Results

Site 994

Bacterial Populations

Bacteria were present in significant numbers in all samples throughout the depth profile (Fig. 2A). Total bacterial populations were highest near the sediment surface (2.50 10⁹ cells mL^-1) and, overall, decreased with depth throughout the core. The bacterial population in the deepest sample (688.4 mbsf), 2.6 10⁶ cells mL^-1, represents a >99.9% decrease of the near-surface population. However, at two points within the depth profile, the trend was reversed and numbers of bacteria significantly increased, at 355.7 (P < 0.02) and 439.5 mbsf (P < 0.001). The greatest increase of five fold occurred at 439.5 mbsf, the sample corresponding most closely to the assumed position of the BSR (440 ± 10 mbsf). A similar trend occurred in the numbers of dividing and divided cells, which represented, on average, 9.5% of the total bacterial population at Site 994.

Populations of viable sulfate-reducing bacteria decreased rapidly from a near-surface maximum (1.1 10⁶ cells mL^-1 at 0.1 mbsf, Fig. 2A) and were not detected below 1.6 mbsf, apart from the deepest sample at 641 mbsf, where a small number of viable cells (2.0 cells mL^-1) were present. However, at several points downcore, populations of viable methane-oxidizing, sulfate-reducing bacteria were detected, with maximum numbers coinciding with a peak in headspace methane (26,000 ppmv at 32 mbsf). Viable nitrate-reducing bacteria were present in all samples, in low numbers, showing a uniform depth distribution (data not shown). Fermentative anaerobic heterotrophs were present in high numbers in all the samples, with large numbers of viable cells present near the sediment surface (2.8 10⁶ cells mL^-1 at 0.5 mbsf). A significant peak (P < 0.05) in viable fermentative heterotroph numbers was present deep in the sediment, at 566 mbsf, where numbers reached 2.6 10⁷ cells mL^-1.

Rates of bacterial growth, as measured by incorporation of methyl[³H]thymidine into DNA, decreased with depth in near-surface sediment from a maximum of 1.14 pmol [mL^-1 per day] at 0.1 mbsf (Fig. 2A) to 0.55 pmol mL^-1 per day at 1.6 mbsf. Below this depth, there was a broad, significant (P < 0.001) peak in thymidine incorporation rates down to 162 mbsf (1.40 pmol mL^-1 per day), below which rates decrease, similar to the depth profile of methane oxidation rates. Rates generally decreased with increasing depth below this, although there was a clear peak at 381.4 mbsf (1.00 pmol mL^-1 per day). Thymidine incorporation rates correlated significantly with microscopically determined numbers of dividing and divided cells at Site 994 (R = 0.945, n = 15, P < 0.002).

Potential Bacterial Activity

During collection and subsequent handling of samples, it is presently impossible to maintain in situ conditions. Although the incubations used for process-rate determinations were carried out at in situ temperatures, facilities do not exist to obtain and subsequently handle samples at in situ pressures. Thus, bacterial activity rates are described as "potential" rates, as they may differ from those in situ.

Bacterial sulfate-reduction rates were high in near-surface sediment (725 nmol mL^-1 per day at 0.1 mbsf; Fig. 2B), increasing to 842 nmol mL^-1 per day at 4 mbsf and, although subsequently decreasing with depth, rates were still high to 8.8 mbsf (269 nmol mL^-1 per day). Below this depth, sulfate-reduction rates were very low (0.8 nmol mL^-1 per day at 29.8 mbsf), although a 17.5 increase, to 14 nmol mL^-1 per day, occurred at 500.8 mbsf. The sulfate reduction rate correlated significantly with the interstitial water sulfate concentration (R = 0.84; n = 15; P < 0.002).

Rates of methanogenesis from HCO₃ were low at Site 994: near the sediment surface they were four orders of magnitude lower than sulfate-reduction rates, and generally decreased with increasing depth. A near-surface maximum was present at 4 mbsf (0.26 nmol mL^-1 per day; Fig. 2B), with further, minor, peaks at 69.3 mbsf (0.041 nmol mL^-1 per day) and a broader increase to 0.026 nmol mL^-1 per day between 456.51 and 565.91 mbsf.

Methane-oxidation rates followed a contrasting trend to both sulfate reduction and methanogenesis, increasing from <0.02 nmol mL^-1 per day in near-surface sediments to a broad peak of ~155 nmol mL^-1 per day at 29.8 and 69.3 mbsf (Fig. 2B). Rates of methane oxidation then decreased with increasing depth below this peak down to 500 mbsf, although there was an increase to 43 nmol mL^-1 per day at 381 mbsf. Below 566 mbsf, rates of methane oxidation began to increase with increasing depth.

Pore-Water and Gas Geochemistry

Interstitial water sulfate concentrations at Site 994 decreased rapidly with depth from 25.6 mM at 1.4 mbsf to 1.2 mM at 20 mbsf, consistent with the high near-surface rates of bacterial sulfate reduction (Fig. 2). Below ~60 mbsf, the sulfate profile was irregular, with low concentrations of sulfate (0-0.2 mM) detected.

Sulfide concentrations (TRIS) were low (<0.1 mM) near the sediment surface at Site 994, reflecting oxidative loss at the sediment/water interface (Fig. 2B). Sulfide concentrations increased rapidly with depth, and reached 11 mM by 30 mbsf, reflecting high rates of sulfate reduction.

Headspace methane increased rapidly from 11 to 29,000 ppmv between 9 and 46 mbsf (Fig. 2B), decreasing to ~3000 ppmv by a depth of 310 mbsf. Below 310 mbsf, headspace methane concentrations remained uniform with increasing depth. The C₁/C₂ hydrocarbon ratios indicated that the methane is microbial in origin (Paull, Matsumoto, Wallace, et al., 1996; Schoell, 1980). Concentrations of CO₂ were low near the sediment surface and generally increased with increasing depth (Fig. 2B). These increases are consistent with continued low levels of organic matter oxidation and methane oxidation within the sediments. Total organic carbon (TOC) in the sediments from Site 994 averaged 0.7% in the uppermost 160 mbsf (Paull, Matsumoto, Wallace, et al., 1996). Below 160 mbsf, organic carbon increases to a mean of 1.4%, with a maximum at 612 mbsf.

Site 995

Bacterial Populations

Near-surface bacterial populations were higher at Site 995 than at the other two sites, at 3.39 × 10⁹ cells mL^-1 (Fig. 3A). Bacteria were present in all samples, and numbers generally decreased from a surface maximum with increasing depth. However, at the BSR (446.2 mbsf), there was a significant peak (P < 0.01) where total numbers increased fivefold, from 7.0 to 7.7 log cells mL^-1. Similar trends were also evident in the numbers of dividing and divided cells, which comprised, on average, 10.8% of the total population. Again, associated with the BSR was an increase in numbers of cells involved in division, from approximately 1.3 10⁶ cells mL^-1 to 4.4 10⁶ cells mL^-1. The deepest sample, 700.8 mbsf, contained 4.6 10⁶ cells mL^-1, 0.13% of the population at the sediment surface.

Viable populations of sulfate-reducing bacteria decreased rapidly with depth from a near-surface maximum (1.1 10⁶ cells mL^-1). In several samples downcore sulfate-reducing bacteria were not detectable, but around the hydrate zone a small, but consistent population of sulfate-reducing bacteria were present, with a maximum at 439 mbsf, close to the BSR. Below this, sulfate-reducing bacteria were undetectable. Sulfate-reducing bacteria oxidizing methane as a carbon source were also found in low numbers in Site 995, with a near-surface maximum at 0.6 mbsf. Numbers of viable fermentative heterotrophs showed a relatively uniform distribution with depth, although there were distinct peaks both near surface and associated with the BSR. Nitrate-reducing bacteria were only present in low numbers at Site 995, and showed a relatively uniform distribution with depth (data not shown).

Thymidine incorporation rates were highest near the sediment surface (4.1 pmol mL^-1 per day) and decreased rapidly with depth to zero at 1.7 mbsf. Between 3.8 and 365 mbsf, low rates of thymidine incorporation occurred (0.2-0.7 pmol mL^-1 per day). In the hydrate zone growth rates increased, reaching a peak at 465 mbsf (1.3 pmol mL^-1 per day). In situ growth rates, as estimated by thymidine incorporation, correlate significantly with numbers of dividing and divided cells determined by microscopy at Site 995 (R = 0.67, n = 16, P < 0.001).

Potential Bacterial Activity

Sulfate-reduction rates were high in near-surface sediment and decreased rapidly with depth from 400 nmol mL^-1 per day at 0.1 mbsf to 0.9 nmol mL^-1 per day at 28.6 mbsf (Fig. 3B). Between 82 and 248 mbsf, no sulfate was detected in pore waters; thus the sulfate-reduction rate was zero. In deep sediment layers, there was an increase in sulfate reduction at, and below, the BSR depth, with a maximum value of 11 nmol mL^-1 per day in the deepest sample at 691 mbsf. Pore-water sulfate concentrations and numbers of viable sulfate-reducing bacteria both correlated significantly (P < 0.002 for both) with measured rates of sulfate reduction.

Rates of methanogenesis from H₂:CO₂ were low in all samples (Fig. 3B). In the uppermost 4 mbsf the rate did not exceed 0.2 nmol mL^-1 per day, and from 4 to 365 mbsf rates were below 0.04 nmol mL^-1 per day. In the deep sediment around the hydrate zone there was a clear increase in rates of methanogenesis to 2.7 nmol mL^-1 per day at 423 mbsf, 27 times the surface rate. This order of magnitude was maintained until 501 mbsf, below which the rate dropped sharply to rates of <0.02 nmol mL^-1 per day.

In contrast, rates of methanogenesis from acetate (Fig. 3B) were at least two, and up to five orders of magnitude higher than H₂:CO₂ methanogenesis. In the near surface (1.65 mbsf), the rate of acetoclastic methanogenesis peaked at 182 nmol mL^-1 per day, and then decreased sharply to 0.3 nmol mL^-1 per day at 3.8 mbsf. A secondary peak occurred at 29 mbsf (58 nmol mL^-1 per day), followed by a decrease to <1 nmol mL^-1 per day at 247 mbsf. At 423 mbsf, there was a 260-fold increase in rates of methane production from acetate, with rates of 209 nmol mL^-1 per day. Activity continued to increase through the BSR (440 ± 10 mbsf) to a maximum of 340 nmol mL^-1 per day at 465 mbsf. Below this depth, there was a decrease in the rate of acetoclastic methanogenesis, to 70 nmol mL^-1 per day at 501 mbsf. However, in contrast to H₂:CO₂ methanogenesis, rates of acetate methanogenesis increased in deeper sediments and reached a maximum of 635 nmol mL^-1 per day at 600 mbsf.

Rates of acetate turnover to CO₂ were generally high, often three orders of magnitude greater than acetate methanogenesis (Fig. 3B). Near the sediment/water interface, the rate of acetate oxidation reached 94 and 72 µmol mL^-1 per day at 0.6 and 1.7 mbsf. Below this rates decreased, remaining reasonably constant until 423 mbsf. Through the hydrate zone there was a significant increase in acetate turnover, with a peak of 190 µmol mL^-1 per day at 465 mbsf, 15 times the near-surface rate. Below the hydrate zone rates remained significant at ~64 µmol mL^-1 per day, before a further increase in the deepest sample, 691 mbsf, to 137 µmol mL^-1 per day.

Methane-oxidation rates were very low (<0.09 nmol mL^-1 per day) in the uppermost 10 mbsf at Site 995 (Fig. 3B). These rates increased by three orders of magnitude to 34 nmol mL^-1 per day by 29 mbsf, and reached a maximum of 174 nmol mL^-1 per day at 82 mbsf, coincident with high concentrations of methane (Fig. 3B). Between 82 and 423 mbsf rates remained generally low, before a peak (19 nmol mL^-1 per day) around the depth of the BSR, and a further increase in the deepest samples (to 57 nmol mL^-1 per day at 691 mbsf).

Pore-Water and Gas Geochemistry

Concentrations of pore-water acetate were low in near-surface sediment and remained low (<68 µM) to a depth of 154 mbsf (Fig. 3B). Below this, acetate concentrations begin to increase dramatically, reaching 2198 µM by 423 mbsf, and continued to increase below the hydrate zone. The maximum acetate concentration was in the deepest sample at 691 mbsf, 14,922 µM, an increase of four orders of magnitude from the sediment surface. These results were confirmed by ion-exclusion chromatography (Wellsbury et al., 1997).

Sulfate concentrations at Site 995 decreased rapidly with depth from 26.4 mM at 1.45 mbsf to 2.1 mM at 20.25 mbsf, consistent with the high near-surface rates of bacterial sulfate reduction (Fig. 3B). Below approximately 22 mbsf, the sulfate profile was irregular with low concentrations of sulfate (0-0.2 mM) detected. Concentrations of sulfides were low (<0.5 mM) near the sediment surface at Site 994, reflecting oxidative loss at the sediment/water interface (Fig. 2B). Sulfide concentrations increased rapidly with depth, and reached 8.1 mM by 28.6 mbsf, reflecting high rates of sulfate reduction.

Sediment TOC was generally below 1% in the uppermost 119 mbsf at Site 995, averaging 0.72% (Paull, Matsumoto, Wallace, et al., 1996). Deeper in the sediment, TOC increased to an average of 1.32%, with uniform depth distribution.

Headspace methane increased from 3.2 to 31,000 ppmv between 0 and 52 mbsf (Fig. 3B), and subsequently decreased to ~<5000 ppmv by 171 mbsf. Between 171 and 514 mbsf, headspace methane concentrations remained uniform with increasing depth. Between 514 and 608 mbsf, there was a zone containing two maxima (11,500 and 9000 ppmv at 533 and 563 mbsf respectively). Below 608 mbsf, headspace methane concentrations increased with depth, and reached 12,200 ppmv in the deepest sample, 699 mbsf.

The CO₂ concentrations were relatively uniform with depth (Fig. 3B), although there were peaks within the profile, consistent with deep bacterial activity. Below 650 mbsf, for example, increasing CO₂ concentrations coincided with an increase in the rate of methane oxidation.

Site 997

Bacterial Populations

Bacteria were present in all samples to a depth of 748.49 mbsf (Fig. 4A). Total bacterial populations decreased rapidly from a near-surface maximum (2.42 10⁹ cells mL^-1 at 0.025 mbsf). Substantial numbers remained even at 748 mbsf (1.8 10⁶ cells mL^-1), however, despite the 99.9% decrease in population size. Above this decreasing trend with depth there were some increases in bacterial numbers, notably at 310-331 mbsf, with a peak (2.13 10⁷ cells mL^-1) at 331 mbsf, the depth from which a "massive" methane hydrate deposit was recovered (Paull, Matsumoto, Wallace, et al., 1996). A further, significant (P < 0.002) increase in bacterial populations at 451-473 mbsf was associated with the BSR, with a 19-fold increase at 451 mbsf and an associated maximum at 469 mbsf (2.43 10⁷ cells mL^-1). In the depth interval between 338 and 443 mbsf, a zone of relatively low bacterial numbers occurred, bounded by the "massive" hydrate deposit and the BSR and free-gas zone. However, within this interval there was an increase (of one order of magnitude) at 388 mbsf, which coincides with a substantial deep subsurface peak in headspace methane (Fig. 4B). The depth distribution of dividing and divided cells was similar to that of the total bacterial population (Fig. 4A) and represented, on average, 13.1% of the total population. Clear increases in the numbers of cells involved in division were associated with both the "massive" hydrate deposit (331 mbsf) and the BSR.

Only six WRC samples were taken at Site 997, spanning the interval 331-469 mbsf. Small numbers of viable sulfate-reducing bacteria were present in all six samples, with a small increase at 444 mbsf associated with the BSR (Fig. 4B). Viable methane-oxidizing, sulfate-reducing bacteria were only detected in the 444-mbsf sample, close to the BSR, and in very low numbers (1.0 10¹ cells mL^-1). Nitrate-reducing bacteria were present in all samples, with a uniform depth distribution (data not shown). Fermentative heterotrophic bacteria were present in all samples, with the greatest numbers at 331 and 492 mbsf (2.9 10⁶ cells mL^-1) and a minimum at 451 mbsf (9.8 10⁴ cells mL^-1).

Thymidine incorporation rates were low (Fig. 4A) even compared to the same depths at the other sites, ranging from zero to 0.2 pmol mL^-1 per day. Maximal growth rates were present at 372 and 492 mbsf, with rates at the other depths at or near the detection limit.

Potential Bacterial Activity

Sulfate-reduction rates were low (<0.1 nmol mL^-1 per day) between 331 and 372 mbsf, although similar to rates at similar depths at the other sites. Below this, rates were generally an order of magnitude higher (range 0.2-1.0 nmol mL^-1 per day) and there was a clear peak at 451 mbsf (2.2 nmol mL^-1 per day) associated with the BSR. Rates of bacterial sulfate reduction correlate significantly with pore-water sulfate concentrations (R = 0.84, n = 6, P < 0.05).

Rates of H₂:CO₂ methanogenesis in the six deep samples were highest from 331 to 372 mbsf, where the maximum rate was 5.6 nmol mL^-1 per day (Fig. 4B). Below 372 mbsf, rates were relatively low (<0.5 nmol mL^-1 per day), with a minimum (0.02 nmol mL^-1 per day) at 451 mbsf.

A clear maximum rate of anaerobic bacterial methane oxidation was measured at 416 mbsf (109 nmol mL^-1 per day). In the remaining deep samples, methane oxidation rates were generally in the range 4-14 nmol mL^-1 per day, although rates again increased to 61 nmol mL^-1 per day in the deepest sample analyzed (492 mbsf).

Pore-Water and Gas Geochemistry

Interstitial water sulfate concentrations decreased rapidly within the uppermost ~20 m of sediment, from 26.8 mM at 1.45 mbsf to 1.5 mM at 21.3 mbsf (Fig. 4B). As at the two other sites, below this the sulfate concentration remained low, with small increases in deep sediment, which may reflect either contamination or a source of sulfate. Sulfide was present in all samples from Site 997 in higher concentrations than at similar depths at the other two sites, ranging from 11.5 to 20.6 mM between 331 and 491 mbsf. A maximum (20.6 mM) was detected at 444 mbsf, close to the position of the BSR.

Sediment TOC was generally below 1% in the uppermost 96 mbsf at Site 995, averaging 0.82% (Paull, Matsumoto, Wallace, et al., 1996). Deeper in the sediment, TOC increased to an average of 1.38%, with uniform depth distribution. Gas samples taken at Site 997 showed some irregularities because of air contamination during sampling (Paull, Matsumoto, Wallace, et al., 1996). Headspace methane increased from 3.2 to 56,200 ppmv between 0 and 50.4 mbsf (Fig. 4B), and subsequently decreased to ~ <5000 ppmv by 149 mbsf. Below this, the concentrations were uniform with depth, with a high degree of scatter. Concentrations of CO₂ broadly increased and then decreased with depth, with a large amount of scatter (Fig. 4B). Within the profile there was a degree of consistency with deep bacterial activity, and particularly methane oxidation (Fig. 4B), with a minimum in both headspace CO₂ and bacterial methane oxidation at 372 mbsf, and a maximum at 416 mbsf.