Oxygen profiles and organic carbon fluxes in Laurentian Trough sediments

95

Netherlands Journal of Sea Research 21 (2): 95-105 (1987)

O X Y G E N P R O F I L E S A N D O R G A N I C C A R B O N F L U X E S IN L A U R E N T I A N T R O U G H S E D I M E N T S

N. SILVERBERG 1, J. BAKKER 2, H.M. EDENBORN 1, 3 and B. SUNDBY 2

ABSTRACT

Sediment trap samples and undisturbed cores of bottom sediment were obtained during spring and summer from a 350-re.deep site in the Laurentian Trough, Gulf of St. Lawrence, and used to determine the flux of organic carbon to the sediment surface, and oxygen and organic carbon concentration profiles close to the sediment-water interface. Oxygen uptake rates by the sediment, calculated from micro-electrode profiles and a one.dimensional molecular diffusion transport model, can only account for 20% of the organic carbon that is mineralized in the sediment. Mechanisms other than molecular diffusion must dominate the transport of oxygen across the sediment-water interface in these deep coastal sediments.

1. INTRODUCTION

Oxygen is consumed by many of the important biogeochemical processes that take place during early diagenesis of sediments. Close to the sediment-water interface, aerobic microbial degradation of organic matter and respiration by the macro- and meiobenthos communities are the greatest consumers of oxygen. Within anoxic microenvironments in the surface layer of the sediment and in the mass of underlying anaerobic sediment, anaerobic processes pro- duce reduced compounds which, upon transport into the oxygenated zone by diffusion, bio- irrigation and bioturbation, are subject to re- oxidation. The oxygen required by all of these

reactions can only be supplied through transport across the sediment-water interface.

The development of electrodes that are able to record oxygen concentrations in sediments with a sub-millimeter spatial resolution (REVSBECH et al., 1980; J(~RGENSEN & REVSBECH, 1985; REIMERS et al., 1984; HELDER & BAKKER, 1985; REIMERS & SMITH, 1986) now permits the concentration gradient of oxygen across the sediment-water interface to be measured with a high degree of detail. In principle, the flux of oxygen into the sediment, which is a function of the concentration gradient, can therefore be calculated. In practice, however, because one rarely knows all of the mechanisms of oxygen transport across the interface and because the composition of bottom sediments varies both laterally and temporally, the interpretation of data obtained with such fine-scale probes is not always straightforward (see J@RGENSEN & REVSBECH, 1985, on this point).

The complex interrelationships between the oxygen concentration profile in the sediment, the flux of organic carbon from the water column to the sediment, the carbon gradient in the sediment, the carbon degradation rate and bioturbation have been discussed for the deep-sea environment by EMERSON et al., (1985). These authors have shown that, despite the relatively low sedimentation and bioturbation rates in the deep sea as compared to coastal environments, the bulk of the organic carbon which settles is degraded within the bottom sediments rather than at the sediment-water interface. A similar conclusion has been reached for the deep coastal sediments of the Laurentian Trough

7 Departement d'Oceanographie, Universit# du Quebec, Rimouski, Qu6bec, G5L 3A 1, Canada

2Netherlands Inst i tute for Sea Research, P.O. Box 59, 1790 AB, Den Burg, Texel, The Netherlands

3present adress: Oak Ridge Research Institute, 113 Union Valley Rd., Oak Ridge, TN 37830, USA

96 N. SILVERBERG, J. BAKKER, H.M. EDENBORN & B. SUNDBY

(SILVERBERG et al., 1985). In this paper we present profiles obtained from

sediment cores recovered at a 350-m-deep loca- tion in the Laurentian Trough of the Gulf of St. Lawrence, est imate oxygen fluxes across the sediment-water interface using a one-dimensional dif fusion model, and compare these fluxes with the f luxes of fresh organic carbon to the sediment surface, and with carbon mineralization rates within the sediment.

Acknowledgements.--We express our thanks to the off icers and crew of the CSS "Louis M. Lauzier" and to Nelson Belzile and Charles Gobeil for their able assistance during the mis- sions at sea, and to Brigitte Leblanc and Frans Wetsteyn for their dedicated work on the core scraping and carbon analyses. We thank M. Rutgers van der Loeff for his helpful review of the manuscript and Wim Helder for his comments and encouragement. This research was sup- ported by grants (A9177, A8849) from the Na- tional Science and Engineering Research Council of Canada, and FCAR grant EQ-1109 from the Ministere d'Education du Qu#bec.

2. METHODS

During May and July 1985, mult iple cores were recovered from a 350-m-deep station in the central portion of the Laurentian Trough in the Marit ime Estuary of the St. Lawrence off Rimouski (Fig. 1) with an instrument modelled after that described by PAMATMAT (1971). The corer was lowered at 0.5 m-s-1 or less during its final 10 m of descent. The init ial raising of the corer out of the bottom was performed very slowly, after which the instrument was quickly raised to the surface and gently lowered onto a cradle on the ship's deck. The sediment-water interface in each of the 4 core tubes was examined, and samples were rejected if there was any sign of sediment resuspension. Those cores displaying the most regular and horizontal sediment sur- faces (one or two), were used for oxygen probe study. A sample of the overlying water in these tubes was taken for Winkler t i trat ion of oxygen. Each core was then transferred to a constant temperature bath, held at the in situ bottom water temperature of 4°C.

Micro-electrode oxygen measurements were

.50 °

7~o

Quebec

I

60° f y-

4 5 ° •

. / \ -

~ 710 ~" - - . \ ,zoOO m J J

o , ' / ~ 65 ° i - i j r h / f r . -

Newfoundland /

%

45 °

o krn 25o

60 °

Fig. 1. Chart showing the Laurentian Trough and sampling site.

OXYGEN PROFILES AND CARBON FLUXES 97

generally begun within 30 minutes of the t ime the corer had tripped on the seafloor. The polargraphic oxygen sensor (POS) used was the Model 760 needle-POS (Diamond Electro-Tech, Inc.), with a 90 ° tip, modified as described by HELDER & BAKKER (1985). Initial calibration of the electrode was done with air-saturated and nitrogen-bubbled, 4°C seawater. The electrode was ult imately calibrated using the electrode response within the air-saturated surface microlayer, and with several air-bubbled samples of the 4°C overlying water. The POS response using this procedure was very consistent, with a mean current of 408+21 pA for this water saturated at 320 #mol O2.dm-3 The zero oxygen value was taken as the constant needle-POS response at depth within the sediment (generally between 4 and 8 pA), and the response was assumed to be linear between the two extremes (REVSBECH et al., 1980).

The needle-POS was lowered through the overlying water by a mechanical device equipped with a vernier scale, accurate to within about 0.1 mm. The sediment surface was always used as the reference level for these measurements. A lamp and white background paper were used to aid in the determination of the sediment-water interface. However, because of the natural ir- regularit ies of the sediment surface and visual distort ions caused by the plastic tube walls and waterbath, the determination of the interface position was probably only accurate to +0.5 mm. Since the overlying water was not stirred, a stagnant sublayer about 2 mm thick developed above the sediment surface (Figs 3 and 4). Because the linear oxygen gradient in the sublayer did not change signi f icant ly for some distance below the sediment-water interface, knowledge of the exact position was not crit ical for calculating the diffusive flux of oxygen across the interface.

The needle-POS response was read 30 to 60 seconds after each incremental lowering of 0.5 to 1.0 mm. This t ime was suff icient to allow stabi l ization of the electrode response. The drift of the electrode during the t ime required to record the profile generally represented less than 5%, as indicated by the response in the overlying water after removal from the sediment. Only the data obtained during the init ial lowering were used for the profiles. Profiles were made at several different locations within each of the core tube studied.

At the end of the oxygen profile determina-

tions, most of the overlying water was siphoned off, the cores were frozen in liquid nitrogen and preserved at -20°C. These cores were sub- sampled by warming the plastic core liner and extruding the sti l l frozen sediment. The cap of ice (frozen overlying water) was reduced by chip- ping away and melting with a hair dryer to within 1 to 2 mm of the sediment surface. The core was then replaced in the freezer to resolidify any sur- f icial water. Subsequently the core was placed in a nitrogen-fi l led glove box in a 4°C cold room. The remaining icecap was removed with a stainless steel scalpel. Twenty successively deeper subsamples were carefully scraped from the top 10 mm of the slowly thawing cores over a 2 to 3 hour period. The samples obtained represented approximately 0.5 mm depth inter- vals from the cores.

Sealed centrifuge tubes containing the scrap- ings were then weighed, allowed to thaw, and centrifuged at 15,000 RPM for 5 to 7 minutes. The supernatant water was removed for metal analysis and weighed. There was thus some dissolved organic carbon lost from the total sample, but DOC concentrations are assumed to be insignif icant in relation to total organic carbon (BARCELONA, 1980; FROELICH, 1980). The remaining wet sediment was again weighed, and each sample was freeze-dried. The porosit ies were then calculated, corrected for the presence of salt.

The dried sediments were ground to a uniform powder. Splits were placed in nickel cups and analysed for total carbon and nitrogen in bat- ches with a Perkin-Elmer 640 CHN Analyser equipped with an automatic sample feed. Previous studies (BOUCHARD, 1983) have shown that the inorganic carbon component is constant with depth and represents less than 0.04%C of the dry sediment weight in Marit ime Estuary muds. We therefore refer to the total carbon measured as organic carbon. Tripl icate analyses were run for all depths in several cores, with the order of sample insertion varied from one run to another in order to minimize errors due to signal drift. These procedures, with all manipulat ions done by the same operator, reduced the analytical error to a standard deviation of 1 to 2% of the mean, equivalent to a concentration of _+ 0.02%C.

Samples of the material sett l ing through the water column were obtained at the site during the sampling programme with a free-drifting sediment trap. These provided daily measures of


the total solid and organic carbon flux to the underlying sediment. The methods used are described in SILVERBERG et al. (1985).

3. RESULTS

3.1. OXYGEN DATA

Table 1 l ists the codes for the different core tubes examined (of the 4 tubes recovered on each cast) for the various dates. The oxygen profi les for the different spots in each of the cores are presented in Fig. 2 (these are coded sequen- t ia l ly through time, e.g. 2 - 117 would be the profi le obtained at the second spot in core tube O, 117 minutes after the corer struck the seafloor). It is apparent that there are considerable varia- t ions in both the shapes of the profiles and oxygen penetration depths from spot to spot in the same core and from core to core. Some of this variabi l i ty is due to the natural heterogeneity of the seafloor, as has been demonstrated to occur in the deep sea by simultaneously obtained in si tu measurements (REIMERS et al., 1986).

Because these measurements were not performed in si tu it is important to evaluate their representativity. Although the recovery of cores from the Laurentian Trough involves much less physical disturbance of the samples than occurs in deep-sea situat ions (the samples remain in their original tubes, the sediment-water interface is recovered intact, there is l i t t le t ime for temperature changes to occur, and any decompression effects are much smaller), the measurements were obtained under incompletely controlled conditions. Thus, we did not isolate the overlying water from the atmosphere, the lack of stirring induced an unnaturally thick "di f fusive boundary

TABLE 1 Core tubes examined for oxygen and carbon profiles.

Core code Day Cast

L May 18 1 M May 18 1 N May 19 1 O May 19 1 P May 20 1 Q May 20 1 R July 12 1 S July 12 2 T July 12 2 U July 13 1 V July 13 1

layer" above the sediment surface, and the dura- tion of the experiments may have been of suffi- cient length to cause important changes in the profiles.

To evaluate the possible influence of experimental art i facts, we examined the evolution of the profiles in both the water column and the sediment as a function of t ime after the corer tripped on the bottom. Fig. 3 shows a typical set of profiles. It can be seen that the concentration gradient at the air-water interface extends to several cm depth within the water overlying the cores. There is a zone below this, however, over which the oxygen concentrations are very uniform. These concentrations (80 to 110 #M) are similar to those measured in the samples taken from the overlying water immediately upon recovery, and to those previously reported for the deep Laurentian Trough (D'ANGLEJAN & DUNBAR, 1968). Below this uniform zone is a linear gradient in oxygen concentration, representing the diffusive sublayer overlying the sediment surface (J(~RGENSEN & REVSBECH, 1985). It is apparent that the effect of oxygen contamination from the air was minor during the time period of these unstirred experiments.

The increase in the boundary layer thickness during the unstirred experiments, however, might decrease the flux of oxygen across the sediment- water interface (RUTGERS VAN DER LOEFF et al., 1984; SUNDBY etal . , 1986). It is possible that, with a reduced flux, the oxygen penetration depths in- to the sediment may have been decreased from in situ depths. In order to est imate the possible importance of this experimental artifact, we performed an addit ional experiment. About one half gram of freeze-dried Tubifex worms (enough to form a layer about 0.5 mm thick) were crushed up and dispersed in the overlying water of tube V after the original experiment was completed, and allowed to settle. Micro-POS (REVSBECH et al., 1980) profiles were then measured at the same spot repeatedly over the next few hours. The rapid decrease in the oxygen concentrations, both in the overlying water and in the sediment, are very apparent (Fig. 4). The sudden addition of a large quantity of metabolizable organic matter produced an upward shift in the depth of oxygen penetration of approximately 1 mm.h-1. For the normal concentrations of organic matter occur- ring in these sediments the change in penetration depth would have been considerably slower. This is in accord with REIMERS et al., (1986), who report l i t t le change in penetration depths be-


Depfh 0

( r a m ) -1

-2

-3

- 4

-5

-6

-7

-8

-9

-10

Dissolved Oxygen juM

0 20 40 60 80 I00

, , ,

t °

-'- N 1-30

-o- N2-65

-I- 01-90

o- 02-117

-~ P 1-30

-~ P2-45

-x- Q 1-70

-x- Q2-85

120

0

-I

-2

-3

-4

-5

-6

-7

-8

-9

-10 o

0 20 40 60 80

I I ~ ~ x ,

_ s

_El / /

b

100 120

July

-o- R 1-40

-o R4-70

I- $I-30

-o- $2-40

-A- T-50

-~- UI-140

x- U3- 185

z- V 1-420

Fig. 2. Profiles of dissolved oxygen (#M) in the sediment cores. The legend code refers to core tube (Table 1), spot in core, and approx, time in minutes after corer

touched the bottom; a) observations in May, b) observations in July.

tween in s i tu and box-core oxygen profiles. We therefore conclude that the difference in oxygen penetration depth among the various measured profiles reflects mostly natural sediment heterogeneity. In this experiment, where the added organic matter provided a strong oxygen sink at the interface, there was very l i t t le change in the oxygen gradient itself, although the levels drop- ped rapidly. This suggests that the f luxes may also be l i t t le affected during the core retrieval

and oxygen measurements. The flux of oxygen across the sediment-water

interface was calculated by assuming transport by simple molecular diffusion, according to Fick's law of diffusion,

J = eDsAC/AX (1)

where J is the oxygen flux in #mol-cm-2.d -1, AC/AX is the oxygen concentration gradient

100 N. SlLVERBERG, J. BAKKER, H.M. EDENBORN & B. SUNDBY

Depth (ram)

0

80

70

60

50

40

30

20

10

0

-10

4O 8O I I

-o- NI-30

-o 02-I 17

-I- PI-30

°- Q 1-70

-A- $2-40

, ,

Dissolved Oxygen NM 120 160 200 240 280 320

- - ~ - - I I . . . . . 4 - - 4 - - I

sediment-water interface

Fig. 3. Examples of oxygen concentration (#M) profiles, including the overlying water.

4

Depth (ram)

2

-2

-4

-6

pico Amperes 50 100 150 200

& lO J& /

/ = - ~ . ~ - P ~ = / ~ minutes

~ ~ Oxygen profile evolution 240

Fig. 4. Evolution of the oxygen profiles (expressed in picoAmperes) with time in minutes following the experimental addition of organic matter to the sediment

surface.

across the sediment-water interface, represented by the l inear port ion of the profi le closest to the sediment surface, e is the mean porosity measured over the same depth interval, and D s is the bulk sediment d i f fus ion coefficient. D s is assumed to be equal to ~2.D 0 for high

porosity coastal sediments (ULLMAN & ALLER, 1982), and D o is the free solut ion di f fusion coefficient; 1.128 cm2.d -1 for oxygen at 4°C.

For the purpose of this study we shall present only a single f lux est imate from each core tube, based upon the earl iest regular profi le obtained,


TABLE 2 Oxygen flux calculations.

Core Depth interval mean porosi ty D s 02 flux prof i le (ram) (e) (cm2.d 1) (l~mol.cm - 2"d- 1)

M-1 0.0 - 1.0 0.908 0.930 0.283 N-1 0.5 - 1.0 0.888 0.889 0.367 O-1 0.0 - 2.0 0.913 0.940 0.168 P-1 1.0 - 2.5 0.880 0.874 0.178 Q-1 0.0 - 1.5 0.875 0.864 0.332 R-2 0.0 - 1.0 0.914 0.942 0.196 S-1 0.5- 1.5 0.926 0.967 0.300 T 0.5 - 3.5 0.910 0.934 0.161 U-1 0.0- 1.5 0.916 0.946 0.220 V-1 0.0 - 1.0 0.920 0.955 0.257

MEAN 0.246

as this will minimize temporal artifacts. The range of these results (Table 2) is probably a good indication of environmental variabil i ty.

Oxygen f luxes calculated for Laurentian Trough sediments, assuming transport by molecular dif fusion across the sediment-water interface, range from 0.16 to 0.37 /~mol.cm-2.d -1. Individual profi les reveal a high degree of lateral variabi l i ty with a spatial scale as small as several cm, and with oxygen penetration depths varying between 2 and 10 mm. The calculated fluxes are within the range reported from various deep-sea environments, viz. 0.03 to 0.36 ~mol.cm-2.d -1 (REIMERS et al., 1986), but are considerably lower than those reported for

shallow, near-shore marine sediments, viz. 0.7 to 8.1 ~mol.cm-2.d -1 (RUTGERS VAN DER LOEFF et al., 1984; SQRENSEN et al., 1979; REVSBECH et al., 1980; HOPKINSON & WETZEL, 1982). The uptake rates thus generally fol low the expected pattern of increasing oxygen demand with typical ly increasing supply of organic matter from the deep- sea to near-shore environments. In the latter environment higher temperatures also contr ibute to increased biological demand.

3.2. ORGANIC CARBON DATA

The distr ibut ions of organic carbon with depth at the top of the cores are presented in Fig. 5. Cores

1.75

0

D e p t h

( m m )

-2

-4

-G

-8

- l o

P e r c e n t C a r b o n

2 .00 2 .25 2 .50 2 .75 ~ .00 3.25

. t z X ~ I I I •

x •

X . •

x x ] o

o :: x 0 0 t~ N

• • x

X • •

I I 0 X T

0 • • & U EX • •

,o.x • • o v

Fig. 5. Vertical distributions of the total organic carbon concentration (%C) in the sediment cores. Each sampling point represent a depth interval of approx. 0.5 mm.

Least-squares regression lines are drawn.


L and M show a marked decrease in gradient below 2 to 3 mm depth, and core V below 6 mm depth, perhaps indicative of the shift from aerobic to anaerobic degradation, and only the upper port ions were used for cores L and M when the gradients (Table 3) were calculated from the least-squares linear regression lines shown in Fig. 5. Although the top 0.5 mm of core V were not included in carbon profi les we cannot be cer- tain that some degree of bioturbat ion had not oc- curred during the course of the experiment. The experiment lasted only a few hours, however, and the occurrence of higher gradients in core L (which had not been contaminated by experimental addit ion of organic matter) suggests that the few organisms present in the small surface of the cored sediment had litt le t ime to seriously af-

fect the subsurface carbon gradient. The carbon gradients range from 0.07 to 4.98%C.cm -1 These are an order of magnitude greater than the gradients measured over 35 cm depth in box cores. Similar measurements from the upper 3 mm of cores obtained at this site in 1984 had a range of 0.42 to 1 .31%C.cm-1 (SILVERBERG et al., 1985).

3.3. SEDIMENT TRAP DATA

Table 4 shows the sedimentat ion rates, carbon content of the sett l ing material and the carbon flux to the sediment during the sampling programme. The higher carbon contents and lower flux rates in July as compared with May are part of a seasonal pattern in the St. Lawrence, and is

TABLE 3 Carbon gradients at the tops of the cores.

Core Date carbon gradient Corr. coefficient Carbon loss* (%C.cm- 1) (l~mol.cm-2.d 1)

L 05/18/85 4.98 0.96 (first 5 pts) 1.939 M 0.34 0.94 (first 8 pts) 0.132 N 05/19/85 0.16 0.94 0.062 O 0.21 0.86 0.082 P 05/20/85 0.07 0.67 0.027 Q 0.13 0.77 0.051 T 07/12/85 0.07 0.61 0.027 U 07/13/85 0.10 0.80 0.039 V 1.07 0.97 0.417

MEAN: 0.792 (185.4 #mol.cm-4) 0.308

Range for Box-cores: 0.012 to 0.018 %C.cm -1

*calculation based upon average porosity of 0.894 and sedimentation rate of 601.3 d-cm-

TABLE 4 Sediment trap measurements and carbon fluxes.

Date Total flux Carbon content (#g.cm-2.d ~) (%C)

Total

Carbon flux (#mol.cm -2.d- 1)

Mine ra lize d TM

May 18 649 3.88 2.10 1.320 May 19 775 3.94 2.55 1.615 May 20 830 3.18 2.20 1.204 July 12 245 6.31 1.29 0.994 July 13 336 4.66 1.31 0.902

M EAN 567 4.39 1.89 1.207 MEAN (b) 469 4.23 1.65 1.090

la).calculated by applying correction for refractory carbon burial (C content at 35 cm depth = 1.44%) (b)-based upon 32 sediment trap measurements for 1980-1985 (and 21°pb dating).


related to the dominance of continental runoff in spring and increased biological production in the estuary during summer and fall (SILVERBERG et al., 1985, 1986a).

4. DISCUSSION

The greatest proportion of the organic matter f lux that reaches the surface of deep-sea sediments is degraded in the oxygenated zone of the sediment (BENDER & HEGGIE, 1984; EMERSON et al., 1985), and most of this degradation can be accounted for by estimates of oxygen uptake using a simple one-dimensional diffusion model (EMERSON et al., 1985; REIMERS & SMITH, 1986). At a site in the Santa Catalina Basin, however, there was some discrepancy between oxygen uptake rates, calculated from micro-electrode profiles, and direct measurements of sediment respiration rates using a benthic f lux chamber (REIMERS et al., 1986), suggesting that mechanisms other than molecular diffusion may contribute to the transport of oxygen into sediments. In general, however, work in the deep-sea has shown relatively good agreement between organic carbon fluxes, carbon gradients in the top layer of the sediment and oxygen uptake rates.

In shallow environments bio-irrigation has been shown to be an important factor in sediment-water exchanges (ALLER & YINGST, 1985; RUTGERS VAN DER LOEFF et al., 1984). In Laurentian Trough sediments, bioturbation has been shown to be moderately high (SUNDBY & SILVERBERG, 1985; SILVERBERG et al., 1986b) and biological transport mechanisms have been in- voked to account for the exchange of both electron acceptors and mineralization products (BOUCHARD, 1983; SUNDBY et al., 1983).

If one-dimensional molecular diffusion along the measured concentration gradients across the sediment-water interface were an adequate representation of the transport of oxygen into these sediments, then the diffusive flux of oxygen should be able to account for the mineralization of organic carbon. We have therefore calculated the average rate of production of mineralized carbon in the sediment column by two different methods: from the carbon fluxes to the sediment, measured with sediment traps; and from profiles of organic carbon in the sediment. In making these calculat ions we made the simpl i fy ing assumptions that the carbon flux to the sediment consists of a refractory and a degradable component, that the refractory com-

ponent is buried, and that the degradable component consumes oxygen on a mole for mole basis. The data used and the results of these calcula- t ions are shown in Tables 3 and 4.

The flux of organic carbon mineralized in the sediment was estimated by mult iplying the difference between the carbon content measured in the traps and the refractory carbon which is buried (taken as 1.44%C, the average concentration at 35 cm depth in the box cores) by the sedimentation rate. The average flux of mineralized carbon during May and July is thus 1.21 ~mol .cm-2.d- 1 with individual values rang- ing from 0.9 to 1.6/~mol.cm-2.d -1. These values are almost 5 t imes greater than the average diffusive flux of oxygen into the sediment (0.246 #mol.cm -2.d - 1, Table 2).

This method assumes that all of the settl ing organic matter is mixed into the sediment before it is degraded. This assumption can be roughly tested, assuming diffusion-analogous particle mixing, by applying the best est imate for a bioturbation coefficient, viz. 1.3x10 -2 cm2.d -1 (SILVERBERG et al., 1986b) to the average of all of the organic carbon gradients near the sediment- water interface (185.4 #moles C.cm-4), to calcu- late the average bioturbational flux of organic carbon into the sediment. This biological transport mechanism can account for a carbon flux of 2.41 #moles C.cm-2.d -1, which is sl ightly greater than the average flux of organic carbon through the water column measured with the sediment trap in May and July (1.89 /~moles C.cm-2.d-1), or the overall average for 32 trap measurements over a 5-year period (1.65 #moles C.cm-2d-1) . Thus the assumption that l i tt le of the sedimenting material is degraded either during the final 200 m of settl ing in the water column below the depth of the trap, or at the sediment- water interface itself before it is mixed down into the sediment, is probably valid.

Since there is some uncertainty as to the precise value or the representativity through t ime of the mixing coeff icient used for the calcula- t ions of bioturbational flux of carbon, it is in- structive to estimate, using the second method, the carbon mineralization flux in the total absence of bioturbation, and compare this est imate with the diffusive oxygen flux. Here the carbon mineralization rates over the top cen- t imeter (Table 3) have been calculated by mult iplying the carbon gradient near the sediment-water interface (least squares fit, Fig. 5) by the sedimentation rate, assuming an


average porosity of 0.894, a sol id component densi ty of 2.65 g.cm -3, and a mean sedimenta- t ion rate of about 600 d.cm -1 (based upon 32 sediment trap measurements and 210pb dating; SILVERBERG et al., 1986b). The results (Table 3), which are highly variable, show that in some cores the di f fusive oxygen f lux (Table 2) is suffi- cient to account for the carbon loss and that in other cores it is not. The average values of the di f fusive oxygen f lux and the carbon mineraliza- t ion f lux, 0.246 and 0.308 #mol .cm-2.d-1, respectively, are very comparable. This apparent agreement between the f luxes is s imi lar to what has been described for deep-sea sediments, where biological mixing is minimal. The magnitude of these est imates also fal ls into the range of those reported from the deep sea. This is, of course, not reasonable for a coastal en- v i ronment where both primary product ion and biological mixing are orders of magnitude greater than the deep sea.

For these reasons the method based upon the di f ferences between carbon input and burial is much more appropriate. The approximate difference, by a factor of 5, between the rate of carbon mineral izat ion estimated by this method and the f lux due to molecular d i f fus ion is s imi lar to the transport est imated for non-molecular diffusion mechanisms, incuding bio-irr igation, in other coastal environments (ALLER, 1980; ANDERSEN & HELDER, 1987; CHRISTENSEN et al., 1984; RUTGERS VAN DER LOEFF et al., 1984). Accor- dingly we conclude that, in contrast to many deep-sea sediments, molecular d i f fus ion is not an adequate representation of oxygen transport across the sediment-water interface in Lauren- t ian Trough sediments.

This implies that the use of oxygen microelectrode measurements is of l imited value for direct ly est imat ing oxygen f luxes in densely populated coastal sediments. As BERNER (1976) has pointed out, in such environments it is better to obtain direct f lux measurements, e.g. bell-jar experiments, rather than relying on molecular d i f fus ion calculat ions based upon individual profiles. On the other hand, if bottom chambers are relied upon exclusively, there is the danger of in- c luding the oxygen respiration of macrobenthic organisms, which exchange oxygen and mineralized carbon direct ly with the overlying water. Such direct exchange would be most ly decoupled from the early diagenet ic processes we hope to describe wi th in the sediment.

5. CONCLUSIONS

Average oxygen respiration rates in Laurentian Trough sediments, calculated by the dif ference between the rate of input of fresh organic matter from the water column and the burial rate of refractory carbon, are about 5 t imes greater than the oxygen f luxes calculated on the basis of one- d imensional molecular d i f fus ion and porewater oxygen profiles. In the absence of detai led knowledge of transport mechanisms, neither oxygen profi les nor organic carbon profi les provide suf f ic ient information about the true extent of respiratory f luxes in coastal sediments.

6. REFERENCES

ALLER, R.C., 1980. Relationships of tube-dwelling ben- thos with sediment and overlying water. In: K.R. TENORE & B.C. COULL. Marine benthic dynamics. Belle W. Baruch Library in Marine Science 11: 285-310.

ALLER, R.C. & J.Y. YINGST, 1985. Effects of marine deposit-feeders Heteromastus fi l i formis (Polychaeta), Macoma balthica (Bivalvia), and Tellina texana (Bivalvia) on averaged sedimentary solute transport, reaction rates, and microbial distributions.--J, mar. Res. 43: 615-645.

ANDERSEN, F.O. & W. HELDER, 1987. Comparison of oxygen microgradients, oxygen flux rates and electron transport system activity in coastal marine sediments.--Mar. Ecol Prog. Ser. (in press).

BARCELONA, M.J., 1980. Dissolved organic carbon and volatile fatty acids in marine sediment pore waters.--Geochim, cosmochim. Acta 44: 1977- 1984.

BENDER, M.L. & D.T. HEGGIE, 1984. Fate of organic carbon reaching the deep sea floor: A status report.--Geochim, cosmochim. Acta 48: 977-986.

BERNER, R.A., 1976. The benthic boundary layer from the viewpoint of a geochemist. In: I.N. MCCAVE. The benthic boundary layer. Plenum Press, N.Y.: 33-55.

BOUCHARD, G., 1983. Variations des parametres biogeochimiques dans les sediments du Chenal Laurentien. M. Sc. thesis, Universite du Quebec & Rimouski: 1-161.

CHRISTENSEN, J.P., A.H. DEVOL & W.M. SMETHIE, 1984. Biological enhancement of solute exchange between sediments and bottom water on the Washington continental shelf.--Continental Shelf Res. 3: 9-23.

D'ANGLEJAN, B.F. & M.J. DUNBAR, 1968. Some observations on oxygen, pH and total alkalinity in the Gulf of St. Lawrence 1966, 1967, 1968. McGill Universi- ty, Marine Sciences Center, Manuscript Report no. 7: 1-50.


EMERSON, S., K. FISCHER, C. REIMERS & D. HEGGIE, 1985. Organic carbon dynamics and preservation in deep-sea sediments.--Deep-Sea Res. 32: 1-21.

FROELICH, P.N., 1980. Analysis of organic carbon in marine sediments.--Limnol. Oceanogr. 25: 564- 572.

HELDER, W. & J.P. BAKKER, 1985. Shipboard comparison of micro- and minietectrodes for measur- ing oxygen distribution in marine sediments. - -L imnol. Oceanogr. 30: 1106-1109.

HOPKINSON, C.S. & R.L. WETZEL, 1982. In-situ measurements of nutrient and oxygen fluxes in a coastal marine benthic community.--Mar. Ecol. Prog. Ser. 10: 29-35.

J(~RGENSEN, B.P. & N.P. REVSBECH, 1985. Diffusive boundary layers and oxygen uptake of sediments and detr i tus.--Limnol. Oceanogr. 30: 111-122.

PAMATMAT, M.M., 1971. Oxygen consumption by the seabed. IV. Shipboard and laboratory measurements.--Limnol. Oceanogr. 16: 536-550.

REIMERS, C.E. & K.L. SMITH, 1986. Reconciling measured and predicted fluxes of oxygen across the deep-sea sediment-water interface.--Limnol. Oceanogr. 31: 305-318.

REIMERS, C.E., S. KALHORN, S.R. EMERSON & K.H. NEALSON, 1984. Oxygen consumption rates in pelagic sediments from the Central Pacific: First estimates from microelectrode prof i les.-- Geochim. cosmochim. Acta 48" 903-910.

REIMERS, C.E., K.M. FISCHER, R. MEREWETHER, K.L. SMITH & R.A. JAHNKE, 1986. Oxygen microprofi les measured in-situ in deep ocean sed imen ts . - Nature 320: 741-744.

REVSBECH, N.P., B.B. J(~RGENSEN & T.H. BLACKBURN, 1980. Oxygen in the sea bottom measured with a microelectrode.--Science 207: 1355-1356.

RUTGERS VAN DER LOEFF, M.M., L.G. ANDERSON, P.O.J. HALL, A.K. IVERFELDT, A.B. JOSEFSON, B. SUNDBY & S.F.G. WESTERLUND, 1984. The asphyxiation technique: An approach to distinguishing between molecular diffusion and biological ly mediated transport at the sediment-water interface.--Limnol. Oceanogr. 19: 675-686.

SILVERBERG, N., H.M. EDENBORN & N. BELZILE, 1985. Sediment response to seasonal variations in organic matter input. In: A.C. SIGLEO & A. HAT- TORI. Marine and estuarine geochemistry. Lewis Publ. Inc., Chelsea, MI: Ch. 5: 69-80.

SILVERBERG, a., C. GOBEIL & H.M. EDENBORN, 1986a. On-station versus seasonal variations in early diagenesis parameters in sediments of the Laurentian Trough. Abstr. EOS. 66: 1307.

SILVERBERG, N., H.V. NGUYEN, G. DELIBRIAS, M. KOIDE, B. SUNDBY, Y. YOKOYAMA & R. CHESSELET, 1986b. Radionuclide profiles, sedimentation rates, and bioturbation in modern sediments of the Lauren- tian Trough.--Oceanol. Acta 9: 285-290.

S(~RENSEN, J., B.B. J~RGENSEN & N.P. REVSBECH, 1979. A comparison of oxygen, nitrate and sulphate respiration in coastal marine sediments.-- Microb. Ecol. 5: 105-115.

SUNDBY, B. & N. SILVERBERG, 1985. Manganese fluxes in the benthic boundary layer.--Limnol. Oceanogr. 30: 372-381.

SUNDBY, B., G. BOUCHARD, J. LEBEL & N. SILVERBERG, 1983. Rates of organic matter oxidation and carbon transport in early diagenesis of marine sediments. Advances in organic chemistry, 1981: 350-354.

SUNDBY, B., L.G. ANDERSON, P.O.J. HALL, A.K. IVERFELDT, M.M. RUTGERS VAN DER LOEFF & S.F.G WESTERLUND, 1986. The effect of oxygen on release and uptake of cobalt, manganese, iron and phosphate at the sediment-water interface.-- Geochim cosmochim. Acta 50: 1281-1288.

ULLMAN, W.J. & R.C. ALLER, 1982. Diffusion coeffi- cients in nearshore marine sediments.--Limnol. Oceanogr. 27: 552-556.

(received 13-3-1987; revised 24-4-1987)

Oxygen profiles and organic carbon fluxes in Laurentian Trough sediments

Documents