Bacterial cells were present in all samples, although absolute numbers were low (Fig. F1). The highest cell concentrations of 7.2 x 106 cells/cm3 were detected near the surface (1.4 mbsf), then populations declined rapidly over the upper 10 m, followed by a much slower rate of decrease for the rest of the core, to 7.2 x 105 cells/cm3 at 172 mbsf. The smallest population of 2.1 x 105 cells/cm3 occurred at 75 mbsf. When compared to the general equation of Parkes et al. (2000) that describes depth distributions for global deep-sediment bacterial populations, it is clear that the population profile at Site 1149 is well below the predicted distribution, and in fact, bacterial numbers are also below the lower prediction limits of this equation (Fig. F1). Whereas low bacterial populations have been observed at other ODP sites such as the eastern equatorial Pacific (EEQ) (Cragg and Kemp, 1995), this is the first time that such persistently low counts have occurred.
In sediments, the substrate for bacterial growth and maintenance is normally organic material input from the euphotic zone. At extreme water depth such as that found at Site 1149, by the time sinking organic matter has reached the sediment surface, much of the labile material has already been mineralized, resulting in a lower, more recalcitrant, organic carbon supply (J鷨gensen, 1983). This effectively restricts bacterial population size and growth rates. It is unfortunate that total organic carbon measurements were not made at Site 1149 to directly confirm this.
Another factor controlling concentrations of buried organic matter, and therefore bacterial population size, is the sedimentation rate. Deposited organic detritus that remains at the sediment surface for a longer period of time will generally be more efficiently mineralized under oxic and suboxic conditions by aerobic heterotrophic and fermentative bacteria, thereby reducing the amount of organic matter buried (Hartnett et al., 1998). At Site 1149, the average sedimentation rates were particularly low at ~18 m/m.y., however, this average is misleading as, particularly in the upper layers, these rates were boosted by volcanic ash layers. A more realistic average is 7-13 m/m.y. (Shipboard Scientific Party, 2000). This compares to sites with high-organic and high-sedimentation rates, such as the Peru margin (80 m/m.y) and to EEQ sites with low organic, low sedimentation rates (35 m/m.y.) (Suess, von Heune, et al., 1988; Mayer, Pisias, Janacek, et al., 1992). It is interesting to note that in the upper 80 m of sediment at the EEQ site (Site 851) sedimentation rates were very much lower at ~20 m/m.y. (Mayer, Pisias, Janacek, et al., 1992). This coincided with particularly low bacterial populations compared to the distribution predicted by Parkes et al. (2000), with many data that also fall below the lower prediction limit (Cragg and Kemp, 1995). Hence, the low sedimentation rates at Site 1149 probably contributed to the consistently low bacterial populations.
With the presumed low levels of recalcitrant organic carbon, rates of bacterial activity would also be expected to be low. The sulfate profile in the uppermost 5 m at this site, where the sulfate concentration falls from 30 to 28 mM, is the most rapid rate of sulfate removal, and hence sulfate reduction, at this site. Below 5 mbsf, much lower rates of sulfate reduction occur, based on the smaller decrease in IW sulfate with depth (Fig. F1). Between 5 and 170 mbsf, sulfate removal is relatively small, decreasing from 28 to 24.5 mM, and even at 407 mbsf, the sulfate concentration remains over 19 mM (Shipboard Scientific Party, 2000). The removal of only about one-third of the sulfate in the pore fluids of sediments deposited for some 130 m.y. (Shipboard Scientific Party, 2000) provides strong circumstantial evidence that the sediment column is relatively depleted in bioavailable organic matter. High concentrations of sulfate throughout the sediment column would also explain the apparently low activity of bacterial methanogenesis in these sediments (Fig. F1), where maximum methane concentrations reach only 5 ppmv (Shipboard Scientific Party, 2000). It is surprising that any methane is present in this core, as in most marine sediments sulfate concentrations >3 mM inhibit methanogenesis (Capone and Klein, 1988). Additionally, methane can be used as a substrate for sulfate reduction (Hinrichs et al., 1999; Boetius et al., 2000; Nauhaus et al., 2002). However, a similar coexistence of sulfate and methane has been observed at other low-organic carbon ODP sites (e.g., EEQ) (Cragg and Kemp, 1995). The reason for this is unclear.
Consistent with low levels of bacterial sulfate reduction and methanogenesis, the geochemical evidence indicates that other bacterial activities occur at low levels. Ammonium concentrations increase from ~20 然 near the surface to a broad maximum of ~175 然 between 20 and 70 mbsf (Fig. F1), indicating low levels of organic matter degradation. The subsequent slow decrease in ammonium between 70 and 170 mbsf may well reflect uptake during clay diagenesis rather than declining levels of bacterial activity with increasing depth (Shipboard Scientific Party, 2000). Between ~35 and 70 mbsf, bioavailable acetate shows local increases from a background of 30 然 to a maximum of 170 然 at 57 mbsf (Fig. F1). These data suggest there may be localized elevated rates of bacterial acetogenesis. Such elevated acetate concentrations are unexpectedly high compared to other low-organic carbon sediments. In the sediments of the Southern Ocean (Leg 177), acetate concentrations were in the range of 0-15 然 (Wellsbury et al., 2001), although many localized spikes (to 110 然) were present in sediments containing diatom-rich lamellae. Additionally, in the western Woodlark Basin (Leg 180), acetate concentrations were again in the range of 1-20 然 (Wellsbury et al., 2002).
Site 1149 has received considerable inputs of hydrothermal plume material in the past; thus, there are significant metalliferous components within the sediments (Shipboard Scientific Party, 2000). The two most microbiologically important metals for sediment bacteria are iron and manganese (Lovley and Chappelle, 1995). Whereas there was no evidence for iron reduction, as dissolved iron concentrations were always below detection (Shipboard Scientific Party, 2000), there was clear evidence for manganese reduction with significant increases of reduced manganese in the interstitial water (Fig. F1). The maximum Mn(II) concentration was 528 然 at 2.9 mbsf. Below this, concentrations declined to a minimum of 110 然 at 26 mbsf. However, below 26 mbsf reduced manganese concentrations exhibited a broad increase to ~100 mbsf, with a local maximum value of 403 然 at 56 mbsf. This broad subsurface dissolved manganese profile corresponds with high ammonium and acetate concentrations.
Despite the presumed low levels of recalcitrant organic matter at this site, there is clear geochemical evidence for bacterial activity at all depths with maximum activity restricted to the top ~5 m and much lower activity below. This is consistent with the high, but rapidly decreasing, bacterial populations in the top ~10 mbsf, and smaller populations below, that decrease more slowly. This type of distribution has been observed previously at several ODP sites (e.g., Amazon Fan and Santa Barbara Basin) (Cragg et al., 1996, 1995) and probably reflects the removal of the most degradable organic matter fractions followed by very slow utilization of recalcitrant organic matter. It seems improbable that any organic matter would be degradable after millions of years of burial, but Cretaceous marine sediments have been shown to support bacterial activity, including acetate formation and sulfate reduction (Krumholz et al., 1997). At Site 1149, deep manganese reduction is also occurring (Fig. F1). Manganese reduction is energetically more favorable than sulfate reduction (Madigan et al., 2000), and hence it is surprising that both soluble manganese is produced and sulfate is removed in the top ~10 m. However, this might just be an artifact of the limited depth resolution of the samples (first sample taken at 1.4 mbsf). Nevertheless, there is clear evidence of simultaneous manganese and sulfate reduction in deeper layers. This situation may reflect heterogeneous conditions in the sediment, the use of noncompetitive substrates by the different metabolic types, or cooperative metabolism between the different bacterial types. Additionally, the possibility that manganese reduction is occurring through the chemical oxidation of sulfide must be considered (Schippers and J鷨gensen, 2001, 2002). It is difficult to demonstrate that this is not occurring, as the maximum increase in Mn(II) between 25 and 100 mbsf (~300 然) would be difficult to detect as equivalent replenished sulfate; 300 然 would represent only an ~1% increase in the sulfate concentration; and sulfate reduction is simultaneously occurring. Hydrogen sulfide concentrations were not measured in Hole 1149A, although H2S is unlikely to be present in any significant quantity as these sediments are only slightly suboxic throughout (Shipboard Scientific Party, 2000). However, the continued gradual decrease in sulfate concentration over this depth range, the increase in ammonium, and the local increases in bioavailable acetate in low-temperature sediments all indicate that bacterial activity, albeit at low levels, is occurring (Fig. F1). Given that bioavailable acetate is present and that manganese reduction by acetate metabolism is considerably more favorable energetically than sulfate reduction, it seems probable that a significant amount, if not all, of manganese reduction will be by way of bacterial activity. Alternatively, if some of the manganese reduction is due to a biological reduction via H2S, this suggests greater rates of bacterial sulfate reduction than indicated by sulfate removal. Whatever the explanation, there is clear geochemical evidence for bacterial activity in the deep subsurface of this deepwater site, which is the deepest so far analyzed for sediment microbiology.