| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Natural Resource Sci. and Landscape Architecture, Univ. of Maryland, College Park, MD 20742 USA
pedon{at}dnamail.com
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
These newly formed submerged-upland tidal marsh soils are characterized by properties that reflect this shift in pedogenesis and a new set of soil-forming conditions. Among the acquired properties reflecting such a change include the formation of organic horizons and the accumulation of S in both organic and mineral forms. Pyrite (FeS2) is the principal mineral phase and forms through the process of sulfidization. Factors necessary for the formation of pyrite have been described by others (Rickard, 1973; Pons et al., 1982) and recently reviewed by Rabenhorst and James (1992) and are ideally met in coastal marsh soils.
If aerobic conditions are induced in sulfide-bearing soils, acid-sulfate weathering can occur. The drainage or dredging of sulfidic soil materials has commonly caused oxidation of pyrite and the development of extreme acidity (Van Breemen, 1982). Two moles of sulfuric acid are produced from each mole of pyrite if Fe is completely oxidized and hydrolyzed (Nordstrom, 1982). The potential for sulfide-bearing soils to become acidic when oxidized has been recognized through the use of Sulf and Sulfi great groups in U.S. soil taxonomy (Soil Survey Staff, 1998). Not only can acid-sulfate weathering lead to extreme acidity in soils, but movement of acidity through drainage waters can cause acidification of ditches and tributaries of estuaries with significant environmental effects (Soukup and Portnoy, 1986).
Pedogenesis in submerged-upland tidal marsh systems proceeds through the transformation of Ultisols to Alfisols and eventually to Histosols as the soils become increasingly inundated (Stolt and Rabenhorst, 1991). In a similar setting in South Carolina, Brinson et al. (1995) describe ecosystem changes in response to sea-level rise, and developed a conceptual model for the transformation of a terrestrial forest into a heterotrophic benthic system. Areas of transgressing coastlines in the lower Chesapeake Bay have recently been described as ideal for quantifying chronological development in marsh pedogenesis by using the chronosequence/chrono-continuum, which is defined by relating marsh age to elevation of the submerged landscape (Rabenhorst, 1997). This is possible because the mineral soil profile beneath the organic horizons in submerged-upland marshes provides a stable elevational control. While the upper organic horizons possess properties of the newly formed marsh soil, the underlying mineral horizons retain properties of the original upland soils such as argillic horizons, high bulk density, low n-value, and high bearing strength. This stands in contrast to estuarine-type marshes that possess mineral substrata that have accumulated under a submerged environment, causing them to have high n-value and low bearing strength (Darmody and Foss, 1979). Thus, where submerged-upland marshes maintain a stable elevation at the mineral surface, the sediments in estuarine marshes are easily consolidated and can become compressed with time.
Although estimates of future sea-level rise vary greatly, many investigators believe that the sea level may rise 1 m by the year 2100 (Revelle, 1983; Thomas, 1986; Robin, 1986). Even modest estimates of the effects of warming due to greenhouse gases presently in the atmosphere suggest a 0.33-m sea-level rise by 2100 (IPCC, 1990). Hoffman and Titus (1983) identified major factors influencing rates of sea-level rise, from which they derived four projected worldwide scenarios for the next century ranging from conservative (0.6 m) to high (3.4 m). While it is unlikely that marsh vegetation could keep pace with the higher rates of sea-level rise, marsh accretion under slower scenarios could dramatically increase the quantities of sulfur accumulating in coastal marsh soils.
The objective of this study is to develop empirical models that describe S accumulation in submerged-upland tidal marsh soils of Chesapeake Bay. These models can be used to estimate the future sequestration of various forms of S under different scenarios of sea-level rise.
|
Methods and materials |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
|
|
5 cm) were used. Upon extraction, the tubes were filled with water and sealed to minimize oxidation during transport. To collect samples from organic horizons at depths >125 cm, a McCauley sampler was employed, which caused negligible vertical compaction. In the field, samples below 125 cm were sectioned into 25-cm increments (unless horizon morphology suggested otherwise), placed into plastic bags, sparged with N gas, and frozen using dry ice. Upon arrival at the laboratory, soil cores and bagged samples were frozen at -15°C until they were analyzed. A McCauley sampler was used to collect separate cores intended for 210Pb dating. Recognizing that microtopographic variability is common to marsh systems, three cores representing different physiographic positions and vegetative cover were collected from each marsh for 210Pb dating. In addition, five basal peat samples for 14C dating were collected from each marsh immediately above the dense, low n-value buried mineral soil using a McCauley sampler.
Laboratory Procedures
The most recent marsh accretion rates (for the past 150 yr) were determined using 210Pb following standard methods (Flynn, 1968; Benoit and Hemond, 1988). Cores were sectioned into 3-cm increments and weighed to determine bulk density and moisture content. Accretion rates before the last few hundred years were estimated using 14C dating of five basal peat samples from each transect. The 14C analyses were carried out by Beta Analytic (Miami, FL).
In preparation for S fractionation, frozen cores were extruded and sectioned into 25-cm increments (unless morphology suggested other horizonation). Sample preparation was done inside a glove bag under a N atmosphere to avoid oxidation. The total weight of the wet samples was recorded to determine the bulk density, and in this regard the small amount of compaction was assumed to have been distributed linearly along cores. A subsample of at least 0.5 kg of the wet sample was dried to determine the gravimetric moisture content. The peat subsamples were oven-dried at 60°C (
48 h) and were then ground using a stainless-steel plant mill. Samples to be analyzed for total S and total C were homogenized and ground to pass a 60-mesh sieve. Total S and total C analyses were done using high-temperature combustion with infrared detectors (Searle, 1968; Tabatabai and Bremner, 1970).
After centrifuging the wet soil samples at 3500 rpm for 30 min, the SO4 concentration in the supernatant was determined by ion chromatography. The centrifuged soil sample was then used to determine acid-volatile S, chromium-reducible S, and elemental S. Quantifying the various S species was carried out using a modified version of a Johnson and Nishita digestion–distillation apparatus (Johnson and Nishita, 1952). In this method, different S fractions were specifically reduced to H2S gas. The H2S gas was transported using an N2 carrier and trapped in a sulfide antioxidant buffer (SAOBII) (Cornwell and Morse, 1987). The content of S in the SAOBII was determined by potentiometric titration with Pb(ClO4)2, using a AgS electrode and double-junction reference electrode for end-point detection. For acid-volatile S, samples were digested following addition of 6 M HCl (Cornwell and Morse, 1987). The chromium-reducible S (which included both pyrite and elemental S) was determined on the same sample by digesting a second time after adding ethanol and acidic CrCl2 solution, and boiling (Canfield et al., 1986). A separate sample was used to determine elemental S following extraction with acetone (Wieder et al., 1985). Organic S was determined by subtracting chromium-reducible S, acid-volatile S, and sulfate S from total S.
|
Results and discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
The marsh portion of the two transects was mapped in a consociation named for the Honga series (Terric Sulfihemist). The slope of the mineral soil surface from Site 1 to Site 13 at Cedar Creek transect was 0.1% and was as steep or steeper beyond Site 26, but was essentially level between Site 14 and Site 26. Along the Hell Hook transect, the slope of the mineral soil surface from Site 1 to Site 8 was 0.2% and was steeper but variable beyond Site 14, but remained nearly level between Site 8 and Site 13. Along Cedar Creek Marsh, the organic horizons were underlain by submerged mineral soil and were generally uniform in thickness between Site 12 and Site 26 (averaging 23 cm), due to the level topography of the submerged surface. From Site 26 to Site 33, the surface elevation of the marsh tends to rise gradually towards the stream edge. This unusual shape of the surface topography could be attributed to a higher rate of vertical accretion near the streamside marsh than that of the inland marsh, which may be related to a natural levee effect (Hatton et al., 1983; Kearney et al., 1994). In the Hell Hook Marsh between Site 7 and Site 16, the organic horizons were also underlain by a submerged mineral soil; but beyond Site 17 (to Site 19), the organic horizons were underlain by high n-value estuarine sediments (n > 1 as estimated in the field). The micro-scale undulations of the marsh surfaces, which can be observed in Fig. 1 and 2, could be attributed to differences in plant species and vegetation density causing uneven distribution of organic matter and inorganic sediment accumulation. In situ translocation of sediment by invertebrates and other sorts of faunal pedoturbation may also contribute to marsh microtopography.
Organic horizons of the 26 sampled pedons at Hell Hook Marsh and Cedar Creek Marsh had no recognizable mineral-sediment layers, and were all underlain by dense, low n-value mineral soils. The upper 25 cm of organic horizons was generally fibric, whereas the subsurface and bottom tiers were generally hemic material.
Depth Distribution of Various Sulfur Species
The depth distribution of various S species in the representative pedons (Table 1)
reflects the influence of factors controlling S processes in marsh systems as well as the physiographic position. The absence of data for elemental S and acid-volatile S indicates that these phases were not detected in the 26 pedons. The concentrations of most S species (organic S, and chromium-reducible S) were lower in the surface horizon, probably due to higher redox potentials and the periodic flushing by meteoric water. Sulfate S was generally the form present in lowest concentrations in all pedons. Organic S was the dominant S fraction in the organic horizons. In Site HH18, where high n-value mineral sediment of estuarine origin lay beneath the organic materials, chromium-reducible S rather than organic S was the dominant S fraction in the mineral sediment, presumably because of the higher levels of sorbed reactive Fe and relatively low organic-C content in these sediments. In pedons where the organic horizon was underlain by submerged-upland soils, all S species decreased abruptly at the organic–mineral contact into these low n-value materials, due to the low organic-matter content and the high density of the submerged-upland soil (Rickard, 1973; Pons et al., 1982; and Griffin and Rabenhorst, 1989; Rabenhorst and Haering, 1989). In Pedon HH19, the depth distribution of organic S and chromium-reducible S was somewhat irregular, reflecting the dominant impact of fluvial processes in this section of the marsh.
|
(C/S ratio is 20:1). Freney (1986) indicated that about 90% of the organic S compounds found in plants and microorganisms are C-bonded S. Marsh plants and microorganisms are considered to be the major source of organic C and organic S in marshes. Soluble inorganic S may react with organic compounds to form organic S (Howarth et al., 1992).
The Rate of Sea-Level Rise and Age–Elevation Relationship
The apparent sea-level rise for the most recent period (the last 150 yr) was estimated using 210Pb dating techniques. Based on the excess 210Pb activity, the rate of marsh accretion (or apparent sea-level rise) at Cedar Creek Marsh ranged between 3.2 and 1.4 mm yr-1, averaging 2.5 ± 0.9 mm yr-1, while at Hell Hook Marsh the rate of marsh accretion (and apparent sea-level rise) ranged between 3.2 and 1.5 mm yr-1, averaging 2.2 ± 0.4 mm yr-1. In spite of differences in methods of determination and time intervals over which rates have been integrated, these data are reasonably consistent with observations reported by others. Hicks et al. (1983) have indicated that based on tidal records for the last 40 yr, sea-level rise in the Chesapeake Bay has ranged between 2.5 and 3.6 mm yr-1, depending on the location. Kearney and Stevenson (1991) have reported that modern rates of sea-level rise at Baltimore varied between 3.0 and 3.9 mm yr-1 (3.17 ± 0.13 mm yr-1) since 1900.
The long-term rate of sea-level rise (or vertical accretion) over the last one or two millennia was determined using the 14C dates of basal peat samples. The peat samples were collected from organic horizons immediately above the low n-value, high-density, submerged mineral soil surface in each marsh. The 14C age was regressed against the elevation of the basal peat, and the slope of the regression line was used to determine the rate of sea-level rise at each site. The long-term rate of sea-level rise before the last century or two at Hell Hook Marsh and Cedar Creek Marsh was found to be
and
, respectively. In a paleobotanical study, Kearney (1996) reported a very slow overall rate of sea-level rise (0.56 mm yr-1) during the last millennium in the Chesapeake Bay area. Others have concluded that the present rate of relative sea-level rise along the U.S. Atlantic coast is as much as 3 to 4 times higher than the long-term trend during the last several thousand years (Kraft et al., 1987).
While the apparent acceleration of sea-level rise in the last century has gained considerable attention (Gornitz et al., 1982), the use of 14C dating to infer rates of sea-level rise can lead to erroneous conclusions if the elevation of 14C samples is not stable over time. Compression from the weight of overlying sediments and organic matter decomposition may cause autocompaction of the deeper layers of organic matter and high n-value sediments over time, with an accompanying increase in bulk density (Kearney and Ward, 1986) that can lead to erroneously low rates of vertical accretion (or sea-level rise) (Craft and Richardson, 1998). This is most likely to be problematic when samples selected for 14C dating are themselves underlain by organic sediments or high n-value mineral sediments. In such instances, the elevation of the sample is not stationary over time but is subject to lowering due to autocompaction. In contrast, the samples selected for 14C dating in this study were immediately underlain by a dense, low n-value, submerged mineral soil surface. The high bearing capacity of the submerged geomorphic surface has provided a stable elevational control, relative to the sea-level datum. Therefore, the calculated rates of sea-level rise based on regressing the radiocarbon dates vs. elevation of the basal peat samples are not appreciably influenced by autocompaction.
Linking the topographic continuum underlying the marsh portion of Hell Hook and Cedar Creek to the sea-level rise history of the Chesapeake Bay, the age–elevation relationship can be established at each site. The age of the marsh for the past 150 yr was determined using 210Pb data. In this regard, the elevation of the submerged surface at a given point was subtracted from the elevation at the marsh edge, and the result was divided by the average rate of sea-level rise. In this way, the effect of the micro-topography of the marsh on age determination was minimized. For pedons >150 yr old, the age of the marsh was assessed using 14C data. In this regard, the age of the marsh was determined using the derived regression relationship and the elevation of the submerged surface at a given point. The derived marsh ages and the corresponding elevation of the submerged surfaces were used to construct the age–elevation relationship (Fig. 3) . Because the elevation of the submerged surfaces was measured relative to the sea-level datum of 1929, the land surface does not reach mean sea level at time zero (Fig. 3).
|
Pyrite Predictive Model
During the development of submerged-upland tidal marsh soils, pyrite accumulates as a result of sulfidization (Rabenhorst and James, 1992). Conditions necessary for SO4 reduction (organic C source, SO2-4 source, low Eh, SO2-4-reducing bacteria) have been described by Rickard (1973) and Pons et al. (1982). Goldhaber and Kaplan (1982) have indicated that dissolved S concentrations <320 mg L-1 may limit the rate of SO4 reduction, although work by Haering (1986) suggests that S accumulation may not be limited until SO4 levels reach an order of magnitude lower. Thus, soluble SO4, organic C, and pyrite contents (Table 1) suggest a favorable environment for SO4 reduction. Work by others in this general vicinity (Griffin and Rabenhorst, 1989; Rabenhorst and Haering, 1989; Rabenhorst and James, 1992) indicates that within the organic-rich horizons of these types of marshes, available reactive Fe is generally the factor that limits sulfidization and these horizons generally exhibit a high degree of pyritization (most of the reactive Fe is tied up in pyrite). Although Fe was not measured directly in these pedons, the favorable environment for SO4 reduction and the relatively high ratio of organic S to pyrite S (Table 1) would seem to confirm that Fe inputs are generally low and probably the factor that limits pyrite formation.
In these systems, two main sources of Fe for sulfidization are postulated: (i) Fe that is sorbed to sediments carried in by the tidal water, and (ii) Fe in the submerged mineral portion of the soil. The amount of Fe brought with the suspended sediment in tidal water is determined by the nature of the sediment source and the sediment load. The sediment load is a function of tidal energy (Stevenson et al., 1986), which is largely controlled by tidal range. The deposition of the sediment load within the marsh may be controlled by vegetation factors such as plant morphology and density (Gleason et al., 1979) or by hydrologic factors such as the degree of development and integration of the tidal creek network and distance from the stream (Leonard et al., 1995). Previous work indicates that sediment-borne Fe inputs in extensive submerged-upland marshes may be lower than in marshes adjacent to estuarine streams (Rabenhorst and Haering, 1989; Griffin and Rabenhorst, 1989). Nevertheless, one would expect sediment-borne Fe inputs, however small, to continually enter and accumulate in coastal marsh ecosystems. Both characterization data for pedons representing different physiographic positions (Table 1) and field observations indicate that essentially all of the horizons above the submerged mineral surface are organic horizons and thus the mineral component of these horizons is limited.
Because the thickness of organic horizons tends to increase with marsh age as sea-level rises, and because pyrite accumulation in these systems tends to be limited by inputs of Fe that mainly enter gradually or in random storm events sorbed to sediment, we hypothesized that the volumetric pyrite content of organic horizons (g m-3 averaged over the thickness of the organic horizon) should be independent of marsh age. To test this postulation, the amount of pyrite sequestered in the organic horizons was normalized or adjusted for the thickness of the organic horizons and then evaluated according to the age of the pedon. Volumetric pyrite content showed no specific pattern over time (marsh age) (data are not shown). Therefore, we concluded that pyrite accumulation in the organic horizons is driven primarily by sea-level rise, and the rate of pyrite sequestration in the organic horizons could be expected to be controlled by the rate of sea-level rise. Factors controlling the availability of Fe, which may vary with marsh habitat and with random episodic depositional events, are integrated within the volumetric pyrite content and are not time-dependent.
In the submerged mineral portion of the soil, the amount of Fe within the soil would determine the magnitude of pyrite formation in the upper portion of the mineral soil where organic C and SO4 levels may be adequate for sulfidization. Before permanent submergence, the upland mineral soils experience extended periods of aerobic conditions during which sulfides are unstable. Thus, upland mineral soils generally do not contain pyrite and the pyrite content at time zero was considered to be zero g m-2. To investigate pyrite sequestration in the submerged mineral portion of the soils, the pyrite content (g m-2) within the upper 25 cm of the submerged mineral soils was plotted against time (Fig. 4) . The data indicate that during the first 50 to 150 yr, pyrite sequestration tends to increase with time. Beyond this initial period the data contained considerable random variability but did not show any specific trend with time.
|
The predicted pyrite content (g m-2) for the marsh soil was the sum of the pyrite content in the organic horizon and that of the submerged mineral portion of the soil. Using this procedure, the predictive pyrite models were two-step linear functions that followed sea-level rise (data are not shown). In general, the rate of pyrite sequestration is largely controlled by the rate of sea-level rise, availability of Fe, and pyrite formation in the submerged mineral portion. During the last 150 yr, the predicted rate of pyrite accumulation for Hell Hook Marsh was 7.08 ± 1.34 g m-2 yr-1 and for Cedar Creek Marsh was 7.38 ± 1.85 g m-2 yr-1, whereas before the last few hundred years, the predicted rate of pyrite sequestration for Hell Hook Marsh was 1.14 ± 0.17 g m-2 yr-1 and for Cedar Creek Marsh was 0.53 ± 0.06 g m-2 yr-1. To validate the pyrite model, an additional four pedons were used from each marsh that were not included in the model development. The ages of these pedons were also obtained using sea-level rise history of the Chesapeake Bay. Pyrite data from these pedons compared favorably with the predicted values at Hell Hook and Cedar Creek with coefficients of determination equal to 0.84 and 0.99, respectively (Fig. 5) .
|
In most mineral soils from the humid and semihumid regions, total S ranges from 20 to 600 mg kg-1, and organic S accounts for more than 95% of the total S (Tabatabai, 1982). Therefore, at time zero the newly formed marsh soils contained some initial, though small, quantity of organic S. Based on measurements of the A and O horizons of pedons from higher elevation, the organic S content at time zero was estimated to be 75 g m-2. Analysis of organic S sequestration in the mineral portion of the soil showed no specific pattern with pedon age (mean 118 g m-2). Given the magnitude of soil variability, the initial organic S content (75 g m-2) and the mean organic S content within the upper 25 cm of the submerged mineral portion of the soil (118 g m-2) are not considered different. Therefore, the initial organic S content within the upper 25 cm of the submerged mineral portion of the soil was taken to be the average of these values (97 g m-2).
Using the age–elevation relationship (Fig. 3) and the mean volumetric organic S content (1828 g m-3) of organic horizons, the predicted organic S content (g m-2) was calculated. The predicted organic S content (g m-2) for the marsh soil was the sum of the organic S content in the organic horizon and that of the submerged mineral portion of the soil (97 g m-2). The organic S models were also two-step linear functions related to rates of sea-level rise. During the last 150 yr, the predicted rate of organic S sequestration at Hell Hook Marsh was 4.02 ± 0.73 g m-2 yr-1 and at Cedar Creek Marsh was 4.57 ± 1.64 g m-2 yr-1; before the last 150 yr, the predicted rate at Hell Hook Marsh was 2.05 ± 0.31 g m-2 yr-1 and at Cedar Creek Marsh was 0.95 ± 0.11 g m-2 yr-1. Using the model-validation data sets, the predicted organic S levels were in good agreement with the observed values with r2 of 0.97 and 0.98 (Fig. 6) .
|
|
|
Marsh Soils and Potential Acidity
Because pyrite accumulates in significant quantities in tidal marsh soils, it constitutes potential acid-sulfate soils capable of producing sulfuric acid upon drainage. Models of pyrite sequestration in the marshes studies (Fig. 7) suggest that from 11000 to 19000 kg ha-1 of pyrite averaging 15000 ± 4300 kg ha-1 may be expected to accumulate during the next century. Future reclamation of these soils for agricultural or silvacultural purposes will be problematic and could have a dramatic impact on the environment due to acid-sulfate weathering. If one were to assume complete oxidation of S and Fe and hydrolysis of Fe, the potential acidity stored in these marshes during the next century ranges from 370 to 630 kmol ha-1.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
Received for publication April 13, 1998.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMethods and materialsResults and discussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. E. Vaughan, B. A. Needelman, P. J. A. Kleinman, and M. C. Rabenhorst Morphology and Characterization of Ditch Soils at an Atlantic Coastal Plain Farm Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 660 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. H. Hussein, M. C. Rabenhorst, and M. L. Tucker Modeling of Carbon Sequestration in Coastal Marsh Soils Soil Sci. Soc. Am. J., September 1, 2004; 68(5): 1786 - 1795. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. H. Hussein and M. C. Rabenhorst Modeling of Nitrogen Sequestration in Coastal Marsh Soils Soil Sci. Soc. Am. J., January 1, 2002; 66(1): 324 - 330. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.H. Hussein and M.C. Rabenhorst Modeling the Impact of Tidal Inundation on Submerging Coastal Landscapes of the Chesapeake Bay Soil Sci. Soc. Am. J., May 1, 2001; 65(3): 932 - 941. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.H. Hussein and M.C. Rabenhorst Tidal Inundation of Transgressive Coastal Areas: Pedogenesis of Salinization and Alkalinization Soil Sci. Soc. Am. J., March 1, 2001; 65(2): 536 - 544. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
