terrigenous sediments on the Peruvian shelf. At the open-ocean sites, the deepest cores reached basement at 122–420 mbsf with crustal ages reaching late Eocene. At the ocean-margin sites, the deepest cores reached 151–277 mbsf in sediments dating to the Pliocene or Miocene. Comprehensive pore water analyses were conducted for all sites drilled from the sediment surface down to the greatest depths reached. The pore water profiles of the main substrates and products of microbial processes provided information on the progressing mineralization of organic material and on the predominant biological energy metabolism. D'Hondt et al. (2004) used these data in a biogeochemical flux model to calculate the net reduction rates of nitrate, oxidized metals, and sulfate based on pore water concentrations of nitrate, manganese, iron, sulfate, and sulfide. In combination with solid-phase analyses, the data enabled mass balance calculations for the major microbial processes. The very complete chemical analyses made it possible to constrain the potential redox reactions that provide energy for the microbial populations. The geochemical data suggest that the subseafloor biosphere is principally fueled by organic carbon buried from the surface world and provide no evidence for a flux of electron donors coming up from the deep. In particular, fluid flow in the basaltic crust at the open Pacific sites shows that seawater could spread below the sediments without a strong change in concentration of even the most redox-active chemical species such as nitrate. Thus, at Sites 1225 and 1231, where microbial activity in the sediment column is the lowest, oxygen and nitrate remain in the crustal fluid and diffuse up into the overlying deposits. This demonstrates that energy-rich electron donors such as organic material or hydrogen are extremely scarce in the lowermost sediment column and in the basaltic basement. The lack of energy sources is confirmed by the sulfate profiles and 34S and 18O values of sulfate at the two sites showing only marginal sulfate reduction throughout the 120- to 320-m-deep sediment column.
All of the microbial processes identified in the ocean-margin sediments also occur in open-ocean sediments but at a highly extended depth scale. At the ocean-margin sites and the most active open-ocean site, sulfate reduction is overall the dominant mineralization pathway. Bacterial metal reduction is also important in subsurface sediments, as Fe2+ and Mn2+ profiles show active bacterial reduction at all sites. At Peru Basin Site 1231, manganese and iron reduction are even more important than sulfate reduction for the overall mineralization in the sediment column (D'Hondt et al., 2004). Sulfate reduction varies considerably from site to site. The greatest sulfate depletion occurs within the uppermost 9 mbsf at Peru Trench Site 1230, where gas hydrate and high methane flux fuel sulfate reduction. The lowest depletion occurs at Peru Basin Site 1231, where sulfate reduction is too low to measure (Böttcher et al., this volume). Surprisingly, methane is actively produced and consumed at all sites, even in open-ocean sediments where sulfate reduction is marginal. Methane distributions show that low rates of methanogenesis occur widespread in deeply buried marine sediments. The high-resolution pore water data led to the important conclusion that different types of microbial energy metabolism, such as metal reduction, sulfate reduction, and methanogenesis, often co-occur in the sediments (D'Hondt et al., 2004). It is not well understood how these competing processes can coexist for potentially millions of years. Application of radiotracers (35S and 14C) for detection of sulfate reduction or methanogenesis shows that this experimental approach is sufficiently sensitive to provide data on process rates in ocean-margin sediments but mostly falls below the detection limit in subsurface open-ocean sediments.
Leg 201 research explored to what extent modern microbial populations and processes in the deep subsurface are controlled by past oceanographic conditions and by resulting sediment properties. Expedition research found several examples of control exerted by the geological history of the sediments. At Site 1225, a distinct correspondence exists between depth distributions of dissolved iron and magnetic susceptibility. The current rate of iron reduction is thus controlled by the original amount and type of iron minerals deposited millions of years ago. At Site 1226 current manganese reduction is clearly a result of past ocean productivity and burial of manganese oxide. The most active manganese reduction today takes place from manganese oxides in an interval of unusually low organic carbon deposited between 8 and 16 m.y. ago. At the Peru shelf sites there are striking examples of processes focused in narrow subsurface zones associated with sedimentary intervals of high grain density, high natural gamma radiation (NGR), and high resistivity. These thin low-porosity intervals, rich in terrigenous sediment, were probably deposited during sea level lowstands. In particular, the lithologic context for zones of anaerobic methane oxidation at the sulfate/methane interface suggests that their orientation within the sediment is controlled by lithologic properties and thus by depositional history.
At all sites and throughout all sediment cores microbial cells are present at densities that generally range from 108 cells/cm3 at the sediment surface to 105–106 cells/cm3 at depth. At organic-poor open-ocean sites, the number of bacterial cells is lower than the global mean trend, whereas at high-productivity ocean-margin sites it is higher than the mean trend. An exceptional peak in population density of 1010 cells/cm3 was discovered at a sulfate/methane interface on the Peru margin. The available global data on microbial cell numbers in subsurface marine sediments were obtained from ODP cores that were mostly drilled in ocean margin regions. There is consequently a bias in the data set, in which the large midoceanic low-productivity regions are underrepresented. A global extrapolation from mean cell densities in all cores studied until today therefore tends to overestimate the total subseafloor cell number. Future IODP cruises should include microbiological studies of low-productivity regions in order to generate a more accurate global estimate.
Extensive cultivation efforts resulted in a large number of pure cultures of subseafloor bacteria belonging to a range of known phylogenetic lineages. The new isolates have physiological properties that for some strains are in accordance with the physical-chemical environment from which they were retrieved, for others not. Most organisms are heterotrophic anaerobes, but some are aerobes although they were isolated from deep in the anoxic sediment. Most isolates are mesophilic or moderately psychrophilic in accordance with the low-temperature in situ conditions, yet there are also several thermophilic isolates from these sediments. There are no archaea among the new isolates, although both biomarker and genetic analyses indicate their presence.
Cultivation-independent DNA- and RNA-based analyses of the microbial populations provide extensive information on their phylogenetic diversity and on key genes of sulfate-reducing and methanogenic microorganisms. The 16S rRNA clone libraries are dominated by uncultured lineages of bacteria and archaea for which very little can be concluded in terms of their basic physiology and metabolism. The distribution of such lineages in different biogeochemical zones, in which various potential redox couples are available for energy metabolism, would seem to indicate their functional role in the subsurface. Yet, the distribution of specific phylogenetic groups within zones dominated by sulfate reduction, manganese reduction, methanogenesis, or anaerobic methane oxidation does not provide a clear pattern based on our current understanding. It is, for example, surprising that 16S rRNA genes representing known groups of sulfate-reducing bacteria of -Proteobacteria or known lineages of methanogenic archaea are sparse in the clone libraries, even in sediment horizons where sulfate reduction or methanogenesis should dominate microbial processes according to the pore water chemistry. The links between phylogenetic diversity of microbial communities and their functional diversity in the subsurface environment are still poorly constrained. We need to determine the types of energy metabolism and carbon transformations of phylogenetically defined groups, such as the DSAG or MCG, for which there is currently little or no information about their life mode in the deep subsurface. As the database grows and new ideas and approaches are added, the pieces in this great puzzle will gradually fall into place. It should be noted, however, that this is not only a scientific problem related to deep subseafloor ecosystems but is also a limitation in studies of most other environments on the surface of our planet. This multidisciplinary research field is still in its early years but is progressing rapidly.
As one example of methodological progress following Leg 201, Schippers et al. (2005) were able to identify a living fraction of the total prokaryotic cells using of a highly sensitive RNA-based technique (CARD-FISH). Using the presence of cellular ribosomes as an indication of activity, the authors found that at least 1/10 to 1/3 of all cells analyzed are living bacteria. The fact that this is only a minimum estimate means that potentially all counted cells could be alive and active. One of the reasons why the question of the degree of viability is so fundamental to deep biosphere research is that the estimated mean activity and growth of the cells are extremely low. As discussed in this chapter, estimates of the potential generation time of subsurface bacteria range from less than a year to tens of thousands of years. If the mean generation times lie somewhere between tens and thousands of years, we are dealing with cellular levels of energy and carbon turnover that are many orders of magnitude below the range studied by microbiologists so far. We are therefore not able to extrapolate from current data on the maintenance metabolism or regulation of growth yield to these natural populations that may still constitute the majority of prokaryotic cells on Earth. The exploration of life at extremely low energy supply remains one of the most challenging tasks of future deep biosphere research.
