Temporal variations of sedimentary sulfur in a Delaware salt marsh

Marine Chemistry, 23 (1988) 311-327

311

Elsevier Science Publishers B.V., Amsterdam - - Printed in the Netherlands

TEMPORAL VARIATIONS OF S E D I M E N T A R Y S U L F U R I N A DELAWARE SALT MARSH

GREGORY A. CUTTER and DAVID J. VELINSKY*

Department of Oceanography, Old Dominion University, Norfolk, VA 23529-0276 (U.S.A.) (Received June 1, 1987; revision accepted September 21, 1987)

ABSTRACT Cutter, G.A. and Velinsky, D.J., 1988. Temporal variations of sedimentary sulfur in a Delaware salt marsh. Mar. Chem., 23: 311-327. The cycling of sedimentary sulfur was examined over a one year period in the Great Marsh, Delaware (U.S.A.) using newly developed analytical procedures. Iron monosulfide (FeS) and elemental sulfur both display large seasonal changes in concentration and distribution with depth, indicating a coupling with marsh redox conditions. In contrast, the depth distribution and concentration of greigite (Fe3S4) did not show appreciable changes with season. Pyrite (FeS2) underwent large concentration changes in the upper 15 cm of sediment during the spring, but remained relatively constant with respect to concentration and distribution below this zone. Using a mass balance approach in the upper marsh sediment, sulfur needed for rapid pyritization is found to be derived from elemental sulfur, iron monosulfide and sulfate reduction. In the deeper sediments, pyritization occurs through a greigite intermediate, and diagenetic modeling indicates that pyrite formation is limited by the synthesis of greigite, and not by the conversion of greigite to pyrite.

INTRODUCTION

The sulfur cycle in coastal marine sediments has received considerable attention due to the pivotal role of sulfate in anaerobic respiration and the resultant production of authigenic sulfide minerals (Goldhaber and Kaplan, 1974). Thus, the cycling of sulfur in anoxic environments affects the biogeochemical cycling of carbon and trace elements, as well as the maintenance of biological activity. Furthermore, it is proposed that reduced sulfur serves as an energy source for chemolithotrophic bacteria (Howarth, 1984). Bacterially produced sulfide can react with a number of metal cations, iron being the most abundant in marine sediments. Of the possible iron-sulfur compounds, pyrite (FeS2) is the only thermodynamically stable phase in marine sediments (Berner, 1967). A variety of pyrite synthesis schemes have been proposed (see review in Rickard, 1975), including the reaction of mackinawite (FeS0.94) with elemental sulfur or polysulfides, the dissociation of greigite (Fe3S4), and the direct reaction of Fe(II) and polysulfides. As a result, mackinawite and greigite can be important intermediates in pyritization. Moreover, greigite may be a * Present address: College of Marine Studies, University of Delaware, Lewes, DE 19958, U.S.A.

0304-4203/88/$03.50

© 1988 Elsevier Science Publishers B.V.

'312

precursor in the synthesis of the commonly observed 'framboidal' form of' pyrite (Sweeney and Kaplan, 1973). Laboratory studies indicate that pyrite formations, via mackinawite and greigite proceeds more slowly than direct precipitation (Rickard, 1975). However. demonstrating the existence of certain pyrite intermediates in the environment has been hampered by the lack of sensitive and selective analytical techniques. Recently developed methods are now able to discriminate between the various forms of sedimentary sulfur (elemental sulfur and sulfur in mackinawite, greigite and pyrite) and still possess low detection limits for the examination of natural sediments (e.g. Cutter and ()atts~ 1987). A salt marsh site was chosen for a careful examination of sedimentary sulfur cycling using these new methods. In addition to the ease of obtaining sediment cores from this environment, the salt marsh system provides an excellent natural laboratory for examining anoxic processes in marine sediments. Owing to seasonal changes in temperature and the injection of oxygen via photosynthesis by marsh grass (Spartina alterniflora), the redox cycle of marsh sediments is clearly resolved over short time scales (months) and depths (centimeters). However, salt marsh sediments still retain characteristics of other anoxic coastal marine sediments, particularly with respect to the cycling of sulfur and carbon (Howes et al., 1984). Thus, data which can be easily obtained in a marsh system may be relevant to other coastal anoxic systems (e.g. Framvaren Fjord). The role of sulfate reduction in the cycling of both sulfur and carbon in salt marsh environments has been the focus of several studies. Howarth and Teal (1979) found that pyrite formed rapidly in upper marsh sediments, and that it represented the major fraction of sedimentary sulfur. In contrast, King et al. (1985) concluded that acid volatile sulfides (mackinawite and greigite) and elemental sulfur are the short-term products of sulfate reduction m a salt marsh. Both these studies employed radiolabeling with :~'~S,a technique which must be used with caution since isotope exchange between the different sulfur phases has been observed (Jorgensen et al., 1984). In addition, processes were examined only on short time scales (days), and in one study (King et al., 1985) only the upper t0 cm of sediment were investigated. Lord and Church (1983) exploited the seasonal redox cycling of a Delaware salt marsh in order to examine sulfate reduction and pyritization. They utilized diagenetic modeling of porewater and solid phase constituents to obtain the rates of pyritization at various depths in the sediment. This paper examines sulfur cycling over a one year period, and presents data on the actual solid phases of sulfur. This examination of sedimentary sulfur is complimented by a parallel study of dissolved sulfur speciation in the same marsh (Luther and Church, 1988, this volume). In this manner the cycling of sedimentary sulfur is examined, particularly the mechanisms of pyritization, through a careful analysis of sulfur pools in the marsh sediment.

313 STUDY SITE AND METHODS

Study site Sediment cores were obtained from Delaware's Great Marsh, located near Lewes, Delaware on the southeastern shore of Delaware Bay. The sampling site is the same as that used by Lord and Church (1983) and Luther and Church (1988, this volume). Other than the digging of mosquito control ditches during the 1930s, the marsh around the sampling site is relatively pristine. The area is covered by the short form of the marsh grass, Spartina alterniflora. The sediment accumulation rate has been estimated at 0.47 cm y-1 (Church et al., 1981). Tidal inundation of the marsh occurs only during the highest tides of each month, but the marsh sediments generally remain saturated with water. Samples were obtained on April 4, June 19 and December 5, 1985 and March 26 and June 26, 1986. The seasonal biogeochemical cycles in Great Marsh sediments have been described by Lord (1980). While sulfate reduction occurs throughout the marsh sediment, during the spring and early summer, S. alterniflora photosynthesis injects oxygen into the upper 15cm of sediment. This process causes the oxidation of iron sulfides (e.g. mackinawite), which in turn releases protons and sulfate, and causes iron oxides and elemental sulfur to precipitate. In the autumn, oxygen infusion slows and sulfate reduction predominates; oxidized sulfur and iron from the preceding season are reduced and reprecipitated as iron sulfides. During winter both oxidation and reduction rates are slow, but iron sulfides continue to precipitate due to the upward diffusion of porewater Fe(II) and the slow production of hydrogen sulfide. During our sampling period, this general sequence was affected by a drought in the eastern U.S.A. which began in the winter of 1986. Desiccation of the upper marsh sediment caused an increased penetration of atmospheric oxygen, and a resultant decrease in the inventory of iron sulfides.

Methods Sediment samples were obtained by driving a 5.6 cm (i.d.) cellulose acetate/ butyrate tube into the sediment. Before removing the core, compaction was estimated by measuring the distance from the top of the core tube to the sediment surface (both inside and out), and a delrin plug with O-rings and venting valve was inserted into the core top. After manually removing the core, the lower end was capped, and the core was returned to the University of Delaware's nearby laboratory within one hour. Cores were sectioned at 2.53.0cm intervals inside a nitrogen-purged glove box, and pore waters were removed by pneumatic squeezing (Reeburg, 1967). The sediment sections were then placed in polyethylene bags and were immediately frozen. At Old Dominion University, determinations of iron monosulfide, greigite-

314

and pyrite-sulfur were made using the technique of' Cutter and Oatts (1987). In brieE the determination of FeS involves placing 5-20mg of frozen sediment (wet/dry weight ratio determined on a separate aliquot) and l0 ml of distilled water in a glass stripping vessel which is purged with helium. Hydrochloric acid is added to a concentration of 0.5 M. Hydrogen sulfide is purged from the solution and is trapped in a glass U-tube filled with Porapak QS which is immersed in liquid nitrogen. After 15 min of purging/trapping, the U-tube is removed from the liquid nitrogen and the hydrogen sulfide volatilizes and is chromatographically separated from other volatile compounds (e.g. methyl mercaptan) prior to entering a photoionization detector. The U-tube is then re-immersed in liquid nitrogen, and solutions of potassium iodide and sodium borohydride are added to generate hydrogen sulfide selectively from greigitesulfur. The reaction time is 15 min and the resulting hydrogen sulfide is chromatographed and detected as above. The precision (as relative standard deviation) for the determination of FeS is 4% and that for greigite is 7%. For the determination of pyrite, sediments are dried and then ground with an agate mortar and pestle. The homogenized sediment is extracted with carbon tetrachloride to remove elemental sulfur and is then re-dried. Approximately 1 mg of extracted, dried sediment is placed in the stripping apparatus and 0.5M HC1 plus potassium iodide and sodium borohydride solutions are added. The mixture is purged with helium to remove any residual FeS and greigite (the trap is not immersed in liquid nitrogen). The trap is then placed in liquid nitrogen, the acid concentration is adjusted to 3 M HC1, and 10 ml of Cr(II) solution is added. Hydrogen sulfide generated from pyrite sulfur is trapped over a period of 22 min and is determined as above. Precision for the determination of pyrite is 4% (relative standard deviation). Elemental sulfur is determined using a method developed by T. Ferdelman and G.W. Luther (unpublished, 1986) at the University of Delaware. This technique uses the reaction between elemental sulfur and sulfite to form thiosulfate quantitatively, and is adapted from the method of Luther et al. (1985) for the determination of dissolved polysulfides. Approximately 250 mg of frozen sediment are placed in a small test tube, 5ml of 0.1M sodium sulfite solution is added, and the sediment is disrupted with a 2 kHz sonic probe. The test tubes are purged with argon, capped and placed in a sonic bath (water at 80°C) for 8 h. After this period, the solution is filtered and determinations of thiosulfate are made using linear sweep voltammetry. The precision for this method is 7% (relative standard deviation). Total sulfur, carbon and nitrogen were determined using a Carlo Erba ANA 1500 elemental analyzer. For these analyses, sediments which had been dried and then ground with an agate mortar and pestle were used. Precisions for carbon, nitrogen and sulfur average 0.8%, 1.5% and 2.0% (relative standard deviation), respectively. Reactive iron oxides were leached from sediments using the hydroxyl amine/acetic acid treatment described by Tessier et al. (1979). Iron is determined in the leachate using atomic absorption spectroscopy. Since the iron in iron monosulfide would be included in the Tessier et

315 al. (1979) method, the iron oxide concentrations were corrected using the FeS data. RESULTS AND DISCUSSION

Total carbon, nitrogen and sulfur Sulfate reduction in a salt marsh is driven by the large inputs of organic matter to the marsh environment (Howarth, 1984), and consequently the formation of iron-sulfur minerals is also affected by this organic matter input (Berner, 1970). Organic carbon values in the surface sediments (0-3cm) averaged 8.32%, while those below 30cm averaged 4.53%. In contrast with other marine sediments where the major organic carbon input is detritus, organic carbon in salt marsh sediments is formed in situ by S. alterniflora photosynthesis (Valiela et al., 1976). In Fig. 1 the variation of carbon/nitrogen (atomic) ratios with depth are shown for the five sampling periods. As is generally seen in marine sediments, the carbon/nitrogen ratios increase with depth, indicating that more nitrogen-rich organic material is being selectively regenerated. In the upper 15cm of sediment where most of the biological productivity occurs, seasonal trends are not readily apparent in the carbon and nitrogen data. It is unclear why the 12/85 core is inconsistent with the others, but anomalies in this core appear in the other data as well. For this reason, the December 1985 results will only be treated qualitatively. In comparison with nitrogen and carbon, total sulfur (Fig. 2) shows considerably more variation with depth in all cores. These changes in total sulfur are due primarily to variations in pyrite (discussed below). In spite of this variability total sulfur generally increases with depth, averaging 6.3 mg g t in surface sections (0-3 cm) and 13.2 mg g 1 below 30 cm (12/85 data not included). At the sediment surface, carbon/sulfur (atomic) ratios average 36.1, a value considerably higher than that for many marine sediments (7.41, Goldhaber and Kaplan,

Carbon/Ndrogen 0

5

10

15

20

0

5

10

15

20

5

10

15

20

5

10

I5

20

5

10

15

20

10-

20-

1

~ 30-

4050-

4/4/85

6/19185

t2/5/85

3/26/86

6/26/86

Fig. 1. Carbon/nitrogen ratios (atomic) in sediments from the Great Marsh. Sediments were sectioned at 2.6-3.0cm intervals, and C/N data points are plotted at the mid-depth of each section.

316 Totol Sulfur (mg S/g) 5

10

15

20

5

10

15

20

0

5

10

15

20

0

5

~.0

15

20

5

10

~5

20

~o

v E c3

2O

30

40 ¸

50

4/4/85

6/1~/S5

-

t2/5/85

3/26/86

6/26/86

Fig. 2. Depthdistributionof total sulfur in the Great Marsh. Sedimentsweresectionedat 2.6-3.0cm intervals, and concentrations are on a dry weight basis. Data points are plotted at the mid-depth of each section and represent the average total sulfur concentration for each interval. 1974). This elevated ratio is due to the high organic carbon content of the marsh sediment. In the upper 15 cm, sedimentary C/S ratios decrease rapidly and below 30 cm the C/S ratio is relatively constant with an average of 9.31. The increase in sedimentary sulfur with depth (Fig. 2) and the relatively constant C/S ratio below 30 cm suggests that most sulfide incorporation occurs through the oxidation of carbon (via sulfate reduction) in the upper marsh sediment. This conclusion is similar to those of other salt marsh studies (Howarth and Teal, 1979; Lord and Church, 1983; Howes et al., 1984).

Iron monosulfide Iron sulfides which are soluble in weak hydrochloric acid are termed acid volatile sulfides (AVS), and under the conditions of early diagenesis are thought to be primarily amorphous iron sulfide and mackinawite (Goldhaber and Kaplan, 1974). While greigite is typically included in the AVS fraction (e.g. Rickard, 1969; King et al., 1985), the analytical methods used here can discriminate between simple iron monosulfides and greigite. Thus, in this paper the AVS fraction will be referred to as iron monosulfide. Iron monosulfide is the initial product formed by the reaction of hydrogen sulfide ion and Fe(H), and in the presence of elemental sulfur is transformed to greigite (Sweeney and Kaplan, 1973) and pyrite (Berner, 1970). Iron monosulfide may also be lost through oxidation to Fe(II), Fe(III) and elemental sulfur. Since iron monosulfide is thermodynamically unstable under most conditions, it is expected to be a transient intermediate in salt marsh sediments. Indeed, large temporal changes are apparent in the concentration and distribution of FeS in the marsh (Fig. 3). Wintertime FeS concentrations are elevated and closer to the sediment surface than in the summer when oxygen injection via S. alterniflora photosynthesis and desiccation are at a maximum. Thus, the abundance and distribution of FeS appears to be a sensitive integrator of redox

317 FeS (rag S/g)

/

. 0 . I . 2 . 3 . 4 0 . i . 2 . 3 . 4 0

10-

20z: 30-

:

:

:

:

:

:

.

:

:

:

,

:

:

:

:

.1.2.3.4 i , i i

,

l

~

:

I

.1.2.3.4 : : : :

b

~

I

:

~

:

.1.2.3.4 i :

I

/

40-

50-

4/4/85

6/19/85

12/5/85

3/26/85

6/25/86

Fig. 3. Depth distribution of iron monosulfide in the Great Marsh. Sediments were sectioned at 2.6-3.0 cm intervals, and concentrations are on a dry weight basis. Data points are plotted at the mid-depth of each section and represent the average FeS concentration for each interval.

conditions in the marsh. In this manner, the 1986 drought and corresponding oxidation of the upper sediment is apparent in the deeper FeS maximum in March 1986 as compared with April of the previous year. The greatest concentration of FeS occurred in June 1986 (5% of the total sulfur), but generally iron monosulfide accounts for 3% or less of the total sedimentary sulfur. For the five sampling periods, the depth of the FeS maximum coincides with predicted mackinawite saturation using solubility calculations (Berner, 1967; Boulegue et al., 1982; Lord and Church, 1983), and the dissolved iron and hydrogen sulfide data of Luther and Church (1988, this volume).

Greigite Like mackinawite, greigite is not thermodynamically stable under most conditions (Berner, 1967). Sweeney and Kaplan (1973) have implicated a greigite intermediate in the synthesis of framboidal pyrite. Further, the kinetic studies of Rickard (1975) indicate t h a t this synthetic pathway would be slow relative to that of direct precipitation. Using solubility calculations and the identification of framboidal pyrite by electron microscopy in Great Marsh sediments, Lord and Church (1983) postulate that greigite should be present. The data in Fig. 4 confirm their hypothesis, and show that the distribution of greigite is not as temporally variable as that of iron monosulfide (Fig. 3). Indeed, greigite appears to be poised at the interface between the upper sediment which cycles from oxic to anoxic and the deeper, permanently anoxic sediment. The highest concentration of greigite is observed in June 1985 (15% of total sulfur), but otherwise greigite ranges from 5-10% of the total sulfur. The greigite peaks correspond with the deeper pyrite maxima (to be discussed below), and with the exception of the April 1985 core, are at shallower depths than those of FeS (Fig. 3). The positions of the greigite maxima are consistent

318 Greigite (rag S/g)

1

2

0

1

2

2

0

~

2

0

1

2

D

~0-

~~_2°~ ~" # a0.

o

%

40

50

6/19/85

4/4/85

~2/5/85

3/26/86

6/26/86

Fig. 4. Depth distribution of greigite (Fe3S4) in the Great Marsh. Sediments were sectioned at 2.6-3.0 cm intervals, and concentrations are on a dry weight basis. Data points are plotted at the mid-depth of each section and represent the average greigite concentration for each interval.

with solubility predictions (Berner, 1967; Boulegue et al., 1982; Lord and Church, 1983). In this case, porewaters should become saturated with respect to greigite at a lower concentration of dissolved hydrogen sulfide than that for iron monosulfide (i.e. greigite should precipitate closer to the sediment surface).

Elemental sulfur The oxidation of dissolved hydrogen sulfide or solid phase iron sulfides can lead to the formation of elemental sulfur. Further oxidation removes elemental sulfur, as does the formation of greigite and pyrite (Goldhaber and Kaplan, 1974). While some data for elemental sulfur are available for salt marshes (e.g. Elementol Sulfur (mg S/g)

1

2

3

4

5

0

~

2

3

4

5

0

1

2

3

4

5

0

1

2

3

4

5

10~ 2O5

c~ 3040-

50-

4/4/S5

3/26/86

6/26/86

Fig. 5. Depth distribution of elemental sulfur in the Great Marsh. Sediments were sectioned at 2.6-3.0 cm intervals, and concentrations are on a dry weight basis. Data points are plotted at the mid-depth of each section and represents the average S(0) concentration for each interval.

319 King et al., 1985), they are not complete enough for a seasonal description. Lord (1980) estimated the concentration of elemental sulfur in the Great Marsh as the difference between total sulfur and pyrite-sulfur (in the upper 20cm, elemental sulfur ranged from 5-12 mg S g-1). However, the concentration of elemental sulfur was probably overestimated since organic sulfur, FeS and greigite were included in the S(0) fraction. The four profiles of elemental sulfur in Fig. 5 show that this sulfur form is restricted to the seasonally oxic portion of the marsh sediment. Correspondingly, elemental sulfur displays the greatest abundance in the late winter/early spring when oxygen is introduced through photosynthesis as well as through desiccation during low rainfall. In June 1986 S(0) reaches 47% of the total sulfur, with maxima during the other sampling periods ranging from 23-30% of the total sulfur. Qualitatively, these elemental sulfur results are similar to those observed by Troelsen and Jorgensen (1982) in shallow coastal sediments, although the concentrations of S(0) in Great Marsh sediment are approximately a factor of 10 higher. Troelsen and Jorgensen (1982) found elemental sulfur maxima in the sediment's oxidized surface layer, and observed the concentration of S(0) increase as the sediment became more oxidized in the winter. In the Great Marsh the abundance and distribution of elemental sulfur, like iron monosulfide, is coupled to seasonal redox changes, and thus to the formation/ destruction of iron sulfide phases. This conclusion will be examined more thoroughly below.

Pyrite As the thermodynamically stable iron-sulfur compound in marine sediments, the abundance and distribution of pyrite ultimately controls the burial of both elements, as well as many trace metals (Boulegue et al., 1982). With respect to sulfur, the importance of pyrite is clearly seen when comparing

PyrLte (rag S/g) O'

5

t0

15

20

:

:

:

;

5

10

15

20

0

5

10

15

20

5

t0

15

20

0

5 ,

,

~0 ,

15

20

t

> "E

20-

c3

30-

40-

50-

4/4/85

6/19/85

12/5/85

3/2S/B6

6/26/86

Fig. 6. Depth distribution of pyrite (FeS2) in the Great Marsh. Sediments were sectioned at 2.6-3.0 cm intervals, and concentrationsare on a dry weight basis. Data points are plotted at the mid-depth of each section and represent the average pyrite concentration for each interval.

:{20

the total sulfur profiles (Fig. 2) to those in Fig. 6 for pyrite. The variations in total sulfur with depth are primarily due to variations in pyrite (with the one exception being the upper sulfur maximum in 3/86 which is due to elemental sulfur). Below 15 cm pyrite comprises an ave.rage of' 56.9/,) " o, of the total sedim(m tary sulfur for all cores. Like total sulfur, pyrite concentrations increase with depth (aw~rage of 1.96mg S g ~ for 0 acre. and 8.04mg S g ) below 30cm). However in the permanently anoxic sediment below 15 cm, a broad pyrite minimum is observed in all five cores (minimum centered at --~24 cm). In view of the sediment, accumulation rate in the Great Marsh (0.47 cm year ~, Church eta]., 1981) and the slow rates of pyritization at this depth (Lord and Church, 1983), the existence of a pyrite minimum probably reflects a depositional artifact, r a t h e r than a diagenetic effect. On the basis of the sediment accumulation rate, the deep pyrite minimum corresponds to the 1930s when mosquito ditches were dug in the marsh, and thus may be a result of oxidative pyrite loss due to this disturbance. In contrast with the deep sediment (below 15 cm), pyrite in the upper marsh sediment shows more variation with time (Fig. 6). Between April and ,June 1985 a pyrite maximum develops and persists into December. Although these data show a one point maximum, the analyses were repeated in triplicate with the same result. While a duplicate core was not taken in June 1985, analyses of duplicate cores from other periods show excellent agreement (i.e. pyrite concentrations agree within 10%). Thus we are reasonably confident that this pyrite maximum is not an artifact. By March 1986 the shallow pyrite maximum has disappeared, the loss mechanism presumably being oxidation (see Luther et al., 1982 and L ut her and Church, 1988, this volume). Such rapid rates of pyrite formation and oxidation have been reported in other marshes (e.g. Howarth and Teal, 1979), and in the Great Marsh (Lord and Church, 198;~). In June 1986 there is an indication of anot her subsurface pyrite maximum (5.9 8.8cm section). On the basis of the data of L ut her and Church (1988, this volume), this slight subsurfa(> maximum on June 26, 1986 is actually the remnants of a larger pyrite peak which underwent oxidative degradation during the monthly tidal inundation of the marsh, Thus, their results indicate that the June 1985 subsurface pyrite maximum is not an isolated phenomenon, but one which is repeated the following year. In this respect it is important to note that the ,June 19, 1985 core was obtained before the monthly flooding. Furthermore, the pyrite data presented by Luther and Church (1988, this volume) demonstrate that pyrite undergoes even more rapid formation and destruction than the results in Fig. 6 would suggest. As t~ord and Church (1983) have noted, the formation of pyrite in the Great Marsh occurs in two distinct regimes. The upper sediment (0 15 cm) displays large variations in pyrite which reflect rapid rates of formation and destruction. The deeper pyrite maxima (below 15 cm) are indicative of slower pyritization. This observation corresponds to the two pathways of pyritization proposed by Goldhaber and Kaplan (1974), in which single pyrite crystals are formed rapidly through direct precipitation of Fe(II) and polysulfides, and

,, .1

0

~.O-

"E~

20

,2

.3

321 Sulfur (rag/g) .4

.5

1 :

2

B

4

5

:

:

:

;

,

,

,

1 i

2

5

,

~0

5

10

~

x_

30-

40 ~

5(

FeS

S (01

Gre~gite

Pyrite

Organic

S

Fig. 7. Depth distributions of all sulfur species in a sediment core (April 4, 1985)from the Great Marsh. Sediments were sectioned at 2.6-3.0cm intervals, and concentrations are on a dry weight basis. Data points are plotted at the mid-depthof each section and represent the average concentration for each interval. framboidal pyrite is produced by a slower reaction with a greigite intermediate. Rickard (1975) has examined the mechanisms and kinetics of pyrite formation in marine sediments. His work shows t h a t rapid pyritization has a second-order dependence on iron monosulfide surface area, and a first-order dependence on protons, total dissolved sulfide and the surface area of elemental sulfur. The proposed mechanism first entails the dissolution of iron monosulfide and elemental sulfur to form Fe(II) and polysulfides, with the subsequent precipitation of pyrite as follows: Fe 2+ + $5S 2- + HS

-~ FeS2 + $4S 2

+ H÷

In this manner, the upper marsh sediment is well-suited for rapid pyritization since all of the r e a c t a nt s are in abundance. To show this, the various sulfur phases for the April 1985 core are plotted side-by-side in Fig 7. The overlapping greigite and pyrite peaks in the deeper sediment (Fig. 7) tend to confirm the slow pyritization mechanism proposed by Sweeney and Kaplan (1973). In this reaction sequence mackinawite combines with elemental sulfur (in the form of polysulfides) to produce greigite (Berner, 1967): 3FeS + S(0) -~ Fe3S4 Greigite in t u r n produces pyrite, either t hrough dissociation or by further reaction with elemental sulfur (as polysulfides): Fe3S4 ~

2FeS + FeS2

Fe3S4 + 2S(0) --* 3FeS2 Beyond these qualitative observations, a quantitative examination of pyrite formation in the marsh is possible.

~22

A qua ntitatt ve assessmettt of pyritization

The rates of pyritization ill the G r e a t Marsh have been determined by Lord and C h u r c h (1983) using a modeling approach. Ill t h e i r work th~, rate of subsurface pyritization is t a k e n to be equal to the sulfate r e d u c t i o n rate (i.~. the p r o d u c t i o n rate of h y d r o g e n sulfide is assumed to be limiting). In the deeper sediment (z . 15cm) the availability of Fe(II) limits the rate of pyritization (i.e. the rate of f o r m a t i o n of iron monosulfide and greigite is limiting). The data presented here allow the net rates of sulfur phase t r a n s f o r m a t i o n s in t;h~ upper sediment to be calculated. This e x a m i n a t i o n will p a r t i c u l a r l y focus on the p r o d u c t i o n of subsurface pyrite between April and J u n e 1985. As noted above, the rapid f o r m a t i o n of pyrite requires a source of iron and sulfur. In the upper 15 cm of sediment, sulfur for pyritization is available from the inventories of elemental sulfur and iron monosulfide, as well as from in situ sulfate reduction. Correspondingly, the sources of pyritic iron can be iron oxyhydroxides, iron monosulfide and dissolved p o r e w a t e r Fe(II), all of which are found in the upper sediment. F r o m April to J u n e 1985 the growth of a subsurface pyrite maximum (Fig. 6) is accompanied by a decrease in elemental sulfur and FeS c o n c e n t r a t i o n s (Figs. 3 and 5). By i n t e g r a t i n g the concentrations of' S(0) in the upper 15cm of sediment for each core and c o m p u t i n g the c o n c e n t r a t i o n changes over 76 days (April 4 to 19 June), one obtains a loss rate for elemental sulfur of 3.2 ltmol S cm '~ day ~ (assuming a sediment density of 1.~ g c m :~; T.M. Church. personal c o m m u n i c a t i o n , 1986). In a similar manner, the loss of iron monosulfide is computed to be 0.6 itmol S cm :~day 1 Over this same depth range, the rate of pyrite increase is 4.2/~mol S cm :~day With respect to a sulfur mass balance, the gain in pyritic sulfur exceeds the losses of elemental sulfur and FeS-sulfur (i.e. 4.2 versus 3.2 4 0.6 : 3.81~mol S cm :~ (lay ~). This difference must r e p r e s e n t pyritic sulfur t h a t comes from sulfate r e d u c t i o n during the 76 day period. Thus, the estimated sulfate r e d u c t i o n rate for this spring period is 0.41Lmol Scm :~ day ~ This rate is similar to the y e a r l y a v e r a g e c a l c u l a t e d by Lord and C h u r c h (1983) for the same site (0.131Lmol cm :~day '), as well as the sulfate r e d u c t i o n rates for o t h e r salt marshes (e.g. H o w a r t h and Teal, 1979: H o w a r t h and Giblin, 1983). The observed pyritization rate (4.2 ~mol S cm :~day ~) is over all order of m a g n i t u d e h i g h e r than t h a t of Lord and C h u r c h (1983) tot the upper sediment. This a p p a r e n t d i s c r e p a n c y may be explained by the fact t h a t the rate estimated here is fbr a specific time period (76 days), and is not a y e a r l y average as r e p o r t e d by Lord and Church (i.e. the pyritization rate may be slower t h r o u g h the rest of the year). In support of this contention, the rate of pyritization in the upper t5 cm of sediment does slow considerably between J u n e and December 1985 (0.39 l~mol S cm :~day ~. c a l c u l a t e d as above). However, it is also i m p o r t a n t to r e m e m b e r t h a t the p y r i t i z a t i o n rates observed in this study are likely to be u n d e r e s t i m a t e s due to the large temporal changes in pyrite. The data p r e s e n t e d by L u t h e r and C h u r c h (1988, this volume) show large losses of pyrite on time scales of several

323 days. Thus, between our sampling periods pyrite could have formed and been recycled several times. Based on pyrite stoichiometry (1Fe:2S) and the rate of pyritization, an integrated iron oxide loss of 2.1#mol F e c m -3 day 1 would be anticipated between April 4 and June 19, 1985. Data for iron oxides are presented in Table I, and in a manner similar to that for sulfur phases, an iron loss of 2.5gmol Fecm 3 day ' is computed. While this value is close to stoichiometric predictions, the contribution of iron from monosulfide must also be considered. If the monosulfide loss is included, the total loss of reactive iron from phases other than pyrite is 3.1 #mol Fe cm 3 day 1. Since the loss of iron from reactive oxides and monosulfide is greater than the increase of pyritic iron, a gain in porewater iron of 1.0~mol cm 3 day ' would be expected between April and June 1985. Luther and Church (1988, this volume) report elevated porewater iron concentrations in June 1985, but data are not available for April 1985. Since dissolved ions can migrate by diffusion and advection, an exact estimate of the porewater iron increase would be difficult. Overall, changes in sulfur and iron inventories of the upper sediments appear to match the rapid pyritization schemes postulated by other workers (e.g., Goldhaber and Kaplan, 1974; Rickard, 1975). While the focus here has been the formation of pyrite in salt marsh sediments, its removal through oxidation is also apparent in our data. This aspect of the sulfur cycle is covered in the companion paper by Luther and Church (1988, this volume). Below 15 cm pyrite accumulates more slowly, and the relationship between greigite and pyrite in this zone was noted earlier. The greigite data in Fig. 4 allow this relationship to be quantitatively examined by comparing the rate of greigite loss to that of pyrite formation. The average concentration of iron monosulfide is ~ 10% that of greigite in sediments below 15 cm. Thus for the calculations below, only greigite will be considered in the formation of pyrite. The rate equation for slow pyrite formation given by Lord and Church (1983) depends on the concentrations of protons, dissolved sulfide and refractory iron. Using the iron data in Table I, and the hydrogen sulfide and pH data from Luther and Church (1988, this volume), the rate of pyritization below 15 cm is calculated to be 0.96 pmol S cm 3 year-1. To estimate the rate of greigite loss, the diagenetic model of Burdige and Gieskes (1983) can be applied to the greigite data where concentrations exponentially decrease with depth (ca. below 15cm, refer to Fig. 4). In this application, it is assumed that the only processes which affect greigite below 15 cm are burial via sedimentation and loss through pyritization. Further, the loss of greigite via pyritization is assumed to be first order with respect to greigite (Rickard, 1975). At steady state these processes can be described using the following diagenetic equation -w(dC/dz)

-

kredC

= 0

(1)

where kred is the first-order removal rate constant (year-I), w is the sedimentation rate (cm year- 1), C is the concentration of greigite (mg S g- 1) and z is the

324 TABLE l (]real, M a r s h sulfi~r d a t a

Depth cm)

Org. C

Total N

(%)

(%

Total S [~i)

FeS (mgSg

0.86 1.02 1.38 0.67 11.57 0.39 0-37 11.42 0.38 0.29 0.32 0.30

0.65 0.97 101 1.18 1.32 1.53 1.0~ 0.88 1,06 I. 12 1.32 1.49

ND 0.14 0.29 NA 0.35 0.24 I).24 0.20 l) 14 NA 0.04 0.08

003 0.02 0.13 002 0.06 1.19 13.3I 0.25 0 24 NA 0.39 NA

0.40 2.91 2.67 2.69 3.93 0.61 1/.44 056 0.68 NA 0.64 NA

1.57 /,65 3.32 3.98 4.93 8.23 6.84 5.04 6.51 7.17 7.30 8.85

13.3 532 N ,~, 2 01 i I0 1.4!3 225 i.!t9 I. 2.1 i 48 2.18 2 19

0,76 I)82 0.77 0.41 0.65 I) 58 0.36 0.38 0.57 0.41 0.39 0.30 0.24 0.21

0.53 (I.70 0.73 1.72 0.89 1,16 1.64 0.92 1,16 0.94 1.32 L42 0.75 6.91

0.03 0.03 0.05 0.03 0.05 0.07 0.05 0.14 0.25 0.03 0.02 0.05 0.06 0. ll3

0.02 0.07 0.08 0.01 0.01 1.76 086 0.71 007 NA 0.41 NA NA NA

0.34 0,93 0.98 0.97 2.44 2.67 0.57 071 0.7:3 NA 0.44 NA NA NA

209 t).85 1.07 12.6 1.86 2.55 12.2 6.34 473 5.55 921 10.3 5.07 7,75

11.~) 2.90 1.70 3,70 1.80 11.80 2.20 I 70 1.20 1.40 1..90 3.(}0 1.~1 1N5

1.23 1.15 1.94 1.10 0.95 1.42 1.64 1.17 0.79 0.71 1.43 1.31) 1.23 1,07 148

0.003 0.06 0-05 0.01 0.02 0.02 0.02 0.04 0.15 0.08 0.04 0.04 0.02 0.02 0.02

0.07 11.05 0,03 0.17 0.06 11.17 11,77 lag 0.44 NA (}.42 NA NA NA NA

NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

3.06 2,28 9.83 3.90 3.08 6.33 9.59 3.97 2.28 :1.25 7..90 8.28 8.61 9,98 5.63

3,90 140 150 I50 1.10 0.90 1.40 1.70 1 ,10 1.10 1.90 2,60 1.80 1.60 2.10

0.55 0,61 1.70 1.21 0.95 1.58 1.50 0.64 1.35 1.34 1.98 1.70 1.22 1.13

ND NI) 0.03 0.02 ND 0,09 0.05 0.20 0.18 0.10 0.01 0.02 0.01 0.04

0.02 0.02 0.06 0.06 0.(12 1.35 0.51 0.50 027 0.51 0 32 0.32 0.88 0.70

0.34 0.77 4.71 3.42 1.06 0.46 0.35 0.97 0.94 NA 076 NA NA NA

1.12 0.51 1.75 2.64 2.9B 11.2 10.2 :5.45 4.~i 4.70 12.1 9.91 6.8(5 6.16

7.83 3.30 2.91 2.07 1.38 2.00 1.95 1.74 1.79 1.17 2.06 3.01 2.02 1.84

Sampling data: 4 April 1985 O 2.9 9.43 2.9 5.8 11.7 5.8 8.6 12.7 8 . 6 11,5 9.44 11.5 14.4 7.62 14.4 173 5.57 17.:~ 20.2 5.~2 20.2 23.0 6,27 23 .(g 25.9 6.16 25.9 28.8 4.11 28.8 232.3 4.48 32.2, 35.7 4.32 Sampling date: 19 J u n e 1985 (4 2,7 8.08 2.7 5.5 9.54 5.5 8.2 9.15 8.2 10.9 5.45 10.9~ 13.7 8.05 13.7 16.4 7.43 16.4 19.1 5.26 19.1 21.8 5.76 21.8 24.6 8.94 24.6 2 7 3 6.00 27.3 30.0 5.61 30.0 34.4 4.19 34.4~ 38.8 3.40 38.842.0 3.20

Sampling date: 5 December 1985 0 3.2 7.70 0.61 3.2 6 4 7, [6 11.83 6.4 9.6 6.04 0.75 9.6 12.8 5.10 0.40 52.& 16.1 6.56 0.56 16.1 19.3 6.08 0.52 19.3 22.5 5.23 0.40 22.5 25.7 6.76 0.49 25.7 28.9 5.08 0.40 28.9 32.1 4.86 0.43 32.1 35.3 4,94 0.43 35.3 38.6 330 0.;51 38,o 41.7 3.33 0,311 41.7 48.3 312 0.31 48.1 54.5 3.60 0.41 Sampling date: 26 M a r c h 1986 0 31/ 7.78 0.73 3.0 6.0 8.52 0.80 6.0 9.0 10.1 0.96 9.0 12.0 9.11 0.79 12.0 15.0 776 0.62 15.(1 18.0 6.30 0.51 18.0 21.0 519 0.38 21.0 24.0 4.74 0.32 24.0 27.0 8.88 0.56 27.0 30.0 7.14 0.48 30.0 ;53.0 4.65 0.33 33.0- 36.0 5.01 037 36.0, 39.0 4.21 0.31 39,1t 45.0 3,71 0,26

11

Fe:~S 4 (mgSg

i}

S(0I (mgSg

t)

FeS 2 (mgSg

I

I:'~,. ox a (mgFeg

325 TABLE l (Continued) Depth (cm)

Org. C (%)

Sampling date: 26 June 1986 0-2.9 7.99 2.9-5.9 8.94 5.9-8.8 6.63 8.9-11.7 8.31 11.7-14.7 6.09 14.7-17.6 4.83 17.6-20.5 4.82 20.5-23.4 5.54 23.4-26.4 6.90 26.4-29.3 5.34 29.3-31.6 5.98 31.6-34.0 6.12

Total N (%)

Total S (%)

FeS (mgSg 1)

Fe3S 4 (ragSg 1)

S(0) (mgSg 1)

FeS 2 ( m g Sg 1)

0.76 0.80 0,54 0,71 0,55 0.37 0.33 0,41 0.44 0.36 0.42 0.42

0.77 0.86 0.98 0.92 1.04 1.54 0.81 0.65 0,92 1.76 1.61 1.16

0.004 0.004 0.01 0.01 0.01 0.02 0.10 0.35 0.14 0.004 ND ND

0.03 0.03 0.02 0.03 0.13 0.78 0.41 0.19 0.06 0.05 0.07 0.05

3.67 1.06 1.36 1.22 0.53 0.64 0.38 0.57 0.76 NA 0.42 NA

1.97 1.11 2.49 0.88 2.14 8.05 4.31 1.93 4.44 10.8 8.68 5.45

Fe, ox a (rngFeg-1)

5.66 2.93 1.59 1.82 1.08 1.87 1.79 1.75 1.25 1.62 2.01 0.87

All values on a dryweight basis; NA, not analyzed; ND, not detectable. aReactive iron oxides.

depth (cm, positive downward). The solution to this equation is

C = Coe-s(~-15)

(2)

where C = Co at z = 15 cm (i.e. the boundary condition), C = 0 as z goes to infinity, and B is equal to kred/w. Employing a modified form of eq. 2, a plot of ln[Fe3S4) versus depth yields a straight line whose slope is equal to kred/w. Using greigite data below 15 cm from all but the 12/85 core, such a best fit line gives Co = 0.8mg S g-l, B = 0.107 cm -1, and a linear correlation coefficient of 0.61 for n = 24. With a sedimentation rate of 0.47 cm year 1(Church et al., 1981), kred then equals 0.051 year 1. Integrating the greigite removal rate (= kredC) from 15--30cm results in a depth averaged greigite removal rate of 1.14gmol S cm -3 year -1. Lord and Church (1983) report that pyrite in the deeper marsh sediment has a framboidal texture, and since the rate of greigite loss is roughly equivalent to the rate of pyrite formation, these results support the laboratory studies of Sweeney and Kaplan (1973). Specifically, a tight coupling between greigite and framboidal pyrite would be expected during slow pyritization. CONCLUSIONS The use of newly developed analytical techniques has allowed a detailed examination of sedimentary sulfur cycling in Delaware's Great Marsh. Within the marsh two well-resolved zones of pyritization are observed. In the upper sediment where seasonal redox cycling takes place, the co-occurrence of elemental sulfur, iron monosulfide, low pH and reactive iron, lead to high rates of pyrite formation during certain times of the year. The speciation data for sedimentary sulfur in this zone support the rapid pyritization mechanism proposed by Rickard (1975). In the deeper sediment where pyrite formation

326 slows, d i a g e n e t i c m o d e l i n g i l l u s t r a t e s the i m p o r t a n c e of a greigite interm e d i a t e d u r i n g pyritization. While the focus of this p a p e r has been on i n o r g a n i c sulfur specms, the a m o u n t a n d b e h a v i o r of o r g a n i c sulfur can be a p p r o x i m a t e d by the difference between t o t a l sulfur and the sum of the i n o r g a n i c sulfur fractions. In Fig. 7 a plot of o r g a n i c sulfur ibr April 1985 is shown. U n l i k e the i n o r g a n i c forms, o r g a n i c sulfur shows little v a r i a t i o n with depth. This s u g g e s t s t h a t o r g a n i c sulfur in the m a r s h is r e l a t i v e l y r e s i s t a n t to d e g r a d a t i o n . H o w e v e r , o r g a n i c sulfur can consist of m a n y c o m p o u n d s (e.g. sulthte esters, sulfur a m i n o acids), and only a gross c h a r a c t e r i z a t i o n is provided by these o p e r a t i o n a l l y defined results. Detailed a n a l y s e s of the a c t u a l o r g a n i c sulfur species would result in a c l e a r e r u n d e r s t a n d i n g of o r g a n i c sulfur cycling. T h e G r e a t M a r s h findings are also a p p l i c a b l e to o t h e r c o a s t a l a n o x i c m a r i n e s y s t e m s such as F r a m v a r e n Fjord. While the F r a m v a r e n s e d i m e n t s are perm a n e n t l y anoxic, e l e m e n t a l sulfur and iron monosulfide, both of w h i c h are required for rapid pyritization, could be supplied by p a r t i c l e s p r o d u c e d n e a r the o x i c / a n o x i c i n t e r f a c e within the w a t e r c o l u m n (e.g. see J a c o b s et al., 1985). Similarly, the e x i s t e n c e of f r a m b o i d a l pyrite in F r a m v a r e n suspended p a r t i c l e s (Skei, 1988, this volume) suggests t h a t greigite is also produced in the w a t e r column. In anoxic sediments, p y r i t e f o r m a t i o n r e q u i r e s e l e c t r o n a c c e p t o r s (oxidants) such as Fe(III), e l e m e n t a l sulfur and greigite. Thus, an i m p o r t a n t difference in sulfur d i a g e n e s i s b e t w e e n salt m a r s h and fjord s e d i m e n t s is the s o u r c e of these oxidants. In a salt m a r s h sediment, o x y g e n from p h o t o s y n t h e s i s is a c o m m o n electron acceptor, w h e r e a s in s e d i m e n t s o v e r l a i n by a n a n o x i c w a t e r column, o x i d a n t s (e.g. iron oxides, e l e m e n t a l sulfur) c a n be supplied via the flux of p a r t i c l e s from oxic s u r f a c e waters. H o w e v e r , the a c t u a l p y r i t i z a t i o n m e c h a n i s m s are likely to be s i m i l a r in b o t h systems. ACKNOWLEDGMENTS We t h a n k o u r colleagues, T. C h u r c h a n d G. L u t h e r ( U n i v e r s i t y of Delaware), for the use of t h e i r field site and l a b o r a t o r y in Lewes, and J. S c u d l a r k for his i n v a l u a b l e a s s i s t a n c e in the field. We also t h a n k D. B u r d i g e for his advice on d i a g e n e t i c m o d e l i n g a n d for m a n u s c r i p t review, and R. K l u c k h o h n , C. K r a h f o r s t and J. B u s a for field and l a b o r a t o r y assistance. REFERENCES Berner, R.A., 1967. Thermodynamic stability of sedimentary iron sulfides. Am. J. Sci., 265:773 785. Berner, R.A., 1970. Sedimentary pyrite formation. Am. J. Sci., 268:1 23. Boulegue, J., Lord, C.J., III and Church, T.M., 1982. Sulfur speciation and associated trace metals (Fe, Cu) in pore waters of Great Marsh, Delaware. Geochim. Cosmochim. Acta, 46: 453464. Burdige, D.J. and Gieskes, J.M., 1983. A pore water/solid phase diagenetic model for manganese in marine sediments. Am. J. Sci., 283: 2~47. Church, T.M., Lord, C.J., II! and Somayajulu, B.L.K., 1981. Uranium, thorium and lead nuclides in a Delaware salt marsh sediment. Estuarine Coastal Shelf Sci., 13: 267-275. Cutter, G.A. and Oatts, T.J., 1987. Determination of dissolved sulfide and sedimentary sulfur speciation using gas chromatographyrphotoionization detection. Anal. Chem., 59: 717-721.

327 Goldhaber, M.B. and Kaplan, I.R., 1974. The sulfur cycle. In: E.D. Goldberg (Editor), The Sea, Vol. 5. Wiley-Interscience, New York, pp. 56~655. Howarth, R.W., 1984. The ecological significance of sulfur in the energy dynamics of salt marsh and coastal marine sediments. Biogeochemistry, 1: 5-27. Howarth, R.W. and Teal, J.M., 1979. Sulfate reduction in a New England salt marsh. Limnol. Oceanogr., 24: 999-1013. Howarth, R.W. and Giblin, A.E., 1983. Sulfate reduction in salt marshes at Sapelo Island, Georgia. Limnol. Oceanogr., 28: 7(~82. Howes, B.L., Dacey, J.H. and King, G.M., 1984. Carbon flow through oxygen and sulfate reduction pathways in salt marsh sediments. Limnol. Oceanogr., 29: 1037-1051. Jacobs, L., Emerson, S. and Skei, J., 1985. Partitioning and transport of metals across the O2/H2S interface in a permanently anoxic basin: Framvaren Fjord, Norway. Geochim. Cosmochim. Acta, 49: 1433-1444. Jorgensen, B.B., Fossing, H. and Thode-Anderson, S., 1984. Radio-tracer studies of pyrite and elemental sulfur formation in coastal sediments. Trans. Am. Geophys. Union, 65: 906. King, G.M., Howes, B.L. and Dacey, J.W.H., 1985. Short-term endproducts of sulfate reduction in a salt marsh: formation of acid volatile sulfides, elemental sulfur and pyrite. Geochim. Cosmochim. Acta, 49:1561 1566. Lord, C.J., III, 1980. The chemistry and cycling of iron, r~anganese and sulfur in salt marsh sediments. Ph.D. Dissertation, University of Delaware. Lord, C.J., III and Church, T.M., 1983. The geochemistry of salt marshes: sedimentary ion diffusion, sulfate reduction and pyritization. Geochim. Cosmochim. Acta, 47: 1381-1391. Luther, G.W., III and Church, T.M., 1988. Seasonal cycling of sulfur and iron in porewaters of a Delaware salt marsh. Mar. Chem., 23: 295-309. Luther, G.W., III, Giblin, A.E., Howarth, R.W. and Ryans, R.A., 1982. Pyrite and oxidized iron mineral phases formed from pyrite oxidation in salt marsh and estuarine sediments. Geochim. Cosmochim. Acta, 46: 2671-2676. Luther, G.W., III, Giblin, A.E. and Varsolona, R., 1985. Polarographic analysis of sulfur species in marine porewaters. Limnol. Oceanogr., 30: 727--736. Reeburg, W.S., 1967. An improved interstitial water sampler. Limnol. Oceanogr., 12: 163-165. Rickard, D.T., 1969. The chemistry of iron sulfide formation at low temperatures. Stockholm Contrib. Geol., 20: 67-95. Rickard, D.T., 1975. Kinetics and mechanism of pyrite formation at low temperatures. Am. J. Sci., 275: 636-652. Skei, J., 1988. Formation of framboidal iron sulfide in the water of a permanently anoxic fjord Framvaren, South Norway. Mar. Chem., 23: 345-352. Sweeney, R.E. and Kaplan, I.R., 1973. Pyrite framboid formation: laboratory synthesis and marine sediments. Econ. Geol., 68: 618~34. Tessier, A., Campbell, P.G.C. and Bisson, M., 1979. Sequential extraction procedure for the speciation of particulate trace metals. Anal. Chem., 51:844 851. Troelsen, H. and Jorgensen, B.B., 1982. Seasonal dynamics of elemental sulfur in two coastal sediments. Estuarine Coastal Shelf Sci., 15: 255-266. Valiela, I., Teal, J.M. and Persson, N.Y., 1976. Production and dynamics of experimentally enriched salt marsh vegetation: below ground biomass. Limnol. Oceanogr., 21: 245-252. -

-