Organic geochemistry of sediments from chemosynthetic communities, Gulf of Mexico slope

Geo-Marine Letters (1994) 14:il0 119

© Springer-Verlag 1994

R. Sassen • I. R. MacDonald • A. G. Requejo • N. L. Guinasso, Jr. • M. C. Kennicutt II • S. T. Sweet • J. M. Brooks

Organic geochemistry of sediments from chemosynthetic communities, Gulf of Mexico slope

Received: 19 November 1993 / Revision received: 26 July 1994

Abstract We used a research submersible to obtain 33 sediment samples from chemosynthetic communities at 541-650 m water depths in the Green Canyon (GC) area of the Gulf of Mexico slope. Sediment samples from beneath an isolated mat of H2 S-oxidizing bacteria at GC 234 contain oil (mean = 5650 ppm) and C1-C» hydrocarbons (mean = 12,979 ppm) that are altered by bacterial oxidation. Control cores away from the mat contain lower concentrations of oil (mean = 2966 ppm) and C1-C s hydrocarbons (mean = 83.6 ppm). Bacterial oxidation of hydrocarbons depletes 02 in sediments and triggers bacterial sulfate reduction to produce the H 2S required by the mats. Sediment samples from GC 185 (Bush Hill) contain high concentrations of oil (mean = 24,775 ppm) and C1-C » hydrocarbons (mean = 11,037 ppm) that are altered by bacterial oxidation. Tube worm communities requiring H2S occur at GC 185 where the sea floor has been greatly modified since the Pleistocene by accumulation of oil, therrnogenic gas hydrates, and authigenic carbonate rock. Venting to the water column is suppressed by this sea-floor modification, enhancing bacterial activity in sediments. Sediments from an area with vesicomyid clams (GC 272) contain lower concentrations of oil altered by bacterial oxidation (mean = 1716 ppm) but C1-C5 concentrations are high (mean = 28,766 ppm). In contrast to other sampling areas, a sediment associated with the methanotrophic Seep Mytilid I (GC 233) is characterized by low concentration of oil (82 ppm) but biogenic methane (Ct) is present (8829 ppm).

R. Sassen - I. R. MacDonald • A. G. Requejo • N. L. Guinasso, Jr. - M. C. Kennicutt II • S. T. Sweet • J. M. Brooks Geochemical and Environmental Research Group (GERG), Texas A&M University, College Station, Texas 77845, USA

Introduction The main structural feature of the US gulf area is the Gulf of Mexico salt basin. The basin formed during Late Triassic rifting of the Pangea supercontinent, and then was floored by thick salt during middle Jurassic marine incursions (Salvador 1987). The structural style of the basin has been influenced by salt movement and faulting driven by rapid deposition of siliciclastic sediments during the Tertiary. The gulf continental slope, in particular, is affected by large salt thrusts (Worrall and Snelson 1989). The bulk of thermogenic hydrocarbons (oil and gas) produced from Miocene to Pleistocene reservoirs of the gulf slope offshore Louisiana are derived from deeply buried marine source rocks of Mesozoic age (Kennicutt et al. 1992a). The thermal history models of Nunn and Sassen (1986) suggest that vertical migration from Mesozoic source rocks (> 6 km depth) is geologically recent or ongoing because of rapid Tertiary burial. Overpressured fracture zones that surround moving salt diapirs and sheets, as well as active growth faults, provide efficient conduits for vertical migration from depth to shallow reservoirs and to the sea floor (Sassen et al. 1993a). Since salt and fault moyement is active, the sea floor over shallow salt displays markedly irregular topography from deformation, faulting, fracturing, and slumping. Hydrocarbon seeps and vents are concentrated in these areas (Sassen et al. 1993a). Seepage, in itself, creates characteristic sea-floor features. CO 2 from microbial oxidation of hydrocarbons precipitates as large volumes of authigenic carbonate rock that strongly affects sea-floor topography (Behrens 1988; Roberts et al. 1989, 1990a,b). Gas hydrates form as a consequence of the high pressures and low temperatures (Brooks et al. 1984, 1986), offen as mounds that actually breach the sea floor (MacDonald et al. 1994). Seepage also creates localized habitats on the sea floor that are colonized by chemoautotrophic fauna that take advantage of localized energy sources including H2S and C 1 on an otherwise barren sea floor. Chemosynthetic c o m -

111

munities of the gulf slope include bacterial mats, tube worms, clams, and mussels (Brooks et al. 1987; Kennicutt et al. 1992b). Although the makeup of chemosynthetic communities changes rapidly over short distances on the sea floor (MacDonald et al. 1989, 1990a,b), the reason for this is not well understood. Venting of hydrocarbons to the water column is thought to occur at specific sea-floor features such as mud volcanoes where elevated pressures drive rapid fluid flow (Roberts and Neurauter 1990). Imagery from the Space Shuttte indicates that natural oil slicks from sea-floor vents are common at the sea surface across the gulf slope offshore Louisiana (MacDonald et al. 1993). To place this observation in volumetric context, these authors suggest that up to 20,000 m 3 of oil vents to the deep gulf water column per year in that area alone. This study focuses on core samples of surficial sediments at four chemosynthetic communities in the Green Canyon area of the gulf slope, offshore Louisiana. Samples were acquired using a research submersible in 541-650 m water depths. The study area is near the center of a larger belt characterized by numerous hydrocarbon seeps with chemosynthetic communities, gas hydrates, and subsurface hydrocarbon accumulations (Fig. 1). Our main objectives are to: (1) document the composition of hydrocarbons in sediments from chemosynthetic communities, and (2) suggest that hydrocarbon characteristics could impact the biologic makeup and Iong-term development of chemosynthetic communities.

Experimental Samples Samples of sediments (45-cm push-cores) were acquired in 1993 at four known chemosynthetic comrnunities (GC 234, 185, 272, 233) in the Green Canyon area of the gulf slope by the Johnson Sea-Link (JSL) research submersible (Fig. 1). Samples were transferred to cans after collection, sodium azide bactericide was added, and the cans were then sealed under N 2. Samples were frozen until analysis. Samples GC 234-1 through -12 were acquired during JSL dive 3525 (27°44.77'N, 91°13.34'W) at a water depth of about 543 m. Small-diameter core tubes (2 cm ID) were used to collect densely spaced samples within a 1-m a grid straddling an isolated orange and white Be99iatoa mat (Fig. 2). Sediments from beneath mats were dark-colored and smelled of H2S. Be99iatoa are H2S-oxidizing bacteria that are widely distributed at seep localities (Sassen et al. 1993b). The larger GC 234 locality is described by MacDonald et al. (1990a). Samples GC 185-1 through -12 were acquired during JSL dive 3529 at Bush Hill (27°46.92'N, 91°30.46'W). Small-diameter core tubes (2 cm ID) were used to collect sediment samples within a 1-m 2 area beneath part of a Beggiatoa mar adjacent to a vestimentiferan tube worin cluster. Sediments from beneath mats were dark-colored and smelled ofH2S. Water depth at the sample site is about 541 m. The tube worms (Lamellibrachia n. sp.) contain bacterial symbionts that oxidize H2S (MacDonald et al. 1989). Sample GC 185-13 was acquired from a gas hydrate that breached the sea floor nearby on Bush Hill (MacDonald et al. 1994). No other sample of "outcropping" hydrate has been analyzed previously. Samples GC 272-1 through -7 were acquired using

Fig. 1 Map showing Green Canyon study areas (inset) in context of reported gulf slope hydrocarbon seeps with chemosynthetic communities, gas hydrates, and subsurface hydrocarbon accumulations

98ow

96°W

94°W

92°W

88oW

90ow

86oW

~ GC 185

30°N -

GC 272

3

2

84ow

O

82ow

N

GC 233

B

GC 234 30°N

ù ù.

\,

\

O ~" _«/

"-\ \,

\

"'ù

Q

',

.-'-' •

'. " . "'-.

28°N -

28°N

200 m

1000 m

--"-" "........,':,.' "-', ."-"-:, ':"

:

26°N

;

i

~ ,' '

0

.-

i

2000 "m ,,'"- - -"--""" "" - -" "'"" -'" "" 3000 m

- -,,,,:,......

/'~" /

~,'~-,.~ ,./

.-'""

-"......

Seeps/Chemosynthetic

"..........................

0

Gas Hydrates

26°N

• Oil & Gas Accumulations 2TO

4OOrkm

112 Bacterial Mat:

I [][] Orange Mat I White Mar

i

i

10cm

O

O

0

Core Locations:

I (~ • Mat Control

O

Fig. 2 Detai!ed locations of core samples taken across a zoned orange and white Beggiatoa mat at the GC 234 locality

large-diameter core tubes (6.5 cm ID) during JSL dive 3535 in an area (27°41.28'N, 91°32.45'W) characterized by darkcolored, H2S-rich sediments with epifaunal vesicomyid clams. Water depth at the site is about 575 m. The clams contain bacterial symbionts that oxidize H a S (MacDonald et al. 1990a). Sample G C 233-1 was acquired using a large-diameter core tube (6.5 cm ID) during JSL dive 3539 (27°43.43'N, 91°16.78'W) adjacent to an anoxic brine pool rimmed with methanotrophic mussels (Seep Mytilid I) at a water depth of about 650 m. The brine pool site is described by M a c D o n a l d et al. (1990b).

Methods We have relied on gas chromatography to provide preliminary insight to oll and C1-C5 hydrocarbons of sediments in this reconnaissance study. Methods used for geochemical analysis are similar to those described by Kennicutt et al. (1988). Dried and powdered sediment samples were Soxhlet extracted with hexane for 24 h with internal standards. After reduction in volume, extracts were subjected to C15 + gas chromatography using a Hewlett-Packard 5880 chromatograph equipped with a 30-m DB-5 capillary column. This approach yields the total hydrocarbon concentration in sediments. Chromatograms of recent sediments that do not contain oil typically consist of n-alkanes with strong odd-numbered carbon predominance, and a flat baseline because the hydrocarbons that characterize oll tend to be absent or in low abundance. If oll is present, chromatograms typically show an elevated baseline. The area beneath the elevated baseline is called the unresolved complex mixture (UCM) because naphthenes and other compounds are present that a r e difficult to resolve by gas chromatography. The n-alkanes and isoprenoids normally prominent in unaltered oils are absent or in low abundance in our samples because these com-

pounds have been destroyed by microbial oxidation. Moreover, oil is present in such great concentration as to generally overprint any higher-plant waxes coextracted from sediments. For these reasons, no data on n-alkanes or isoprenoids are presented. The concentrations of U C M are expressed as patts per million by sediment weight. We analyzed the free C1-C» hydrocarbons in the headspace of canned samples to define hydrocarbon types, and because they are sensitive indicators of bacterial oxidation (James and Burns 1984; Sassen et al. 1988). After shaking canned samples at 50°C for 12 hr, the C1-C» hydrocarbon gases in the can headspace were separated using a HewlettPackard 5710 chromatograph equipped with a 2-m packed alumina column. A standard gas mixture was used for calibration. Concentrations of individual C~-C» hydro-

Table 1 Geochemical data on sediment samples from Green Canyon 234 bacterial mat and control sites, Green Canyon 185 bacterial mat and tube worm sites (Bush Hill), Green Canyon 272 vesicomyid clam sites, and the Green Canyon 233 brine pool with Seep Mytilid I" Gas

Sample Green Canyon 234 GC 234-1 GC 234-2 GC 234-3b GC 234-4b GC 234-5 GC 234-6 GC 234-7 GC 234-8b GC 234-9b GC 234-10b GC 234-11b GC 234-12 Green Canyon 185 GC 185-1 GC 185-2 GC 185-3 GC 185-4 GC 185-5 GE 185-6 GC 185-7 GC 185-8 GC 185-9 GC 185-10 GC 185-11 GC 185-12 GC 185-13c Green Canyon 272 GC 272-1 GC 272-2 GC 272-3 GC 272-4 GC 272-5 GC 272-6 GC 272-7 Green Canyon 233 GC 233-1

UCM 2,380 2,684 ND 3,928 2,226 3,393 4,513 7,809 ND 4,168 6,693 2,597

Total C1-C 5 wetness

i-C4/n-C 4

11.5 20.1 169.6 104.1 45.9 165.2 128.7 1,257.9 97.6 23,316.1 52,925.9 130.1

11.7 10.8 9.8 14.7 2.7 1.7 2.6 2.9 3.0 0.1 0.4 2.6

0.52 0.51 0.52 0.58 0.56 1.77 0.66 0.34 0.48 0.59 37.49 2.04

35,209 33,035.2 38,992 300.8 23,519 363,4 18,702 2,486.1 35,938 6,780.2 21,667 7,180.7 25,802 10,643.3 21,587 12,602.4 23,145 31,136.8 14,849 8,723.9 20,312 18,057.6 17,583 1,138.1 10,311 153,566.0

1.8 15.1 26.5 1.5 69.4 0.5 0.8 1.9 3.4 1.0 7.0 17.1 73.3

39.02 NC 15.32 ND 585.62 NC 37.14 1.14 55.86 2.92 121.75 92.50 0.18

1,060 1,464 3,294 3,312 1,089 328 1,463

47,436.9 26,398.1 39,682.7 38,951.3 23,986.8 7,426.7 17,476.5

8.5 0.2 0.1 0.3 0.3 5.6 0.3

0.20 8.47 8.51 30.88 5.53 3.76 5.25

82

8,834.2

0.1

1.88

« Gas wetness = ( C 2 - C 4 / C 1 - C4) x 100. Note that n = normal, i = iso, and NC = not calculated. b Mat samples Gas hydrate

113

carbons are expressed as parts per million by sediment volume.

Results

Results of analysis of 33 sediment samples from the four chemosynthetic communities sampled in Green Canyon are shown in Table 1 (UCM, total C I - C » gas ratios) and Table 2 ( C I - C » compositions). Isolated bacterial mat at GC 234 The C 1s+ chromatograms of hexane extracts from G C 234 show a broad U C M characterized by the absence of

n-alkanes and isoprenoids (Fig. 3). UCM concentrations (Table 1) are in the 2226-7809 ppm range (mean 4039 ppm). Concentrations are highest beneath the mat (mean 5650 ppm), and decrease away from the mat (mean 2966 ppm). Total concentrations of C1-C5 hydrocarbons (Table 1) are in the 11.5-52,926 ppm fange (mean 6531 ppm). The total C1-C5 concentrations are rauch higher beneath the mat (mean 12,979 ppm), whereas values drop away from the mat (mean 84 ppm). Gas compositions are dominated by CI, as indicated by gas wetnesses in the 0.1-11.7~o fange. Samples from beneath the orange mat with the highest C~-C5 concentrations show the lowest gas wetness values. The i-C4/n-C 4 ratlos generally vary in the 0.3-2.0 range, with the exception of a single orange mat sample (37.5). The C» hydrocarbons are absent or below detection limits (Table 2).

Table 2 Concentrations of individual C1-C» hydrocarbons in sediment samples from Green Canyon 234 bacterial mat and control sites, Green Canyon 185 bacterial mat and tube worm sites (Bush Hill), Green Canyon 272 vesicomyid clam sites, and the Green Canyon 233 brine pool with Seep Mytilid P Sample

Methane

Ethane

Ethylene

Propane

Propylene

i-Butane

n-Butane

i-Pentane

n-Pentane

Green Canyon 234 GC GC GC GC GC GC GC GC GC GC GC GC

234-1 234-2 234-3 b 234-4 b 234-5 234-6 234-7 234-8 b 234-9 b 234-10 b 234-11 b 234-12 Green Canyon 185 G C 185-1 G C 185-2 G C 185-3 G C 185-4 G C 185-5 G C 185-6 G C 185-7 G C 185-8 G C 185-9 G C 185-10 G C 185-11 G C 185-12 G C 185-13 ° Green Canyon 272 G C 272-1 G C 272-2 G C 272-3 G C 272-4 G C 272-5 G C 272-6 G C 272-7 Green Canyon 233 G C 233-1

10.1 17.7 149.6 82.7 44.5 162.3 125.1 1,218.7 92.9 23,289.5 52,702.7 124.4

0.5 0.8 6.7 5.4 0.9 2.5 2.7 19.5 1.6 23.7 187.7 1.8

0.1 0.2 2.6 1.7 0.1 0.1 0.2 1.9 1.1 0.6 0.4 1.4

0.4 0.6 4.3 3.7 0.2 0.2 0.3 6.1 0,7 1.0 7.2 0.6

0.0 0.1 1.0 5.4 0.1 0.0 0.1 0.9 0.7 0.3 0.2 0.9

0.2 0.2 1.8 1.9 0.0 0.0 0.i 2.8 0.2 0.4 26.9 0.7

0.3 0.5 3.5 3.3 0.1 0.0 0.2 8.1 0.4 0.6 0.7 0.3

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.Õ

31,504.9 225.0 266.9 2,449.0 485.6 7,143.5 10,554.2 12,362.9 28,783.5 8,631.9 14,631.6 942.6 24,132.9

170.0 8.9 43.8 13.7 104.5 34.6 59.7 97.9 307.4 81.3 198.4 92.7 13,011.4

0.2 0.2 0.1 0.3 0.6 0.3 0.3 0.0 0.2 0.6 0.6 0.5 0.0

157.2 10.1 26.7 4.0 74.7 2.3 8.5 36.7 223.8 8.0 173.6 45.0 15,634.3

0.1 35.7 0.0 0.1 0.3 0.0 0.1 0.0 0.0 0.2 0.4 0.2 0.0

235.1 20.8 24.3 18.9 920.5 0.0 19.9 56.0 460.9 1.4 715.6 56.6 6,210.0

6.0 0.0 1.6 0.0 1.6 0.0 0.5 48.9 8.3 0.5 5.9 0.6 35,074.3

961.6 0.0 0.0 0.0 5,192.4 0.0 0.0 0.0 1,352.7 0.0 2,331.7 0.0 57,870.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1,632.9

40,594.9 26,347.0 39,626.4 38,846.6 23,909.4 7,013.2 17,424.3

2,252.5 44.8 49.1 78.3 68.1 105.8 46.5

0.0 0.2 0.1 0.1 0.2 0.6 0.2

607.4 3.0 4.2 6.4 5.1 149.3 3.2

0.2 0.2 0.1 0.2 0.2 0.4 0.2

151.9 2.6 2.5 19.0 3.3 124.4 1.7

759.3 0.3 0.3 0.6 0.6 33.1 0.3

2,218.7 0.0 0.0 0.0 0.0 0.0 0.0

852.1 0.0 0.0 0.0 0.0 0.0 0.0

0.1

0.2

0.0

0.0

0.0

0.0

0.0

8,829.1

a Note that n = normal, i = iso

b Mat samples Gas hydrate

4.7

114 Fig. 3 Examplesof C15+ chromatograms of hexane extracts from the GC 234, GC 185, GC 272, and GC 233 study areas. Sample numbers are shown with each chromatogram. All chromatograms from these areas show a UCM feature, and the absence of the n-alkanes and isoprenoids initially present in oil

GC 234-4

GC 185-4

GC 272-2

GC 233-1

Bacterial mat and tube worms at GC 185 (Bush Hill) C~5+ chromatograms from samples GC 185-1 through -12 show a broad UCM characterized by the absence of n-alkanes and isoprenoids (Fig. 3). The UCM concentrations (Table 1) are in the 17,583-38,992 ppm range (mean 24,775 ppm). Total concentrations of C1-C 5 hydrocarbons (Table 2) are in the 301-33,035 ppm range (mean 11,037 ppm). Gas compositions are generally dominated by C1, but there are exceptions. The range of gas wetness is wide (0.5-69.4yo). In three samples, n-Co is absent or below detection limits. Where i-C4/n-C 4 ratios are calculable in nine samples, they are in the wide 1.1-586 range. Although n-C5 is absent or below detection limits in these samples, i-C» is a significant component in four samples (961-5192 ppm).

i-C4/r/-C 4 ratios vary in the 0.2-30.9 range. The C5 hydrocarbons are either absent or below detection limits in all but one sample (Table 2).

Chemosynthetic mussels at GC 233 The C15 + chromatogram of sediment from the brine pool (Seep Mytilid I) shows only UCM with slight overprinting from coextraction of higher-plant waxes from recent sediment (Fig. 3). The UCM concentration is only 82 ppm (Table 1). The total concentration of C t-C5 hydrocarbons is 4903 ppm. C~ dominates, as indicated by a gas wetness of about 0.1~. The C4 and C5 hydrocarbons are either absent or below detection limits (Table 2).

Gas hydrate at GC 185 (Bush Hill) Discussion

Total C~-C» concentration of the hydrate sample (GC 185-13) is 153,567 ppm (Table 1). The C2-C4 hydrocarbons are major components of the hydrate, as reflected by a gas wetness of 73.3Yo. The i-C4/n-C 4 ratio is 0.18 and the i-C»/n-C5 ratio is 35.4. In comparison to other samples, i-C» and n-C5 are significant components (Table 2).

Chemosynthetic clams at GC 272 C15+ chromatograms of samples from an area characterized by vesicomyid clams show only UCM (Fig. 3). The UCM concentrations (Table 1) are in the 328-3312 ppm range (mean 1716 ppm). Total concentrations of C1-C» hydrocarbons are in the 7427-47,437 ppm range (mean 28,766 ppm). Gas compositions are dominated by C 1, as indicated by gas wetnesses in the 0.1-8.5~ range. The

Hydrocarbons The most basic observation concerning the Beggiatoa mat at the GC 234 sampling locality is that mean concentrations of UCM (Fig. 4) and C1-C 5 hydrocarbons (Fig. 5) are higher beneath the mat than in control cores oft the mat. It should be stressed that the concentrations of C1-C» hydrocarbons beneath the mat are an order of magnitude higher than in control samples only 10 cm from mat margins (Table 1 and Fig. 2). The result is consistent with the tight coupling between mats and hydrocarbon gradients suggested by Sassen et al. (1993b). The effects of aerobic microbial oxidation can be recognized since the prominent n-alkanes and the branchedchain isoprenoids originally present in crude oil are preferentially degraded, leaving a residue of UCM (Sassen et al.

115 1993b). This sequence is illustrated in Fig. 6 by comparison of an unaltered oil from a subsurface reservoir and an degraded seep extract. The C15+ chromatograms of extracts from the GC 234 sampling area are altered by bacterial oxidation, as shown by the example in Fig. 3. =0>,3 The i-C4/n-C4 ratios of low-molecular-weight hydrocarbons can provide a sensitive index of the degree of bacterial oxidation (Sassen et al. 1988). Since the straightchain hydrocarbon n-C 4 is preferentially oxidized by bacteria, the branched-chain hydrocarbon i-C4 tends to beControl come more significant as this process advances (Winters and Williams 1969). The lowest gas wetness values and the single highest i-C4/n-C 4 ratio (37.5) come from samples Q acquired directly beneath the mat (Table 1). Bacterial oxiO~ UCM (ppm) dation effects (as indicated by the C1-C» hydrocarbons) thus appear more advanced beneath the mar than away Fig. 4 Histogram illustrating that the concentrations of U C M in from the mat. core samples of the G C 234 Beggiatoa mat are higher beneath the Hydrocarbons set into motion complex bacterial intermat than in control cores away from the mat actions within sediments of hydrocarbon seeps (Sassen et al. 1993b, and references therein). Hydrocarbon-oxidizing bacteria produce CO2, some of which is likely to precipirate as carbonate rock. When O z is depleted, sulfatereducing bacteria become active and produce H2S. Beg3L 9iatoa mats occupy the interface between anoxic sediments and the oxic water column. The mats oxidize H2 S ~2. to first form granules of elemental sulfur wJthin cells and ultimately oxidize the elemental sulfur to form sulfate. The Beggiatoa mat adjacent to the tube worm cluster at GC 185 and the mar on GC 234 are similar in that thermogenic hydrocarbons altered by bacterial oxidation ntrol are present at both localities. However, the mean conceno °° i tration of U C M at GC 185 is higher (24,775 ppm) than at GC 234 (4039 ppm) (Fig. 7). Mean concentrations of C t C» hydrocarbons at GC 185 are also higher (11,037 ppm) CD than at G C 234 (653 l ppm) (Fig. 8). Total C1-Cs (ppm) Bacterial oxidation decreases gas wetness by degrading the thermogenic Cz-C4 hydrocarbons, a process accentuFig.5 Histogram illustrating that the concentrations of total C1-C 5 hydrocarbons in core samples of the GC 234 Beggiatoa mat are ated in natural systems by addition of biogenic C 1 (James higher beneath the mat than in control cores away from the mat and Burns 1984; Sassen et al. 1988). GC 185 samples show a much wider range of gas wetnesses (0.5-69.4%) than noted at G C 234 (0.1-14.7%). Unusually high gas wetness Fig. 6 Comparison of C15+ chromatograms of a subsurface reser- in some samples argues that another process affects results voir sample of oil from Jolliet Field in GC 184 and a seep sample from the GC 185 area. from GC 185 (Bush Hill) illustrates the effects of bacterial oxidation The i-C4/n-C« ratios from G C 185 samples are quite

Jolliet Field Reservoir Sample

~

BACTI~LOXIDA'I'lON> ù,,,,,

B«

~

- o

GC 185-2 Seep Sample

{I , ,f

~ ......

116

4 IJ. 1-

185

0.

UCM (ppm)

'~

Fig. 7 Histogram illustrating differences in UCM concentrations between the four study areas. Note the relatively high UCM concentrations measured in sarnples from GC 185 (Bush Hill)

!

0>,7-

t"- 6 - ~ 5"

ü. a2 1'

G

d

(Ripmeester and Ratcliffe, 1991). Thus, there is a straightforward explanation for out data on GC 185 sedirnents. Since the GC 185 area contains abundant hydrate (Sassen et al. 1993a), it is likely that a number of sedirnent samples contain srnall arnounts of this type of hydrate. The C15+ chromatograrns of sedirnent extracts frorn a vesicomyid clarn environrnent on GC 272 show that oil has been affected by bacterial oxidation (Fig. 3). The concentration of UCM is lower (rnean 1716 pprn) than noted at GC 234 or GC 185 (Fig. 7). Ct-C5 concentrations (mean 28,766 ppm), however, are higher than at other sites (Fig. 8). C1 dorninates the compositions. The i-Co/n-C4 ratios tend to be lower than noted in sarnples adjacent to tube worin clusters at GC 185, suggesting that bacterial oxidation effects are less advanced. The sample of sedirnent frorn the anoxic brine pool on GC 233 colonized by the Seep Mytilid I is characterized by low UCM (82 pprn). The dominant lower-rnolecularweight hydrocarbon is C1 (8829 pprn). The low gas wetness (0.1%), and the absence of C4 and Cs hydrocarbons in our sediment sarnple suggest a dorninantly biogenic origin. This is consistent with the carbon isotopic composition of C1 (-63.8%o PDB) reported at the site (MacDonald et al. 1990b). Therrnogenic hydrocarbons äppear to be less significant than at other sites. Hydrocarbon cornpositions at the chemosynthetic cornmunities thus appear to be influenced by multiple processes including: (i) gross differences in rates of migration of therrnogenic hydrocarbons to sedirnents, (2) differing levels of bacterial oxidation, and (3) specific effects related to hydrate formation. We emphasize that large differences in hydrocärbon characteristics rnay occur within spatial scales ofless than 1 rn, helping to explain the great heterogeniety orten observed between and within chernosynthetic cornrnunities.

Total C1-Cs (ppm) Seeps and vents Fig. 8 Histogram illustrating differences in total C1-C» hydrocarbon cencentrations between the four study areas

variable (1.1-586). Some i-C4/n-C « ratios frorn GC 185 sarnples are high enough to irnply more advanced bacterial oxidation effects than at the GC 234 area, where values are lower (0.3-37.5). However, unusually low i-C4/n-C,~ ratios in a few GC 185 sarnples suggest that a process other than rnicrobial oxidation affects our results. Relatively high concentrations of the i-C5 hydrocarbon in four GC 185 sarnples also need to be explained. A tore sarnple of gas hydrate that breached the sea floor (GC 185-13) was obtained near out other Bush Hill sarnple site. The abundant C2-C » hydrocarbons (gas wetness = 74.3%) eontrast with previously analyzed hydrate frorn the gulf slope in which C1 is the dominant component (Brooks et al. 1984, 1986). The i-C4/n-C 4 ratio of the hydrate is quite low (0.18), whereas the i-C»/n-C» ratio is high (35.4). It should be emphasized that the i-C s hydrocarbon is strongly favored as guest rnolecules in structure H hydrate

Differences between seeps and vents need to be emphasized. Our data show that cornplex ehemosynthetic comrnunities are associated with sea-floor seeps where high concentrations of oil and gas altered by bacterial oxidation occur in sedirnents. In contrast, chernoautotrophic fauna other than bacterial rnats are not observed near a rnud volcano on GC 143 where oil and gas episodically vent to the water column (Roberts and Neurauter, 1990). Since abundant hydrocarbons are present at both seeps and vents, it is surprising that we do not see cornplex cornmunities in both situations. A C1» + chromatograrn of oil-stained sediments frorn the active GC 143 mud volcano is shown in Figure 9. The chromatograrn has unusual features including: (1) a strongly elevated baseline (UCM) consistent with intense bacterial oxidation of oil, and (2) an envelope of n-alkanes and isoprenoids frorn oil that has not been signifieantly impacted by bacterial oxidation. At present, bacterial oxidation appears unable to keep pace with the high rates of hydrocarbon venting. This pattern is consistent with a

117

n..C17

i Fig. 9 A C~s+ chromatogramof sedimentsfrom a mud volcanoon GC 143. Unusual features include: (1) a strongly elevated baseline (UCM) consistentwith the effectsof bacterial oxidationthat is overprinted by (2) the n-alkanes and isoprenoids characteristic of unaltered oll from recent venting

history of episodic venting of fresh oil followed by bacterial hydrocarbon oxidation.

Sea-floor rnodification at seeps Understanding why oil composition at sea-floor seeps differs frorn that at vents could shed new light on the early development of complex chernosynthetic cornrnunities, particularly at GC 185 (Bush Hill). Frorn the standpoint of the H2 S requirements of chemosynthetic cornrnunities, venting is negative since the hydrocarbons needed to promote bacterial processes in sedirnents are largely lost to the water column. Unless hydrocarbons are retained in sedirnents, hydrocarbon-oxidizing bacteria will not deplete O2, and bacterial sulfate reduction will be inhibited. Beggiatoa rnats appear to be so widely distributed because they are well adapted to rapid colonization of the many srnall or transient seep sites that provide energy and carbon sources on the gulf slope (Sassen et al. 1993b). Our data show a clear link to concentration gradients of therrnogenic hydrocarbons, but mats also occur at seeps of purely biogenic C1 where UCM is present in low abundance (Sassen et al. 1993b). Bacterial oxidation of any hydrocarbon type depletes 0 2 in sedirnents, perrnitting the bacterial sulfate reduction thät provides the mats with H2S. We suggest that the adventitious Beggiatoa rnats (and other bacteria) play a role in starting a process of sea-floor rnodification. Beggiatoa rnats are thought to retard hydrocarbon loss to the water column, a function that could enhance production of the Hz S needed by the mats (Sassen et al. 1993a). This "biologie barrier" at the interface between sedirnents and the water column could also enhance precipitation of authigenic carbonate rock (Sassen et al. 1993b) to further seal the sea floor. Tube worrns, in contrast, appear to occupy environments that already have been modified by thermogenic seepage effects over a long span of tirne. The GC 185 (Bush

Hill) site, for example, is associated with active faults that serve as conduits for vertical rnigration of oil and associated gas frorn subsurface reservoirs ofnearby Jolliet Field (Sassen et al. 1993a). Since subsurface reservoirs of Jolliet Field were charged with oil and gas during the Pleistocene, hydrocarbons probably began to irnpact the sea floor at that time (Sassen et al. 1993a). The abundant authigenic carbonate rock that accurnulated at our tube worin site (MacDonald et al. 1989), as well as over shallow salt and faults north of Jolliet Field (Roberts et al. 1989), provides conclusive evidence of seepage over a long span of time. The highest UCM concentrations rneasured in the present study are adjacent to a vestirnentiferan tube worrn cluster on Bush Hill (GC 185). An empirical association of tube worrns with oily sedirnents at GC 185 has been documented previously (MacDonald et al. 1989). Based on our new data, however, we can move further towards understanding why this relationship exists. Both rnigration and bacterial oxidation of hydrocarbons appear rapid at this site. C15+ chromatograrns only show a UCM residue because the n-alkanes and isoprenoids have been destroyed by bacterial oxidation. The C~-C s hydrocarbons frorn this site are impacted by bacterial oxidation. We suggest that the vestirnentiferan tube worms are so abundant at GC 185 because biologic barriers and relatively irnperrneable seals of authigenic carbonate rock and gas hydrate appear to retain hydrocarbons within sedirnents. Moreover, hydrate bodies could serve as buffers to rnaintain hydrocarbons in sediments for bacterial activity should hydrocarbon migration rates fluctuate over time. The intense bacterial oxidation of abundant hydrocarbons held in sedirnents at GC 185 triggers the rapid production of the H2S required by the tube worrns, favoring their proliferation. Another requirement of tube worrns could help explain why they need a rnore specialized habitat than, for exarnple, the adventitious Be9giatoa rnats. Tube worms appear to need a hard substraturn as a point of attachment (H. H. Roberts personal comrnunication), and authigenic carbonate rock fulfills that need in an otherwise unfavorable mud-dorninated environrnent. A long span of tirne may be required to precipitate enough carbonate rock to serve the large populations of tube worms seen at GC 185. Results frorn the GC 272 site rnove us closer to understanding the relationship of hydrocarbons to the vesicomyid clarn environment. The C1»+ chrornatograms show evidence of bacterial oxidation, but UCM concentrations are low (1716 ppm) in cornparison to GC 185 or GC 234. The Ca-C» hydrocarbon concentrations are high (mean 28,766 pprn), but show low gas wetness values consistent with both bacterial oxidation and strong dilution by biogenic Ca (Table 1). It eould be that the clams persist in a relict oil seep environrnent where the rate of oil seepage has declined frorn higher levels in the past, or oxidation of gas rather than oil triggers H2S production. The cornposition of hydrocarbons sarnpled in association with the Seep Mytilid I (GC 233) is distinct since these organisms are rnethanotrophic. UCM is not present in significant concentration in sedirnents, and the main type

118

of low-molecular-weight hydrocarbon is biogenic C~. The Seep Mytilid I, however, is known to utilize either biogenic or thermogenic C~, as available (MacDonald et al. 1990a).

Conclusions I. Sediment samples from beneath an isolated mat of H 2 •oxidizing bacteria at GC 234 contain oil (mean 5650 ppm) and C1-C » hydrocarbons (mean 12,979 ppm) that are altered by bacterial oxidation. Control cores away from the mat contain lower concentrations of oil (mean 2966 ppm) and C1-C» hydroearbons (mean 84 ppm). Bacterial oxidation of hydrocarbons depletes O » triggering bacterial sulfate reduction to produce the HaS needed for mat development. Moreover, CO2 from the bacterial hydrocarbon oxidation is likely to precipitate as authigenic carbonate rock, beginning a process that modifies the sea floor. 2. Sediments at GC 185 (Bush Hill) contain high concentrations ofUCM (mean 24,775 ppm) and C1-C~ hydrocarbons (mean 11,037 ppm). Tube worin communities requiring HzS occur at GC 185 where the sea floor has already been greatly modified since the Pleistocene by accumulation of oil, gas hydrate, and authigenic carbonate rock in sediments. Venting is thus suppressed over time, favoring slow development ofcomplex communities that depend on abundant hydrocarbons in sediments. 3. Sediment samples from an area with vesicomyid clams (GC 272) contain lower UCM (mean 1716 ppm) but C1-C 5 concentrations are high (mean 28,766 ppm). The clams appear to persist at relict therrnogenic seeps where the rate of seepage has declined from past highs or H2S production is triggered by gas instead of oil. 4. The sediment we sampled in association with the methanotrophic Seep Mytilid I is characterized by low UCM (82 ppm) and by dominance of biogenic C1 (8829 ppm). Acknowledgments Support for research was provided by the NOAA National Undersea Research Center, University of North Carolina at Wilmington, and by the Minerals Management Service. Additional support from BP Exploration and Conoco is acknowledged. We thank H. H. Roberts of LSU for a mud volcano sample and Tim Collett of the USGS for helpful comments on hydrate composition.

References Behrens EW (1988) Geology ofa continental slope oil seep. American Association of Petroleum Geologists Bulletin 72:105 114 Brooks JM, Kennicutt MC, Fay RR, McDonald TJ, and Sassen R (1984) Thermogenic gas hydrates in the Gulf of Mexico. Science 225:409-411 Brooks JM, Cox HB, Bryant WR, Kennicutt MC, Mann RG, and McDonald TJ (1986) Association of gas hydrates and oil seepage in the Gulf of Mexico. Organic Geochemistry 10:221-234

Brooks JM, Kennicutt MC, Fisher CR, Macko SA, Cole K, Childress JJ, Bidigare RR, and Vetter RD (1987) Deep-sea hydrocarbon seep communities: Evidence for energy and nutritional carbon sources. Science 238:1138-1142 James AT and Burns BJ (1984) Microbial alteration of subsurface gas accumulations. AAPG Bulletin 68:957-960 Kennicutt MC, Brooks JM, and Denoux GJ (1988) Leakage of deep, reservoired petroleum to the near surface of the Gulf of Mexico continental slope. Marine Chemistry 24 : 39-59 Kennicutt MC, McDonald T J, Comet PA, Denoux G J, and Brooks JM (1992a) The origins of petroleum in the northern Gulf of Mexico. Geochimica et Cosmochimica Acta 56:1256-1280 Kennicutt MC, Burke RA, MacDonald IR, Brooks JM, Denoux GJ, and Macko SA (1992b) Stable isotope partitioning in seep and vent organisms: chemical and ecological significance. Chemical Geology 101:293-310 MacDonald IR, Boland GS, Baker JS, Brooks JM, Kennicutt MC, and Bidigare RR (1989) Gulf of Mexico hydrocarbon seep communities Il. Spatial distribution of seep organisms and hydrocarbons at Bush Hill. Marine Biology 101:235--247 MacDonald IR, Guinasso NL, Reilly JF, Brooks JM, Callender WR, and Gabrielle SG (1990a) Gulf of Mexico hydrocarbon seep communities: VI. Patterns in community structure and habitat. GeoMarine Letters 10:244-252 MacDonald IR, Reilly JF, Guinasso NL, Broõks JM, Carney RS, Bryant WA, and Bright TJ (1990b) Chemosynthetic mussels at a brine-filled pockmark in the northern Gulf of Mexico. Science 248:1096-1099 MacDonald IR, Guinasso NL, Ackleson SG, Amos JF, Duckworth R, Sassen R, and Brooks JM (1993) Natural oil slicks in the Gulf of Mexico are visible from space. Journal of Geophysical Research 98(C9): 16351-16364 MacDonald IR, Guinasso NL, Sassen R, Brooks JM, and Lee L (1994) Gas hydrate that breaches the sea floor on the continental slope of the Gulf of Mexico. Geology, in press Nunn JA and Sassen R (1986) The framework of hydrocarbon generation and migration, Gulf of Mexico continental slope. Transactions, Gulf Coast Association of Geological Societies 36:257-262 Ripmeester JA and Ratcliffe CI (1991) Solid state NMR studies of inclusion compounds. In: Atwood JL, Davies JED, and MacNicol DD (Eds.), Inclusion Compounds, vol 5, Inorganic and Physical Aspects of Inclusion. New York: Oxford University Press pp 37-85 Roberts HH and Neurauter TW (1990) Direct observations of a large active mud vent on the Louisiana continental slope. American Association of Petroleum Geologists Bulletin 74:1508 Roberts HH, Sassen R, Carney R, and Aharon P (1989) 13C depleted authigenic carbonate buildups from hydrocarbon seeps, Louisiana continental slope. Transactions, Gulf Coast Association of Geological Societies 39:523-530 Roberts HH, Sassen R, Carney R, and Aharon P (1990a) The role of hydrocarbons in creating sediment and small-scale topography of the Louisiana continental slope. In: Schumacher D and Perkins BF (Eds.), Gulf Coast Oils and Gases, Proceedings Ninth Annual Research Conference, GCSSEPM Foundation pp 311-324 Roberts HH, Aharon P, Carney R, Larkin J, and Sassen R (1990b) Seafloor responses to hydrocarbon seeps, Louisiana continental slope. Geo-Marine Letters 10:232 243 Salvador A (1987) Late Triassic-Jurassic paleogeography and origin of Gulf of Mexico basin. Bulletin, American Association of Petroleum Geologists 71:419-451 Sassen R, McCabe C, Kyle JR, and Chinn EW (1988) Deposition of magnetic pyrrhotite during alteration of crude oil and reduction of sulfate. Orgainc Geochemistry 14: 381-392 Sassen R, Brooks JM, MacDonald IR, Kennicutt MC, Guinasso NL, and Requejo AG (1993a) Association of oll seeps and chemosynthetic communities with oll discoveries, upper continental slope, Gulf of Mexico. Transactions, Gulf Coast Association Geological Societies 43 : 349 355 Sassen R, Roberts HH, Aharon P, Larkin J, Chinn EW, and Carney R (1993b) Chemosynthetic bacterial mats at cold hydrocarbon seeps, Gulf of Mexico continental slope. Organic Geochemistry 20:77-89

119 Winters JC and Williams JA (1969) Microbiological alteration of petroleum in the reservoir. Division of Petroleum Geochemistry, New York City Meeting. Washington, DC: American Chemical Society, Preprints pp E22 E31

Worrall DM and Snelson S (1989) Evolution of the northern Gulf of Mexico, with emphasis on Cenozoic growth faulting and the role of salt. In: Bally AW and Palmer AR (Eds.), The Geology of North America--An Overview, The Geology of North America. Geological Society of America pp A97-A138